Volume 75, Issue 4, April 2020
Anaesthesia
Protection you can trust
for all your clinical needs
To find the right choice for your
clinical environment please visit:
www.intersurgical.co.uk/
info/filtrationandhumidification
Quality, innovation and choice
Anae_v75_i4_Cover.indd 1
Anaesthesia Journal of the Association of Anaesthetists Volume 75, Number 4, April 2020 pp 429–564
Peri-operative medicine, critical care and pain
Editorials
433 Restrictive blood transfusion – is less really more? A. Shah, S. J. Stanworth and
A. B. Docherty
438 Allogeneic blood and postoperative cancer outcomes: correlation or causation? E. A. Dickson
and A. G. Acheson
442 From variance to guidance for awake tracheal intubation M. F. Aziz and M. S. Kristensen
447 Fasten your seatbelts: innovation in regional anaesthesia is a bumpy ride E. R. Mariano and
K. El-Boghdadly
451 Abandoning inhalational anaesthesia S. M. White and C. L. Shelton
Original Articles
455 Red blood cell transfusion in surgery: an observational study of the trends in the USA from
2011 to 2016 A. T. Nordestgaard, L. S. Rasmussen, M. Sillesen, J. Steinmetz, A. I. Eid,
Your first K. Meier, H. M. A. Kaafarani and G. C. Velmahos
464 The association of allogeneic blood transfusion and the recurrence of hepatic cancer after
choice in surgical resection Y. H. Tai, H. L. Wu, M. S. Mandell, M. Y. Tsou and K. Y. Chang
472 The association of pre-operative anaemia with survival after orthotopic liver transplantation
P. Lichtenegger, J. Schiefer, A. Graf, G. Berlakovich, P. Faybik, D. M. Baron and
filtration and J. Baron-Stefaniak
479 The role of intra-operative cell salvage in patient blood management for revision hip
humidification arthroplasty: a prospective cohort study A. J. R. Palmer, T. D. Lloyd, V. N. Gibbs, A. Shah,
P. Dhiman, R. Booth, M. F. Murphy, A. H. Taylor, B. J. L. Kendrick and collaborators
487 Prophylactic phenylephrine and fluid co-administration to reduce spinal hypotension
during elective caesarean section in a resource-limited setting: a prospective alternating
intervention study A. S. Buthelezi, D. G. Bishop, R. N. Rodseth and R. A. Dyer
Breathing filters and HMEFs 493 A randomised controlled trial of shoulder block vs. interscalene brachial plexus block for
provide an effective barrier that ventilatory function after shoulder arthroscopy P. Rhyner, K. Kirkham, C. Hirotsu, A. Farron
prevent cross contamination and E. Albrecht
499 Hemidiaphragmatic paralysis following ultrasound-guided anterior vs. posterior
between patients, respiratory suprascapular nerve block: a double-blind, randomised control trial F. Ferré, M. Pommier,
breathing systems, equipment P. Laumonerie, A. Ferrier, R. Menut, L. Bosch, V. Balech, N. Bonnevialle and V. Minville
and the clinical environment. Guidelines
509 Difficult Airway Society guidelines for awake tracheal intubation (ATI) in adults
The Intersurgical range offers I. Ahmad, K. El-Boghdadly, R. Bhagrath, I. Hodzovic, A. F. McNarry, F. Mir, E. P. O’Sullivan,
A. Patel, M. Stacey and D. Vaughan
a choice of electrostatic and
pleated mechanical filters which Review Articles
have been independently tested 529 A systematic review and cost effectiveness analysis of reusable vs. single-use flexible
bronchoscopes J. M. Mouritsen, L. Ehlers, J. Kovaleva, I. Ahmad and K. El-Boghdadly
and proven to be highly efficient 541 Learning from the law. A review of 21 years of litigation for nerve injury following central
in preventing the passage of neuraxial blockade in obstetrics K. McCombe and D. G. Bogod
549 Correction
bacteria and viruses.
550 Correspondence
lnteract with us
www.intersurgical.co.uk
3/2/2020 4:55:09 PM
 Anaesthesia
Peri-operative medicine, critical care and pain
Journal of the Association of Anaesthetists www.anaesthetists.org

Editor-in-Chief Dr A Klein (Cambridge)

Senior Editor Professor A F Smith (Lancaster)
Editors Dr S Agarwal (Manchester), Dr C Bailey (London), Dr J Carlisle (Torbay), Dr L Duggan (Canada),
Dr K El-Boghdadly (London), Professor M G Irwin (Hong Kong), Professor B J Jenkins (Cardiff),
Dr M Kinsella (Bristol), Dr H Laycock (London), Dr A E Vercueil (London), Dr M Wiles (Sheffield)
Anaesthesia Reports Editor Dr K El-Boghdadly (London)
Social Media Editor Dr M Charlesworth (Manchester)
Anaesthesia Trainee Fellows Dr E Gilbert-Kawai (London) and Dr EJ Smith (London)
International Advisory Panel Dr E Albrecht (Switzerland), Dr B Biccard (South Africa), Dr N Chrimes (Australia), Dr I Cortinez (Chile),
Dr A Dennis (Australia), Dr G Djaiani (Canada), Dr R Dyer (South Africa), Dr N B Foss (Denmark),
Dr J Gadsden (USA), Dr M Heesen (Switzerland), Dr H Huitink (Netherlands), Dr M Kertai (USA), Dr J Kurata (Japan),
Professor R Landau (USA), Dr E Mariano (USA), Dr S Marshall (Australia), Professor H Otake (Japan),
Dr A Palanisamy (USA), Dr T Selak (Australia), Professor J-K Shim (South Korea)
Editorial Board Chair/President Dr K Ferguson; President Elect Dr M Nathanson; Honorary Secretary Dr T Meek;
Honorary Treasurer Dr R Gill; Honorary Membership Secretary Dr T Sheraton; Immediate Past Honorary Treasurer
Dr P Barker; Immediate Past Honorary Membership Secretary Dr U Misra; Honorary Secretary Elect Dr M Davies;
Trainee Committee representative Dr K Miller; Elected Association Board Members Dr M Patteril, Dr R Self; Chair
of Education Committee Professor W Fawcett
Statistical Advisor Dr S W Choi (Hong Kong)
Editorial Co-ordinator Miss R Gloag (anaesthesia@anaesthetists.org)
Production Editor Ms Q Li (anae@wiley.com)

Aims and Scope
Anaesthesia is the official journal of the Association
of Anaesthetists and is international in scope and
comprehensive in coverage. It publishes original,
peer reviewed articles on all aspects of general
and regional anaesthesia, intensive care and pain
therapy, including research on equipment. The
journal welcomes submissions of review articles,
special articles, statistical reviews and editorials,
and there is a correspondence section in each issue.
For submission instructions, subscription and all
other information visit: www.anaesthesia-journal.
org/guidance.
Anaesthesia accepts articles for Open Access
publication. Please visit https://authorservices.
wiley.com/author-resources/Journal-Authors/open-
access/index.html for further information.
Association Correspondence
Association of Anaesthetists, 21 Portland Place,
London W1B 1PY, UK.
Copyright and Copying
Copyright © 2020 Association of Anaesthetists. All
rights reserved. No part of this publication may be
reproduced, stored or transmitted in any form or by
any means without the prior permission in writing
from the copyright holder. Authorisation to copy
items for internal and personal use is granted by
the copyright holder for libraries and other users
registered with their local Reproduction Rights
Organisation (RRO), e.g. Copyright Clearance
Center (CCC), 222 Rosewood Drive, Danvers, MA
01923, USA (http://www.copyright.com), provided
the appropriate fee is paid directly to the RRO. This
consent does not extend to other kinds of copying
such as copying for general distribution, for
advertising or promotional purposes, for creating
new collective works or for resale. Permissions for
such reuse can be obtained using the RightsLink
“Request Permissions” link on Wiley Online
Library. Special requests should be addressed to:
permissions@wiley.com.
Information for subscribers
Anaesthesia is published monthly. Institutional
subscription prices for 2020 are: Print & Online:
US$2505 (US), US$2923 (Rest of World), €1726
(Europe), £1359 (UK). Prices are exclusive of tax. Asia-
Pacific GST, Canadian GST/HST and European VAT
will be applied at the appropriate rates. For more
information on current tax rates, please go to www.
wileyonlinelibrary.com/tax-vat. The price includes
online access to the current and all online back files to
January 1st 2016, where available. For other pricing
options, including access information and terms and
conditions, please visit www.wileyonlinelibrary.com/
access.
Delivery Terms and Legal Title
Where the subscription price includes print issues
and delivery is to the recipient’s address, delivery
terms are Delivered at Place (DAP); the recipient
is responsible for paying any import duty or taxes.
Title to all issues transfers FOB our shipping point,
freight prepaid. We will endeavour to fulfil claims
for missing or damaged copies within six months
of publication, within our reasonable discretion
and subject to availability.
Back issues
Single issues from current and recent volumes are
available at the current single issue price from: cs-
journals@wiley.com. Earlier issues may be obtained
from Periodicals Service Company, 351 Fairview
Avenue - Ste 300, Hudson, NY 12534, USA. Tel:
+1 518 822-9300, Fax: +1 518 822-9305, Email:
psc@periodicals.com.
View this journal online at: www.wileyonlinelibrary.
com/journal/anae.
ANAESTHESIA, (Print ISSN: 0003-2409; Online ISSN
1365-2044), is published monthly. US mailing agent:
Mercury Media Processing, LLC, 1850 Elizabeth
Avenue, Suite #C, Rahway, NJ 07065 USA. Periodical
postage paid at Rahway, NJ. Postmaster: Send all
address changes to ANAESTHESIA, John Wiley
& Sons Inc., C/O The Sheridan Press, PO Box 465,
Hanover, PA 17331 USA.
Disclaimer
The Publisher, the Association of Anaesthetists and
Editors cannot be held responsible for errors or any
consequences arising from the use of information
contained in this journal; the views and opinions
expressed do not necessarily reflect those of
the Publisher, the Association of Anaesthetists
and Editors, neither does the publication of
advertisements constitute any endorsement by the
Publisher, the Association of Anaesthetists and
Editors of the products advertised.
Publisher
Anaesthesia is published by John Wiley & Sons Ltd,
9600 Garsington Road, Oxford OX4 2DQ, UK. Tel:
+44 (0) 1865 776868; Fax: +44 (0) 1865 714591.
Printed in the UK by Hobbs the Printers Ltd.
Journal Customer Services
For ordering information, claims and any enquiry
concerning your journal subscription please go to
https://hub.wiley.com/community/support/
onlinelibrary or contact your nearest office.
Americas: Email: cs-journals@wiley.com; Tel: +1 781
388 8598 or +1 800 835 6770 (toll free in the USA &
Canada).
Europe, Middle East and Africa: Email: cs-
journals@wiley.com; Tel: +44 (0) 1865 778315.
Asia-Pacific: Email: cs-journals@wiley.com; Tel:
+65 6511 8000.
Japan: For Japanese speaking support, Email:
cs-japan@wiley.com.
Visit our Online Customer Help available in 7
languages at https://hub.wiley.com/community/
support/onlinelibrary.
Advertising: corporatesales@wiley.com
Commercial reprints: commercialreprints@wiley.
com
Wiley’s Corporate Citizenship initiative seeks to
address the environmental, social, economic, and
ethical challenges faced in our business and which
are important to our diverse stakeholder groups.
Since launching the initiative, we have focused on
sharing our content with those in need, enhancing
community philanthropy, reducing our carbon
impact, creating global guidelines and best practices
for paper use, establishing a vendor code of ethics,
and engaging our colleagues and other stakeholders
in our efforts. Follow our progress at www.wiley.com/
go/citizenship.
Wiley is a founding member of the UN-backed
HINARI, AGORA, and OARE initiatives. They are
now collectively known as Research4Life.org,
making online scientific content available free
or at nominal cost to researchers in developing
countries. Please visit Wiley’s Content Access –
Corporate Citizenship site: http://www.wiley.com/
WileyCDA/Section/id-390082.html.

anae_v75_i4_issueinfo.indd 1 2/20/2020 7:58:36 AM

Anaesthesia 2020 doi:10.1111/anae.14973

Editorial

Restrictive blood transfusion – is less really more?

A. Shah,1,2 S. J. Stanworth3,4 and A. B. Docherty5,6

1 NIHR Doctoral Research Fellow, 3 Associate Professor, Radcliffe Department of Medicine, University of Oxford, Oxford,
UK
2 Specialty Registrar, Nuffield Department of Anaesthesia, John Radcliffe Hospital, Oxford, UK
4 Consultant, Department of Haematology, Oxford University Hospitals NHS Foundation Trust, Oxford, UK
5 Wellcome Clinical Research Career Development Fellow, The Usher Institute, University of Edinburgh, Edinburgh, UK
6 Honorary Consultant, Department of Intensive Care Medicine, Royal Infirmary Edinburgh NHS Lothian, Edinburgh, UK
............................................................................................................................................................................................................................................................................................................
Correspondence to: A. Shah
Email: akshayshah@doctors.org.uk
Accepted: 6 December 2019
Keywords: guidelines; myocardial infarction; peri-operative risk; transfusion mortality: causes
This editorial accompanies an article by Nordestgaard et al., Anaesthesia 2019; https://doi.org/10.1111/anae.14900.
Twitter: @DocAShah, @SimonStanworth, @abdocherty79

In this month’s issue of Anaesthesia, Nordestgaard et al. residual confounding may still persist due to other
provide data on the reduction in the rates of peri-operative unidentified factors. An important omission was the data
red blood cell transfusion in the USA [1]. Following on cell salvage use that were not available to the
adjustment for several confounders, they observed a 45% authors. This was a significant development precisely
reduction in peri-operative red blood cell transfusions over during their study period. The Association of
a 6-year period, which equated to 356,679 fewer red blood Anaesthetists guidelines recommend using cell salvage
cell units. This reduction was not associated with an increase where blood loss greater than 500 ml is anticipated [3].
in peri-operative myocardial infarction, stroke or all-cause Cell salvage can reduce the rate of exposure of red
30-day mortality and resulted in potential cost savings of blood cell transfusion by a relative 38% [3]. Data on the
just over £200 million. use of tranexamic acid and whether hospitals had
These findings could have been due to many layers of established patient blood management programmes are
practice improvements including: better surgical techniques; also not shown. Tranexamic acid safely reduces blood
more laparoscopic surgery; better pre-operative man- loss and red blood cell transfusion requirements across
agement and optimisation of patients at risk of bleeding; and multiple surgical specialties and it is now included in
immortal time bias. Immortal time, sometimes also referred the World Health Organization list of ‘essential
to as survivorship bias, is a period of time in the follow-up medicines’ [4]. However, the optimal dose, route and
period of a study during which an outcome of interest (e.g. timing of administration of tranexamic acid is less clear.
death, stroke) cannot occur [2]. Bias is introduced when this Network meta-analyses are currently underway to
time period is either excluded from the analysis or identify an optimal peri-operative dosing regimen to
misclassified with regards to the treatment status [2]. The standardise clinical care [4]. Both tranexamic acid and
biggest change in the study was observed in patients cell salvage are now key components of a multimodal
undergoing orthopaedic surgery (64% decrease) and this patient blood management strategy. Recent evidence
might be related to factors such as tourniquet use, more suggests that comprehensive peri-operative patient
arthroscopic surgery, routine administration of tranexamic blood management programmes can reduce red blood
acid and increasing cell salvage use [3]. cell transfusion rates by up to 45%, reduce hospital
Despite adjusting for some of these factors in the length of stay and lower total number of postoperative
regression models, not all were accounted for and complications [5].

© 2020 Association of Anaesthetists 1

Anaesthesia 2020
Nordestgaard et al. also speculate that adherence to
restrictive transfusion practices, which have become
increasingly popular since 2010, may have contributed to a
reduction in red blood cell transfusion rates and they point
to a lower pre-transfusion haematocrit at the end of the
study period in Fig. 4c of their paper. However, the pre-
transfusion haematocrit was approximately 33% which
1
equates to a haemoglobin concentration of 110 g.l . This
could be classed as a liberal transfusion threshold.
Although these findings may provide some temporary
reassurance to blood transfusion services in terms of
reducing demand, modelling studies have suggested that
blood availability will need to increase again to meet the
demands of an ageing population [6]. There is already a
growing demand for universal blood groups (e.g. O
negative) and for minor blood groups that may be needed
to support patients requiring multiple transfusions (e.g.
sickle cell disease) [6]. In addition, blood services continue
to encounter problems in attracting and retaining young
donors [7]. These issues were highlighted in a recent James
Lind Alliance Blood Donation and Transfusion priority-
setting partnership exercise where the top three research
priorities were [7]:
1 What would encourage more people (especially Black
and ethnic minority groups or people with a rare blood
type) to donate blood?
2 How can health professionals be discouraged from
using blood inappropriately?
3 How can the wastage of donor blood be minimised?
Peri-operative cardiovascular events
and transfusion
Nordestgaard et al. report no increase in adverse clinical
outcomes associated with restrictive transfusion strategies,
such as myocardial ischaemia, stroke and all-cause 30-day
mortality. The rates observed are comparable with other
large administrative database studies [8]. However, it is
important to note that the event rates in all these studies are
likely to be an underestimate. Such databases are subject to
administrative coding errors, reporting bias and immortal
time bias. The authors do not provide data on the aetiology
of myocardial infarction or on the use of cardiovascular
medical therapies such as beta-blockers, statins and
antiplatelets in the peri-operative period and therefore the
effect, if any, of these interventions cannot be evaluated.
These interventions are likely to contribute to immortal time
bias. Various approaches have been described to eliminate
immortal time bias including using time-dependent Cox
regression analyses, only studying ‘survivors’ of the
2 © 2020 Association of Anaesthetists
Editorial
immortal period by moving the start of the follow-up to the
end of the immortal period or a time-matched, nested case-
control analysis of the study cohort [2].
It is now well recognised that the dominant pathology
in peri-operative myocardial infarction is myocardial oxygen
supply and demand mismatch, and not plaque rupture and
thrombosis. As a result, the majority of ischaemic events in
the peri-operative period are often silent and missed
clinically. A recent prospective, observational study in
critically ill patients found that more than 95% of myocardial
infarctions were undetected by clinical teams [9].
Restrictive transfusion strategies may not always be
indicated or appropriate. Evidence from systematic reviews
suggests that liberal transfusion strategies may reduce
myocardial infarction rates in patients with acute and
chronic cardiovascular disease, and even in those without
known cardiovascular disease [10]. Two recent pilot trials,
conducted in patients with traumatic brain injury [11] and in
those undergoing major vascular surgery [12], observed
harm in patients randomly assigned to lower red blood cell
transfusion thresholds and benefits at higher thresholds.
Participants randomly assigned to the liberal threshold
groups experienced lower mortality, less post-traumatic
vasospasm, improved neurological status [11] and fewer
major vascular complications [12]. Larger trials are
warranted to confirm or refute these early results. The
results of the ongoing myocardial ischemia and transfusion
randomised controlled trial (NCT02981407) are also
eagerly awaited. Although the focus of research so far has
been on cardiovascular events and mortality, the effect of a
restrictive or liberal transfusion strategy on renal function is
not yet known. Peri-operative acute kidney injury is
common, with estimates ranging from 5% to 40%, and
studies are beginning to investigate whether this is
influenced by transfusion strategies [13].
These conflicting results highlight some of the
limitations in major trials and large database studies
undertaken to date, such as: underpowered sub-group
analyses and fixed interventions masking divergent effect in
at-risk sub-groups; effect of immortal time bias and the
absence of important confounders in routine datasets; and
the lack of measurement of long-term patient-centred
outcomes, although data are beginning to emerge on these
[14].
Although restrictive transfusion thresholds lead to
fewer red blood cell transfusions, is haemoglobin or
haematocrit the best indicator for a transfusion? A recent
systematic review found that transfusion did not generally
improve tissue oxygenation or microcirculation in critically
ill patients, unless there was prior evidence of reduced
Editorial Anaesthesia 2020

tissue oxygenation or abnormal microcirculatory indices
using assessment tools such as near infrared spectroscopy,
spectral imaging and tissue microdialysis [15]. These
findings remained constant regardless of which assessment
method was used. Furthermore, recent evidence suggests
that blood donor characteristics (e.g. age, sex), collection
and processing methods and recipient characteristics, such
as age and body mass index, significantly influence changes
in haemoglobin concentrations after transfusion [16].
Further research into these areas may allow for a bespoke
approach towards transfusion in the future, but this may also
have an impact on blood donation strategies.
Another unintended consequence of widespread
adoption of restrictive transfusion practices is that more
patients are likely to be discharged from hospital with
anaemia. A recent large retrospective study of over
445,000 patients showed that the prevalence of moderate
anaemia (haemoglobin between 70 and 100 g.l
increased from 20% to 25% over a 4-year period [17]. In
addition, the proportion of patients whose anaemia had
resolved within 6 months of hospital discharge decreased
from 42% to 34%. Although this was not associated with
increased rehospitalisation or mortality, the impact of
anaemia on physical function and health-related quality of
life was not investigated. The relationship between
persisting, and often untreated, anaemia and fatigue and
poor quality of life has been described in survivors of
critical illness and trials are currently underway to address
this [18].
Guidelines, guidelines and even more
guidelines
In their discussion, Nordestgaard et al. speculate that
adherence to restrictive transfusion guidelines may have
contributed to the observed reduction in red blood cell
transfusion rates. However, some limitations need to be
considered when addressing the role of guidelines.
Guidelines are one of the approaches applied in clinical
practice to bridge the gap between actual and
recommended evidence-based practice. Other strategies
include audit, feedback, quality improvement projects and,
increasingly for transfusion practice, computerised decision
support systems. Traditionally, guidelines may have been
consensus statements driven by the opinions of experts, but
increasingly guidelines are developed with higher
methodological rigour. The emergence of the grading
of recommendations, assessment, development and
1
) evaluation (GRADE) approach has also strengthened quality
but many still continue to suffer from methodological
weaknesses [19].
There has been an exponential increase in the number
of clinical guidelines over the past three decades. The
current total in the International Guidelines Library (https://
g-i-n.net) stands at 6859, and this is likely to be an
underestimate of the actual number. Relevant to transfusion
medicine, a recent systematic review identified 30
guidelines related to the transfusion of red blood cells that
have been developed since 2006 [20]. At least ten of these
are relevant to anaesthesia, peri-operative medicine or

Table 1 Examples of inconsistent recommendations in transfusion guidelines relevant to peri-operative medicine.

Strength of Quality of
Guideline group Recommendation recommendation evidence
Association of Apply restrictive transfusion threshold (Hb 70–80 g.l 1) depending Not reported Not reported
Anaesthetists 2016 on patient characteristics and haemodynamics.
[23] Uncertainty remains for patients with ischaemic heart disease, Not reported Not reported
including acute coronary syndrome and after cardiac surgery, and
higher threshold (Hb < 80 g.l 1) may be more appropriate in such
circumstances
American Association For patients undergoing orthopaedic surgery or cardiac surgery Strong Moderate
of Blood Banks 2016 and those with pre-existing cardiovascular disease, the AABB
[24] recommends a restrictive RBC transfusion threshold
(Hb < 80 g.l 1)
Frankfurt Patient The panel recommended a restrictive transfusion threshold Conditional Moderate
Blood Management (Hb < 80 g.l 1) in patients with hip fracture and cardiovascular
Consensus 2019 [25] disease or other risk factors
Expert panel suggests further research in patients undergoing non-
cardiac or non-orthopaedic surgery
The panel recommended a restrictive RBC transfusion threshold Strong Moderate
(Hb < 75 g.l 1) in patients undergoing cardiac surgery
Hb, haemoglobin concentration; RBC, red blood cell.

© 2020 Association of Anaesthetists 3

Anaesthesia 2020
critical care, and the timeline of the publication of these
guidelines is demonstrated in the Supporting Information
(Fig. S1) of the paper by Nordestgaard et al. [1].
Despite guideline development often being carried out
by reputable professional bodies, concerns regarding their
usefulness, and even trustworthiness, remain [19]. The
development of guidelines on the same topic by several
different organisations can leave the reader confused,
particularly if there are different recommendations based
on the same primary evidence, as highlighted in Table 1. It
leads to unnecessary duplication of effort and is a waste of
time and money, especially when the cost of developing a
guideline is thought to be more than $100,000, which
approximately equates to £77,000 or €90,000 [20].
Furthermore, depending on how quickly the relevant field is
evolving, guidelines can quickly become outdated.
The final recommendations of any guideline are, not
infrequently, a reflection of the composition of individuals on
the working party, including (sometimes heated) discussion
about recommendations in the face of limited, if any, high-
quality evidence. Panels may be unbalanced and include a
disproportionate number of content experts, each with their
own prejudices, bias and conflicts of interest. This is especially
pertinent for guidelines involving pharmaceuticals. Patient
representatives and healthcare economists are often not
included and as a result patient preferences and cost-
effectiveness analysis may not be addressed. Consensus may
be achieved through voting (such as a Delphi) process but the
precise wording of the question can influence interpretation
of the evidence and subsequent decision making. These
processes are often not clear enough to the clinician reading
the final report. It is therefore not surprising that strong
recommendations are made in the presence of weak
evidence and there is disagreement with similar guidelines
written by other groups (Table 1). The widespread adoption
of recommendations based on weak evidence can
accidentally lead to patient harm, reinforce traditional
practices and reduce the drive for research. Many of the
issues discussed above have been highlighted by the
example of the Surviving Sepsis Campaign guidelines, which
continue to divide the intensive care community. Some have
called for them to be retired completely [21].
Collaborative models, where a single, diverse and
multidisciplinary panel takes ownership to develop
statements, have the potential to use standardised
guideline development processes and promote
transparency. In order to quickly incorporate new evidence
into pre-existing guidelines, Intensive Care Medicine and
the British Medical Journal Rapid Recommendation Group
have recently introduced the concept of rapid practice
4 © 2020 Association of Anaesthetists
Editorial
guidelines [22]. Here, researchers perform a timely
systematic review and in parallel, a panel including
methodologists, researchers, clinicians and patients will
choose the most important outcomes. The systematic
review and evidence will be assessed using the GRADE
approach, and recommendations for practice will be
generated. This would then be submitted to the relevant
journal for rapid peer review and publication.
In conclusion, Nordestgaard et al. should be
congratulated on analysing a large, complex national
database and providing important data for blood
transfusion services globally. The limitations of such
retrospective studies are well understood but the authors
have raised important questions on clinically diagnosed
peri-operative cardiovascular events, while also providing
us with an opportunity to discuss the strengths and
limitations of clinical guidelines.
Acknowledgements
AS is being supported by an NIHR Doctoral Research
Fellowship (DRF-2017-10-094). No competing interests
declared.
References
1. Nordestgaard AT, Rasmussen LS, Sillesen M, et al. Red
blood cell transfusion in surgery: an observational study of
trends in the USA from 2011 to 2016. Anaesthesia 2019;
74. https://doi.org/10.1111/anae.14900. [Epub ahead of
print].
2. Ho AM, Dion PW, Ng CS, Karmakar MK. Understanding
immortal time bias in observational cohort studies. Anaesthesia
2013; 68: 126–30.
3. Klein AA, Bailey CR, Charlton AJ, et al. Association of
Anaesthetists guidelines: cell salvage for peri-operative blood
conservation 2018. Anaesthesia 2018; 73: 1141–50.
4. Gibbs VN, Champaneira R, Palmer A, Doree C, Estcourt LJ.
Pharmacological interventions for the preventions of bleeding
in people undergoing elective hip or knee surgery: a systematic
review and network meta-analysis. Cochrane Database of
Systematic Reviews 2019; 3: CD013295.
5. Althoff FC, Neb H, Herrmann E, et al. Multimodal patient blood
management program based on a three-pillar strategy: a
systematic review and meta-analysis. Annals of Surgery 2019;
269: 794–804.
6. Williamson LM, Devine DV. Challenges in the management of
blood supply. Lancet 2013; 381: 1866–75.
7. Hibbs SP, Brunskill SJ, Donald GC, Saunders HD, Murphy MF.
Setting priorities for research in blood donation and
transfusion: outcome of the James Lind Alliance priority-setting
partnership. Transfusion 2019; 59: 574–81.
8. Smilowitz NR, Gupta N, Ramakrishna H, Guo Y, Berger JS,
Bangalore S. Perioperative major adverse cardiovascular and
cerebrovascular events associated with noncardiac surgery.
Journal of the American Medical Association Cardiology 2017;
2: 181–7.
9. Docherty AB, Alam S, Shah AS, et al. Unrecognised myocardial
infarction and its relationship to outcome in critically ill patients
with cardiovascular disease. Intensive Care Medicine 2018; 44:
2059–69.
Editorial
10. Cortes-Puch I, Wiley BM, Sun J, et al. Risks of restrictive red
blood cell transfusion strategies in patients with cardiovascular
disease (CVD): a meta-analysis. Transfusion Medicine 2018; 28:
335–45.
11. Gobatto ALN, Link MA, Solla DJ, et al. Transfusion
requirements after head trauma: a randomized feasibility
controlled trial. Critical Care 2019; 23: 89.
12. Moller A, Nielsen HB, Wetterslev J, et al. Low vs high
hemoglobin trigger for transfusion in vascular surgery:
a randomized clinical feasibility trial. Blood 2019; 133:
2639–50.
13. Garg AX, Shehata N, McGuiness S, et al. Risk of acute kidney
injury in patients randomized to a restrictive versus liberal
approach to red blood transfusion in cardiac surgery: a
substudy protocol of the transfusion requirements in cardiac
surgery 3 noninferiority trial. Canadian Journal of Kidney Health
and Disease 2018; 5: 2054358117749532.
14. Carson JL, Sieber F, Cook DR, et al. Liberal versus restrictive
blood transfusion strategy: 3-year survival and cause of death
results from the FOCUS randomised controlled trial. Lancet
2015; 385: 1183–9.
15. Nielsen ND, Martin-Loeches I, Wentowski C. The effects of red
blood cell transfusion on tissue oxygenation and the
microcirculation in the intensive care unit: a systematic review.
Transfusion Medicine Reviews 2017; 31: 205–22.
16. Roubinian NH, Plimier C, Woo JP, et al. Effect of donor,
component, and recipient characteristics on hemoglobin
increments following red blood cell transfusion. Blood 2019;
134: 1003–13.
17. Roubinian NH, Murphy EL, Mark DG, et al. Long-term outcomes
among patients discharged from the hospital with moderate
© 2020 Association of Anaesthetists
Anaesthesia 2020
anemia: a retrospective cohort study. Annals of Internal
Medicine 2019; 170: 81–9.
18. Shah A, Marian I, Dutton SJ, et al. Intravenous iron to
treat anaemia following critical care (INTACT): a protocol
for a feasibility randomised controlled trial. Journal of the
Intensive Care Society 2019. https://doi.org/10.1177/
1751143719870080. [Epub ahead of print].
19. Iannone P, Costantino G, Montano N, et al. Wrong guidelines:
how to detect them and what to do in the case of flawed
recommendations. Evidence Based Medicine 2017; 22: 4–8.
20. Pavenski K, Stanworth S, Fung M, et al. Quality of evidence-
based guidelines for transfusion of red blood cells and plasma:
a systematic review. Transfusion Medicine Reviews 2018; 32:
135–43.
21. Marik PE, Farkas JD, Spiegel R, et al. Should the surviving
sepsis campaign guidelines be retired? Yes Chest 2019; 155:
12–14.
22. Alhazzani W, Moller MH, Belley-Cote E, Citerio G. Intensive care
medicine rapid practice guidelines (ICM-RPG): paving the road
of the future. Intensive Care Medicine 2019; 45: 1639–41.
23. Klein AA, Arnold P, Bingham RM, et al. Association of
Anaesthetists guidelines: the use of blood components and
their alternatives 2016. Anaesthesia 2016; 71: 829–42.
24. Carson JL, Guyatt G, Heddle NM, et al. Clinical practice
guidelines from the AABB: red blood cell transfusion
thresholds and storage. Journal of the American Medical
Assocation 2016; 316: 2025–35.
25. Mueller MM, Van Remoortel H, Meybohm P, et al. Patient blood
management: recommendations from the 2018 frankfurt
consensus conference. Journal of the American Medical
Association 2019; 321: 983–97.
5
Anaesthesia 2019 doi:10.1111/anae.14965

Editorial

Allogeneic blood and postoperative cancer outcomes:

correlation or causation?
E. A. Dickson1 and A. G. Acheson2

1 Clinical Research Fellow, National Institute for Healthcare Research Biomedical Research Centre, Nottingham Digestive
Diseases Centre, University of Nottingham, Nottingham, UK
2 Associate Professor, Department of Surgery, University of Nottingham and Nottingham University Hospitals,
Nottingham, UK
.................................................................................................................................................................
Correspondence to: E. A. Dickson
Email: edward.dickson@nhs.net
Accepted: 26 November 2019
Keywords: bias; blood transfusion; cancer outcomes; observational study; trial methodology
This editorial accompanies an article by Tai et al., Anaesthesia 2020; https://doi.org/10.1111/anae.14862.
Twitter: @Ed_Dickson, @austin_acheson

Controversy exists relating to the impact of peri-operative
allogeneic blood transfusion on cancer survival and
recurrence across a number of surgical specialties [1–3].
Although allogeneic blood transfusion may, in certain
circumstances, be a life-saving treatment, there is an
increasing awareness of the negative effect, both short- and
long-term, it may have on recipients. Intra-operative blood
loss can be common in hepatobiliary surgery and perhaps
none more so than in resection of hepatocellular carcinoma.
In this month’s issue of Anaesthesia, Tai et al. present a
retrospective analysis of peri-operative transfusion in 1469
patients who underwent hepatocellular carcinoma
resection over a 12-year period [4]. In this study, the authors
aimed to investigate a proposed link between transfusion,
cancer survival and disease recurrence. Firstly, it must be
acknowledged that the circumstances surrounding peri-
operative transfusion are subject to influence by a number
of factors. As such, when investigating any link between
transfusion and adverse outcomes, the issue of causality is
of utmost importance. Cancer recurrence rates are
influenced by a number of poor prognostic factors such as
tumour characteristics and spread; patient comorbidities;
and operative difficulty. Furthermore, the development of
pre-operative anaemia itself is also subject to confounding
variables such as a more aggressive tumour morphology. It
follows that there is often a strong bias in these
characteristics among patients who are transfused.
Therefore, does the frequency of transfusion increase as a
consequence of these pre-existing factors or are there
mechanisms that underpin a direct link to cancer recurrence
and survival?
Much of the current evidence draws from uncontrolled
retrospective studies which lack standardisation of
transfusion practice and are fraught with both known and
unknown confounders for cancer survival. Logically, these
issues could be addressed with a randomised trial but such
a study would raise significant ethical issues in the context of
a potentially life-saving treatment. In the present study, Tai
et al. attempt to replicate the design of a randomised trial
by employing the use of propensity score matching [4]. This
has allowed them to draw balanced groups of those who
did and did not receive peri-operative transfusion which
they define as transfusion intra-operatively or up to 7 days
postoperatively. To further strengthen their methodology,
they also used inverse probability of treatment weighting to
adjust for any clinicopathological bias. Using these
methods, the authors reported that among some 1469
patients, those who were transfused had an increased risk of
recurrence by an estimated 30% and mortality by 90%
based on adjusted hazard ratios.
Transfusion volume
In addition to proposing a causal relationship, Tai et al. also
hypothesised that the number of transfused red cell units
can influence cancer recurrence risk. Using the technique
of restricted cubic splines, they aimed to assess for any

© 2019 Association of Anaesthetists 1

Anaesthesia 2019
non-linear, dose–response association between transfusion
volume and recurrence. Splines have their origins tied to the
engineering world where draftsmen would use weights or
‘knots’ to cantilever a smooth architectural line between two
points. Statistically, this technique allows for the modelling
of an even graphical curve between specific data points
(knots) – in this case, number of units of red cells transfused.
By presenting this non-linear relationship, Tai et al.
observed hazard ratios peaking at the 5–6 unit threshold in
both adjusted and unadjusted models. The effect of
transfusion volume on adverse events has also been
examined in other retrospective studies. In a series of
patients undergoing curative pancreaticoduodenectomy,
Yao et al. [5] concluded that transfusing 3 units of red cells
or more was associated with worse 3-year and 5-year
survival with a relative risk of 2.082 (1.048–4.135, p = 0.036)
compared with recipients of ≤ 2 units. Dose-related effects
have also been demonstrated in a meta-analysis of
colorectal studies where transfusion of just 1–2 units led to a
40% increase in cancer recurrence risk and > 5 units
doubled this risk [6].
Timing of transfusion
In their study, Tai et al. define peri-operative transfusion as
occurring during or up to 7 days after surgery [4]. Here, the
definition of ‘peri-operative’ is highly important and should
be considered when interpreting the results of this study.
There is good evidence that pre-operative anaemia can
lead to adverse postoperative outcomes [7] and efforts to
correct anaemia before surgery should now be common
practice. It is, therefore, important to acknowledge that, in
clinical practice, blood transfusions can fall outside this 7-
day window either pre- or postoperatively and this may, in
turn, modify the risk for certain postoperative outcomes.
Among colorectal cancer patients, one recent meta-analysis
reported that although cancer recurrence risks were highest
with intra- or postoperative transfusion, a 50% increase was
still observed with pre-operative transfusion [6]. Therefore,
perhaps an agreed consensus on the ‘peri-operative’ period
should be sought to allow comparison with similar studies
and to facilitate meta-analysis in this complex area.
Immunological effects
There is general consensus that the adverse survival
outcomes following transfusion relate to its transient
immunosuppressive effects. There has been an awareness
of transfusion-related immunomodulation (TRIM) for some
time, but the exact mechanism remains poorly understood.
Early evidence stemmed from reports of improved renal
allograft survival due to the immunosuppressive effects of
2 © 2019 Association of Anaesthetists
Editorial
transfusion in the early 1970s [8]. By 1981, the first data from
animal studies emerged demonstrating depressed
lymphocyte function and increased tumour growth
following transfusion which raised concerns for cancer
recurrence [9]. This was shortly followed by the first clinical
data supporting a link between transfusion and colorectal
tumour recurrence [10]. At the time, this link to an
immunosuppressive mechanism was also supported by
observations in Crohn’s patients where transfusion
recipients had lower disease recurrence rates compared
with non-transfused individuals [11]. Much of the earlier
understanding of TRIM ascribed this mechanism to
leucocytes present in packed red cells. However, despite
leukoreduction now being implemented as standard
practice, there still remains an association between
transfusion and adverse clinical outcomes. In fact,
randomised trials have demonstrated no difference in
survival or cancer recurrence between groups receiving
blood with or without allogeneic leucocytes [12]. The
current understanding of TRIM links transfusion to a
reduction in T-cell activity, inhibition of the
immunoregulatory cytokine IL-2 and the release of
immunosuppressive prostaglandins [13]. Moreover, after
transfusion, immune function can further suffer due to
decreased maturation of dendritic cells and a reduction in
their antigen-presenting activity [14]. Additional
mechanisms have also been proposed by Proctor et al.,
who demonstrated that packed red cells deplete
extracellular arginine, an amino acid responsible for T-cell
growth and differentiation [15]. Furthermore, the authors
also showed that prolonged storage of packed red cells
may increase the activity of amino acid transporters
responsible for this mechanism.
Despite acknowledging the immunosuppressive
effects of transfusion, it is important to appreciate that
anaemia itself can also be linked to cancer growth and
recurrence. For example, the effects of relative tissue
hypoxia as a consequence of anaemia can render tumour
cells more aggressive [16] and pre-treatment haemoglobin
levels have been associated with certain adverse tumour
characteristics [17]. Anaemia also leads to the increased
expression of transcription factors such as HIF-1 [18] which
may confer certain selective advantages among tumour
cells thus promoting cancer progression. This further
highlights the importance of correcting anaemia by
alternatives to transfusion, such as iron, where possible.
However, the replacement of packed red cells may only play
a part in the negative sequelae of transfusion. Indeed, Tai
et al. acknowledge that one limitation to their study is a lack
of data on the use of other blood products such as fresh
Editorial
frozen plasma (FFP) and platelet concentrate. Again, due to
its immunomodulatory effects, the administration of FFP has
been subject to investigation. In cases of hepatocellular
carcinoma, some authors claim a link between FFP
administration and worsening overall survival [19], although
others evaluating disease-free survival have demonstrated
no effect [20].
Storage of packed red cells
There are other additional factors, perhaps outside the
control of Tai et al., which may have influenced post-
transfusion disease recurrence rates. The storage duration
of blood products has previously drawn attention as a
potential contributor to such outcomes. During packed red
cell storage, certain molecular and biochemical changes
occur leading to the so called ‘storage lesion’. These may
include angiogenic or oncogenic factors and concerns
have been raised that in some circumstances this could
promote cancer recurrence among recipients. Experimental
observations by Benson et al. [21] have shown that packed
red cells stored for up to 42 days may carry elevated levels
of certain pro-cancer cytokines, although pre-storage
leukodepletion did lead to their reduction.
retrospective studies have also examined the effect of red
cell storage on cancer recurrence in the clinical setting.
Kekre et al. reviewed 27,000 cancer patients treated by
means of chemoradiotherapy or surgery of whom 1929 had
been transfused [22]. Although the authors found
significantly higher levels of cancer recurrence among
transfused patients, there was no association with red cell
storage duration (≤ 14 days vs. 14–28 days vs. > 28 days).
Alternatives to transfusion
Given the evidence presented by Tai et al. and the
supporting findings from previous studies, it follows that
alternatives to allogeneic blood transfusion should be
evaluated. Fundamentally, efforts to minimise blood loss
during resection of hepatocellular carcinoma must be
employed, in combination with the other measures
encompassed by the now familiar concept of patient blood
management (PBM). Patient blood management combines
the actions of pre-operative haemoglobin optimisation,
standardisation of transfusion practice and blood-sparing
techniques with the aim of improving outcomes in
transfusion medicine. Studies of PBM have now been
undertaken at an epidemiological level. Meybohm et al.
analysed some 129,719 patients across four German
teaching hospitals, discharged between July 2012 and June
2015, comparing those who had undergone operation at a
time of current PBM practice (75,206 patients) with those
© 2019 Association of Anaesthetists
Anaesthesia 2019
treated pre-PBM (54,513 patients) [23]. Not only was
non-inferiority to standard practice demonstrated but a
reduction in transfusion was observed among the PBM
group. Regarding hepatocellular carcinoma resection,
there are a number of procedure-specific blood-sparing
measures reported in the literature. These include variations
in surgical technique or equipment, the use of haemostatic
agents and the modulation of central venous pressure,
although many of these methods carry low-quality evidence
in recent meta-analysis [24]. In cases of large blood loss, one
alternative to transfusion may be the use of autologous
blood. Controversies relating to autologous transfusion in
the context of cancer have centred on the notion that
surgical resection may increase the volume of circulating
tumour cells at time of operation leading to the potential for
systemic dissemination with auto-transfusion. However, at
present these theories remain unsupported by robust
evidence. In fact, studies have found autologous transfusion
to be safe in the setting of hepatectomy for colorectal liver
metastasis, resection of hepatocellular carcinoma and liver
transplant for hepatocarcinoma with no impact on survival
or cancer recurrence risk [25].
Large
Future directions
Despite the issues previously raised, there may also be
confounding peri-operative factors contributing to
postoperative cancer recurrence which are yet to be fully
understood; for instance, the potential link between certain
anaesthetic or analgesic agents and cancer recurrence [26].
It follows that further work in this field is required to reveal
other influential peri-operative variables apart from
transfusion which may have been unaccounted for in
previously reported studies. It is, therefore, logical to
suggest that future prospective studies may help to address
some of the controversies relating to peri-operative
transfusion. However, it must be recognised that regardless
of any mechanism by which transfusion may be associated
with worse cancer-related outcomes, the importance of
appropriate PBM cannot be overstated. When combining
factors linked to allogeneic blood such as the cost, the risk
of infection, allo-immunisation and adverse postoperative
outcomes, there must be a strong focus on reducing
transfusion. These measures may include the correction of
pre-operative anaemia, operative and anaesthetic
techniques to limit blood loss and perhaps the use of
autologous transfusion where appropriate.
Acknowledgements
AA’s research department has received grant support from
Syner-Med (UK), Vifor Pharma (Switzerland) and
3
Anaesthesia 2019
Pharmacosmos A/S (Denmark). He has received honoraria
and travel support for consulting or lecturing from Ethicon
Endosurgery (UK), Johnson and Johnson Ltd (UK), Olympus
(UK) and Vifor Pharma (Switzerland). ON has received
honoraria and travel support for consulting from
Pharmacosmos A/S, Denmark. No external funding or other
competing interests declared.
References
1. Liu L, Wang Z, Jiang S, et al. Perioperative allogenenic blood
transfusion is associated with worse clinical outcomes for
hepatocellular carcinoma: a meta-analysis. PLoS ONE 2013; 8:
e64261.
2. Wu H-L, Tai Y-H, Lin S-P, Chan M-Y, Chen H-H, Chang K-Y. The
impact of blood transfusion on recurrence and mortality
following colorectal cancer resection: a propensity score
analysis of 4,030 patients. Scientific Reports 2018; 8: 13345.
3. Acheson AG, Brookes MJ, Spahn DR. Effects of allogeneic red
blood cell transfusions on clinical outcomes in patients
undergoing colorectal cancer surgery: a systematic review and
meta-analysis. Annals of Surgery 2012; 256: 235–44.
4. Tai YH, Wu HL, Mandell MS, Tsou MY, Chang KY. The association
between allogeneic blood transfusion with recurrence of hepatic
cancer after surgical resection. Anaesthesia 2020; 75: https://doi.
org/10.1111/anae.14862.
5. Yao HS, Wang Q, Wang WJ, Hu ZQ. Intraoperative allogeneic
red blood cell transfusion in ampullary cancer outcome after
curative pancreatoduodenectomy: a clinical study and meta-
analysis. World Journal of Surgery 2008; 32: 2038–46.
6. Amato A, Pescatori M. Perioperative blood transfusions and
recurrence of colorectal cancer. Cochrane Database of
Systematic Reviews 2006; 1: CD005033.
7. Leichtle SW, Mouawad NJ, Lampman R, Singal B, Cleary RK.
Does preoperative anemia adversely Aaffect colon and rectal
surgery outcomes? Journal of the American College of
Surgeons 2011; 212: 187–94.
8. Opelz G, Sengar DP, Mickey MR, Terasaki PI. Effect of blood
transfusions on subsequent kidney transplants. Transplantation
Proceedings 1973; 5: 253–9.
9. Francis DM, Shenton BK. Blood transfusion and tumour growth:
evidence from laboratory animals. Lancet 1981; 2: 871.
10. Burrows L, Tartter P. Effect of blood transfusions on colonic
malignancy recurrent rate. Lancet 1982; 2: 662.
11. Williams JG, Hughes LE. Effect of perioperative blood transfu-
sion on recurrence in Crohn’s disease. Lancet 1989; 334: 131–3.
12. van dWL, Brand A, Houbiers JG, Klein KW, Hermans J, van de
Velde C. Perioperative blood transfusions, with or without
allogeneic leucocytes, relate to survival, not to cancer
recurrence. British Journal of Surgery 2001; 88: 267–72.
4 © 2019 Association of Anaesthetists
Editorial
13. Vamvakas EC, Blajchman MA. Transfusion-related immunomod-
ulation (TRIM): an update. Blood Reviews 2007; 21: 327–48.
14. Cata JP, Wang H, Gottumukkala V, Reuben J, Sessler DI.
Inflammatory response, immunosuppression, and cancer
recurrence after perioperative blood transfusions. British
Journal of Anaesthesia 2013; 110: 690–701.
15. Procter LD, Meier CF, Hamilton C, et al. y+ cationic amino acid
transport of arginine in packed red blood cells. Journal of
Surgical Research 2013; 179: e183–7.
16. H€ockel M, Vaupel P. Tumor Hypoxia: definitions and current
clinical, biologic, and molecular aspects. Journal of the
National Cancer Institute 2001; 93: 266–76.
17. Khan AA, Klonizakis M, Shabaan A, Glynne-Jones R.
Association between pretreatment haemoglobin levels and
morphometric characteristics of the tumour, response to
neoadjuvant treatment and long-term outcomes in patients
with locally advanced rectal cancers. Colorectal Disease 2013;
15: 1232–7.
18. Semenza GL. HIF-1 and tumor progression: pathophysiology
and therapeutics. Trends in Molecular Medicine 2002; 8: S62–7.
19. Shiba H, Ishida Y, Haruki K, et al. Negative impact of fresh-
frozen plasma transfusion on prognosis after hepatic resection
for liver metastases from colorectal cancer. Anticancer
Research 2013; 33: 2723–8.
20. Tomimaru Y, Wada H, Marubashi S, et al. Fresh frozen plasma
transfusion does not affect outcomes following hepatic
resection for hepatocellular carcinoma. World Journal of
Gastroenterology 2010; 16: 5603–10.
21. Benson DD, Beck AW, Burdine MS, Brekken R, Silliman CC,
Barnett CC. Accumulation of pro-cancer cytokines in the
plasma fraction of stored packed red cells. Journal of
Gastrointestinal Surgery 2012; 16: 460–8.
22. Kekre N, Mallick R, Allan D, Tinmouth A, Tay J. The impact of
prolonged storage of red blood cells on cancer survival. PLoS
One 2013; 8: e68820–e20.
23. Meybohm P, Herrmann E, Steinbicker AU, et al. Patient blood
management is associated with a substantial reduction of red
blood cell utilization and safe for patient’s outcome. Annals of
Surgery 2016; 264: 203–11.
24. Moggia E, Rouse B, Simillis C, et al. Methods to decrease blood
loss during liver resection: a network meta-analysis. Cochrane
Database of Systematic Reviews 2016; 10: CD010683.
25. Kang R, Seath BE, Huang V, Barth RJ Jr. Impact of autologous
blood transfusion on survival and recurrence among patients
undergoing partial hepatectomy for colorectal cancer liver
metastases. Journal of the American College of Surgeons 2019;
228: 902–8.
26. Wall T, Sherwin A, Ma D, Buggy DJ. Influence of perioperative
anaesthetic and analgesic interventions on oncological
outcomes: a narrative review. British Journal of Anaesthesia
2019; 123: 135–50.
Anaesthesia 2019 doi:10.1111/anae.14947

Editorial

From variance to guidance for awake tracheal intubation

M. F. Aziz1 and M. S. Kristensen2

1 Professor, Department of Anesthesiology and Peri-operative Medicine, Oregon Health and Science University,
Portland, OR, USA
2 Consultant, Department of Anaesthesia, Centre of Head and Orthopaedics, Rigshospitalet, University Hospital of
Copenhagen, Denmark
............................................................................................................................................................................................................................................................................................................
Correspondence to: M. F. Aziz
Email: azizm@ohsu.edu
Accepted: 11 November 2019
Keywords: awake intubation; difficult airway; flexible scope; intubation; videolaryngoscopy
This editorial accompanies an article by El-Boghdadly et al., Anaesthesia 2019; https://doi.org/10.1111/anae.14904.
Twitter: @DrMikeAziz

Awake tracheal intubation remains the gold standard for The most notable feature of these guidelines is the
difficult airway management. Since the advent of the flexible relative absence of the highest level evidence behind each
fibreoptic endoscope, anaesthetists have learned to use this recommendation. Although some guidance is supported
technique to safely secure tracheal tubes and avoid critical by grade-A evidence from randomised trials, many of the
respiratory events. Newer modalities for asleep airway recommendations are simply supported by the opinions of
management such as videolaryngoscopes and supraglottic these experts (grade D). These guidelines stand in contrast
airway devices have further advanced safe airway to many other clinical care guidelines that are only
management but have not replaced awake tracheal formulated once the strength of evidence is so compelling
intubation. Since the 1980s, we have learned to perform this that there is an urgent need to standardise care. These
technique from those that taught us or from airway findings highlight that these guidelines should not
management courses. However, there has been no represent standard care for every clinician, but rather a
guidance from airway management guidelines on how this consideration for those unfamiliar with how to approach a
critical technique should be performed. In this issue of step during awake tracheal intubation.
Anaesthesia, Ahmad et al., representing the Difficult Airway So, why should we offer any guidance at all when the
Society, offer much needed guidelines for this approach [1]. evidence is limited or non-existent? Certainly, many
The authors came to this guidance via a robust process. providers have learned to perform this technique and do
They followed recommended standards for conducting not need these guidelines to safely care for their patients.
systematic reviews. In doing so, they evaluated the existing However, standardised care does support patient safety.
evidence and appropriately graded each recommendation Adherence to clinical practice guidelines is associated with
based on the strength of that evidence. They met multiple reduced mortality, reduced cost of care and hospitalisation
times and utilised the Delphi technique to bring value and when studied for cancer, cardiovascular or respiratory care
uniformity to each recommendation. Uniquely, they also [2–5]. Bringing uniformity to a clinical approach that carries
engaged patients for input. This engagement is something patient risk offers the advantage of making each step and
often absent from anaesthetic guidance but vitally subsequent outcome predictable. When the entire clinical
important to this procedure because the patient experience care team has a robust understanding of equipment needs,
during awake tracheal intubation cannot be undervalued. set up and procedural steps, the most efficient help is
Certainly, patients also want the safest approach to airway offered as it is familiar to all. These team members are also
management, but no patient appreciates discomfort. more likely to identify abnormalities because we can clearly
Recognising this important patient perspective adds value identify what is ‘normal’ when we standardise care. So,
to these recommendations. despite some limited evidence basis, we argue that the

© 2019 Association of Anaesthetists 1

Anaesthesia 2019
mere presence of guidelines for care supports patient safety
and helps the infrequent user.
Beyond safety and efficiency, standardised care offers
another important advantage. It serves to advance the
science of the care we provide. When we all approach
awake tracheal intubation differently, we have limited
understanding of what works and how we may improve.
Instead, we have all heard things like, ‘This is how I do it, and
it works well’. We have no idea if that approach is what
works, the providers’ familiarity with their own approach or
the individual provider’s skill set and bed-side manner. With
these guidelines, we now effectively have a control by which
we can conduct clinical experiments. Defining this usual
care serves future investigators’ capacity to advance the
science of awake tracheal intubation.
Why is awake tracheal intubation
underutilised?
Fear of making the procedure uncomfortable for the
patient
Clinicians may find relief in the only controlled study on the
patient’s experience of the procedure [6]. In the awake
group, patients lightly sedated to a Ramsay sedation scale
of 2 (co-operative, orientated and tranquil) or 3 (responding
to commands only), only one in seven patients reported any
discomfort. When the patients were asked what they feared
the most (if they feared anything), almost twice as many
(27%) answered that they feared the surgery compared with
tracheal intubation (17%) [6].
Belief that the procedure will be time consuming
Procedural time may be trivial if the clinician becomes
proficient with the procedure. The median time to perform
awake tracheal intubation is only 8 min longer than for
tracheal intubation after induction [7]. However, time itself
should not be the determining factor in cases where awake
tracheal intubation is considered the safest approach, and
the alternative, tracheal intubation attempt after induction
of anaesthesia in a predicted difficult airway patient may
well take longer, and bears the concomitant increased risk
of having to revert to emergency front-of-neck access.
Indications for awake tracheal
intubation
These guidelines do not include detailed recommendations
on which patient features or conditions merit awake tracheal
intubation. Historically, bed-side examination features
indicating a difficult direct laryngoscopy, such as a high
Mallampati classification score, were common indications
2 © 2019 Association of Anaesthetists
Editorial
for awake tracheal intubation, but evidence suggests that
asleep videolaryngoscopy may often overcome this barrier
[8]. One may suspect that this advancement has resulted in a
reduction in the incidence of awake tracheal intubations
performed, but these presumptions have not been
supported by evidence to date [9]. We opine that certain
patient populations deserve routine consideration for
awake tracheal intubation. First, the patient with neck
pathology from tumour, previous surgery, obstructive
lesions or radiation therapy is known to be at increased risk
for failed facemask ventilation, direct laryngoscopy,
videolaryngoscopy and potentially surgical airway [10, 11].
It seems prudent to consider the nature of the pathology in
these patients and come to a reassurance that asleep airway
management is safe before induction of anaesthesia, and if
not, then choose awake tracheal intubation. We also
suggest that there may be haemodynamic indications for
awake tracheal intubation: once the airway is secured it
allows a gradual transition to positive pressure ventilation
and a slow induction of anaesthesia for those who may
suffer from increased risk of haemodynamic collapse. The
patient with a large anterior mediastinal mass may meet
such an indication. Further considerations are summarised
in Table 1.
Awake tracheal intubation techniques
The reader might get the impression from these guidelines
that only two techniques exist for awake tracheal intubation:
flexible bronchoscopy (FB) and videolaryngoscopy (VL) with
brief mentioning of awake tracheal intubation via a
supraglottic airway or guided by a lighted stylet. We
highlight that grade-A evidence supporting comparable
success rates of awake intubation with VL with FB are limited
by study design to patient cohorts with relatively easy oral
access who are often deeply sedated. We agree that awake
tracheal intubation with FB and VL will cover the vast
majority of situations, but these techniques are not equally
applicable, and in many patients, awake VL will be difficult,
or even impossible (e.g. insufficient mouth opening).
We think that it is essential for anaesthetists to realise
that all tracheal intubation techniques can be applied in the
awake patient. The authors state that bleeding is a relative
contra-indication for awake tracheal intubation – we
strongly disagree! Yes, awake tracheal intubation with FB or
VL may be difficult, or impossible, in the bleeding patient,
but if the airway examination indicates an anatomically
difficult airway then an awake approach, either awake
tracheal intubation or awake front-of-neck access, is
definitively as indicated as it would have been had the
Editorial Anaesthesia 2019

Table 1 Advantages and disadvantages of intubating the patient awake/anaesthetised and breathing spontaneously or anaes-
thetised and apnoeic. Y = Yes, N= No. Reprinted with permission from the Scandinavian Airway management course ‘Airway
management for Anaesthesiologists’ at www.airwaymanagement.dk.
Anaesthetised, breathing Anaesthetised,
Awake spontaneously apnoeic
Advantages of an awake patient
Patent airway preserved Y Y N
Spontaneous breathing preserved Y Y N
Easier to localise glottic opening (air bubbles) Y Y N
The patient can help (e.g. protrude the tongue) Y N N
The patient can be sitting Y N N
Some protection against aspiration Y N N
(patient may be able to cough if content enters trachea)
Observing the patient’s neurological status Y N N
(e.g. In cervical fracture)
Allows decision making (intubate or not?) during the Y N N
endoscopy (e.g. in suspected epiglottitis)
Avoid cardiovascular depression Y N N
Disadvantages of an awake patient
Cardiovascular stimulation Y Y/N N
Patient discomfort? (Y) N N
More time consuming? (Y) (Y) (N)

airway not been bleeding – or even more so [13]. Awake
tracheal intubation can still be achieved, even if severe
bleeding impedes visibility with VL and FB, by using
techniques that are relatively unaffected by the presence of
blood [12]. These techniques include retrograde-[12], blind
nasal-, light- or flexible-video/optical scope-guided tracheal
intubation via a supraglottic airway device (provided the
origin of the bleeding is above the supraglottic airway
device); or awake front-of-neck access [12]. Awake front-of-
neck access remains the most conservative approach to
difficult airway management and should always be
considered.
Route
The authors state that there is no proven benefit of the oral
or nasal route; however, it is obvious that some potentially
severe complications as epistaxis can only happen if the
nasal route is chosen. We thus suggest the oral route to be
chosen as the first choice, all other indications being equal.
oxygenation and performance’ and may prompt the
clinician to perform the tasks in that sequence which could
become dangerous because both sedation and
topicalisation can cause apnoea and obstruction and thus
must be preceded and accompanied by distribution of
oxygen. Furthermore, sedation, if to be used at all, should
come before topicalisation because topicalisation may be
equally uncomfortable for the patient as tracheal intubation
itself, and therefore should be preceded by sedation. So,
from a clinical point of view, and as a help for the infrequent
user, the sequence should be: ‘oxygenation, sedation,
topicalisation and performance’ (OsTP). However, as it has
been shown that abbreviations [14] and acronyms [15] in
medicine and science are often counterproductive and may
present an imminent disaster [14], we suggest using words
spelt out in full. Additionally, this would also prevent
confusion with the use of the word ‘sTOP’ in the
‘unsuccessful awake tracheal intubation’ part of the
guidelines, where it actually means stop!

‘sTOP’ vs. ‘OsTP’: oxygenation, Attempts and unsuccessful awake

sedation, topicalisation and
performance
The authors have given in to the understandable temptation
to use a well-sounding acronym ‘sTOP’ to highlight the steps
for awake tracheal intubation, but we believe that it can be
misleading. Here ‘sTOP’ stands for ‘sedation, topicalisation,
tracheal intubation
The guidelines define an unsuccessful attempt at awake
tracheal intubation as “unplanned removal of flexible
bronchoscope, videolaryngoscope or tracheal tube from the
airway”. The guidelines state that if a total of three to four
attempts at awake tracheal intubation have failed and if

© 2019 Association of Anaesthetists 3

Anaesthesia 2019
abandoning tracheal intubation is not an option and awake
front-of-neck access is inappropriate or unsuccessful then
‘the only remaining option is a high-risk anaesthetic’
involving induction of anaesthesia despite the prediction of
the considerable likelihood of failure. We do not agree
that a ‘high-risk anaesthetic’ is the only option. Following
the definition above, withdrawal of the flexible scope
due to secretions four times would prompt the team to
abandon awake tracheal intubation. If the attempts have
been gentle and atraumatic then additional optimised
attempts by an alternative technique/operator/sedation
can still be the answer and may very well be safer than
the risky induction of anaesthesia with apnoea.
Furthermore, there is an additional option in this
scenario: deep sedation or general anaesthesia with
preserved spontaneous ventilation which bears several
advantages over managing the difficult airway of an
apnoeic patient (see Table 1). In this setting, induction
and maintenance with ketamine [16] can be favourable.
If the ‘high-risk anaesthetic’ is chosen, then this should
only occur after the trachea and the cricothyroid
membrane are identified and properly marked, if
necessary with the aid of ultrasonography [17].
How do we obtain, and conserve, skills
in awake tracheal intubation?
With the advent of simulators that can simulate both
pathology in the airway and the patients’ reaction to
tracheal intubation attempts, it is now possible to train for
awake tracheal intubation even in the absence of patients
whose airways are difficult to manage. Training with these
simulators has been shown to improve tracheal intubation
success (at least in anaesthetised patients [18]). These
simulators have proven their validity [19] and we
encourage their incorporation into training programmes
for residents and for achievement and maintenance of
skills for anaesthetists. New adjuncts, like an emitter of
infrared blinking light placed on the skin on the
cricothyroid membrane, resulting in visible light emerging
from the trachea as guidance for the flexible scope [20],
may further support the safe application of awake
intubation for the infrequent user.
This editorial is intended to help get the most out of
these guidelines with their impressive attention to detail and
clinical usefulness. We encourage all airway practitioners,
both novice and expert, to scrutinise and use these
guidelines for improvement of personal readiness and
capability to perform awake tracheal intubation. We further
encourage regional and departmental standards based on
these guidelines and commentary.
4 © 2019 Association of Anaesthetists
Editorial
Acknowledgements
No competing interests declared.
References
1. Ahmad I, El-Boghdadly K, Bhagrath R, et al. Difficult Airway
Society guidelines for awake tracheal intubation (ATI) in adults.
Anaesthesia 2019; https://doi.org/10.1111/anae.14904.
2. Komajda M, Lapuerta P, Hermans N, et al. Adherence to
guidelines is a predictor of outcome in chronic heart
failure: the MAHLER survey. European Heart Journal 2005;
26: 1653–9.
3. Wilke M, Grube RF, Bodmann KF. Guideline-adherent initial
intravenous antibiotic therapy for hospital-acquired/ventilator-
associated pneumonia is clinically superior, saves lives and is
cheaper than non guideline adherent therapy. European
Journal of Medical Research 2011; 16: 315–23.
4. Wockel A, Kurzeder C, Geyer V, et al. Effects of guideline
adherence in primary breast cancer–a 5-year multi-center
cohort study of 3976 patients. Breast 2010; 19: 120–7.
5. Proietti M, Nobili A, Raparelli V, et al. Adherence to
antithrombotic therapy guidelines improves mortality among
elderly patients with atrial fibrillation: insights from the REPOSI
study. Clinical Research in Cardiology 2016; 105: 912–20.
6. Schnack DT, Kristensen MS, Rasmussen LS. Patients’ experience
of awake versus anaesthetised orotracheal intubation: a
controlled study. European Journal of Anaesthesiology 2011;
28: 438–42.
7. Joseph TT, Gal JS, DeMaria S Jr, Lin HM, Levine AI, Hyman
JB. A retrospective study of success, failure, and time
needed to perform awake intubation. Anesthesiology 2016;
125: 105–14.
8. Aziz MF, Bayman EO, Van Tienderen MM, Todd MM, St AGEIG,
Brambrink AM. Predictors of difficult videolaryngoscopy with
GlideScope(R) or C-MAC(R) with D-blade: secondary analysis
from a large comparative videolaryngoscopy trial. British
Journal of Anaesthesiology 2016; 117: 118–23.
9. Law JA, Morris IR, Brousseau PA, de la Ronde S, Milne AD.
The incidence, success rate, and complications of awake
tracheal intubation in 1,554 patients over 12 years: an
historical cohort study. Canadian Journal of Anesthesia
2015; 62: 736–44.
10. Aziz MF, Healy D, Kheterpal S, Fu RF, Dillman D, Brambrink AM.
Routine clinical practice effectiveness of the Glidescope in
difficult airway management: an analysis of 2,004 Glidescope
intubations, complications, and failures from two institutions.
Anesthesiology 2011; 114: 34–41.
11. Kheterpal S, Healy D, Aziz MF, et al. Incidence, predictors, and
outcome of difficult mask ventilation combined with difficult
laryngoscopy: a report from the multicenter perioperative
outcomes group. Anesthesiology 2013; 119: 1360–9.
12. Kristensen MS, McGuire B. Managing and securing the
bleeding upper airway: a narrative review. Canadian Journal of
Anesthesia 2019; https://doi.org/10.1007/s12630-019-01479-5.
13. Kristensen MS. Airwaymanagement.dk.: Retrograde intubation
with the pulling-the-catheter-method in bloody, bleeding,
patients with difficult airway. 2018. www.airwaymanagement.d
k/retrograde (accessed 08/12/2019).
14. Parvaiz MA, Subramanian A, Kendall NS. The use of
abbreviations in medical records in a multidisciplinary world–
an imminent disaster. Communication and Medicine 2008; 5:
25–33.
15. Hales AH, Williams KD, Richter J. Alienating the audience: how
abbreviations hamper scientific communication. Observer
2017; 30: 22–4.
16. Merelman AH, Perlmutter MC, Strayer RJ. Alternatives to rapid
sequence intubation: contemporary airway management with
Editorial
ketamine. Western Journal of Emergency Medicine 2019; 20:
466–71.
17. Kristensen MS, Teoh WH, Rudolph SS, Hesselfeldt R, Borglum J,
Tvede MF. A randomised cross-over comparison of the
transverse and longitudinal techniques for ultrasound-guided
identification of the cricothyroid membrane in morbidly obese
subjects. Anaesthesia 2016; 71: 675–83.
18. Wong DT, Mehta A, Singh KP, et al. The effect of virtual reality
bronchoscopy simulator training on performance of
© 2019 Association of Anaesthetists
Anaesthesia 2019
bronchoscopic-guided intubation in patients: a randomised
controlled trial. European Journal of Anaesthesiology 2019;
36: 227–33.
19. Baker P. Assessment before airway management.
Anesthesiology Clinics 2015; 33: 257–78.
20. Kristensen MS, Fried E, Biro P. Infrared Red Intubation System
(IRRIS) guided flexile videoscope assisted difficult airway
management. Acta Anaesthesiologica Scandinavica 2018; 62:
19–25.
5
Anaesthesia 2020 doi:10.1111/anae.14992

Editorial

Fasten your seatbelts: innovation in regional anaesthesia is

a bumpy ride
E. R. Mariano1,2 and K. El-Boghdadly3,4

1 Professor, Department of Anesthesiology, Peri-operative and Pain Medicine, Stanford University School of Medicine,
Stanford, CA, USA
2 Chief, Anesthesiology and Peri-operative Care Service, Veterans Affairs Palo Alto Health Care System, Palo Alto, CA,
USA
3 Consultant, Department of Anaesthesia, Guy’s and St Thomas’ NHS Foundation Trust, London, UK
4 Honorary Senior Lecturer, King’s College London, UK
............................................................................................................................................................................................................................................................................................................
Correspondence to: E. R. Mariano
Email: emariano@stanford.edu
Accepted: 18 December 2019
Keywords: diaphragm; hemidiaphragmatic paralysis; interscalene block; phrenic nerve; suprascapular
This editorial accompanies an article by Ferr
e et al., Anaesthesia 2020; https://doi.org/10.1111/anae.14978.
Twitter: @EMARIANOMD; @elboghdadly

achieved by early versions of two-point seatbelts, there was

Seatbelts save lives.
Early studies of motor vehicle accidents demonstrated a
significant reduction in major and fatal injuries attributed to
the use of seatbelts that ranged from 35% to 80%
depending on the study [1]. In the UK, registered cars were
required to have front seatbelts by 1965 although
mandatory use of them did not become law until 1981 [2].
Initially, seatbelts only offered two-point restraint (i.e. lap or
diagonal, but not both). While mortality data strongly
support any seatbelt over no seatbelt, the two-point
diagonal design is considered the least effective and most
dangerous in terms of causing injury [2]. Diagonal two-point
seatbelts can potentially cause soft tissue injury to the neck,
chest and breasts as well as rib and sternal fractures [2].
Potential injuries associated with two-point lap seatbelts
include intra-abdominal
perforation, haematomas from mesenteric tears and vessel
injury, gallbladder rupture, pancreatic damage and closed
loop obstruction [2]. Skeletal injury in the form of pelvic and
spinal fractures or ligament disruptions has also been
associated with lap-only seatbelts [1]. Visible bruising or
friction burns on the abdomen at the site of the seatbelt was
termed a ‘seatbelt sign’ and demanded further investigation
to rule out serious injury [2]. Despite the safety strides
a clear need to improve the design [1].
This year marks the 60th anniversary of the modern
three-point seatbelt [3]. A former aviation engineer, Nils
Bohlin, is credited for developing this innovative seatbelt
design for Volvo, and customers in the Scandinavian
market had access to vehicles with three-point seatbelts as
early as 1959 [3]. Although they have also been associated
with seatbelt-related injuries [4, 5], 1991–1997 data from
the US Fatality Analysis Reporting System on head-on
collisions show a 72% decreased odds of death when
passengers are using a three-point seatbelt compared
with no seatbelt [6]. This means more lives saved over the
previous model of two-point lap-only seatbelts which were
associated with a 57% decreased odds of death [6].
Today, three-point seatbelts are standard, and all
trauma such as intestinal automobile manufacturers have implemented them. Why
are we talking about seatbelts? Three-point seatbelts
represent a modern innovation that has improved
outcomes that matter to people and has changed the
automobile industry.
In this issue of Anaesthesia, Ferr
e et al. present the
results of a randomised clinical trial comparing an anterior
(proximal) with a posterior (distal) approach to the
suprascapular nerve block in terms of hemidiaphragmatic

© 2020 Association of Anaesthetists 1

Anaesthesia 2020
paralysis assessed by ultrasound [7]. All participants in the
study underwent arthroscopic shoulder surgery and also
received an axillary nerve block consistent with the selective
‘shoulder block’ concept introduced by Dr Price [8]. The
only comorbidity exclusion was pre-existing respiratory
failure, and participants were generally healthy and non-
obese. Not surprisingly, the block approach performed
further away from the brachial plexus was associated with
a lower rate of sonographically-demonstrated hemi-
diaphragmatic paralysis [7]. This study, while meticulously
performed, raises two important questions: is this a
patient-centred outcome; and will the results change
clinical practice?
Patient-centred outcomes
The Patient-Centred Outcomes Research Institute (PCORI)
defines patient-centred as being “meaningful and important
to patients and caregivers” [9]. Using the seatbelt example,
investigating innovations to decrease the potential
outcomes of death and major injury from a motor vehicle
accident fit the PCORI definition of patient-centred since
“the focus includes outcomes that matter to patients” [9].
This is an oft-discussed topic in many modern anaesthesia
studies [10], and similarly, we question whether the
sonographic detection of hemidiaphragmatic paralysis, the
primary outcome of the present study [7], meets this
criterion.
The assessment of hemidiaphragmatic function was
performed by clinicians skilled in point-of-care ultrasound.
High-end portable ultrasound equipment was used to
visualise the diaphragm and estimate the degree of
diaphragmatic motion using M-mode. Study participants
were asked to breathe normally and perform a ‘sniff test’ (i.e.
forced inhalation through the nose) on command.
Consistent with the methodology used in previous studies
[11, 12], hemidiaphragmatic paralysis was considered
complete if the reduction in movement was > 75% and
partial if reduction was between 25% and 75% [7]. This
detailed measurement is appropriate for a research study
but is not easy to implement in routine clinical practice. In
the study by Kim et al. which evaluated the outcome of
hemidiaphragmatic paralysis in a comparison of superior
trunk and interscalene blocks, investigators noted difficulty
in visualising the left hemidiaphragm [12]. They stated,
“Visualisation of the left-sided hemidiaphragmatic motion is
more challenging because of the smaller window of the
spleen, especially in patients with high body mass indexes”
[12]. For this reason, they excluded patients from the study
2
who had a BMI over 35 kg.m and developed an alternate
technique for visualising the left hemidiaphragm in the case
2 © 2020 Association of Anaesthetists
Editorial
of a difficult scan [12]. Determining the implications of
research studies that primarily report changes in
sonographically-detected diaphragmatic excursion from
the patient perspective is nearly impossible. The clinical
skills to obtain this measurement are not universally
available in the peri-operative setting, and rarely do patients
exhibit any symptoms from hemidiaphragmatic paralysis.
Dyspnoea is a patient-centred outcome measure that
can be a marker of underlying involvement of the phrenic
nerve, pneumothorax or other sinister complications, and
can indicate further treatment and prolonged hospital stay.
However, even with sonographic evidence of hemi-
diaphragmatic paralysis, as well as a change in forced
expiratory volumes and vital capacity, dyspnoea remains
uncommon in modern practice [13]. In patients who do
report dyspnoea after interscalene or supraclavicular
brachial plexus blocks following shoulder surgery, only a
third to three-quarters actually have objective evidence of
hemidiaphragmatic paralysis [11, 14]. There are several
reasons that patients may experience dyspnoea after
regional anaesthesia, general anaesthesia and shoulder
surgery; thus, this symptom is neither sensitive nor specific.
Overall, the relationship between ultrasound-diagnosed
hemidiaphragmatic paralysis and clinical symptomatology
is unclear.
The quantification of diaphragmatic motion by
ultrasound is a decidedly clinician-focused metric and a
distant surrogate marker for the clinically-relevant outcome
of respiratory distress. Although it is measurable and
therefore permits the estimation of a relatively smaller
sample size for a clinical trial, the outcome that matters to
patients is more likely to be clinical respiratory distress
requiring intervention, emergency care or hospital
admission. This occurred in only one participant from the 84
person study [7], an incidence of 1.1%. To reduce this
incidence by half (0.55%) in a randomised clinical trial study
(ɑ 0.05 and b 0.2), 3460 per group are required with a total
enrolment of 6920 participants. Whilst not impossible, the
feasibility of such a study is limited, and the rarity of this
complication means that this may not be worth exploring.
Indeed, serious postoperative complications after arthro-
scopic shoulder surgery are uncommon with an estimate of
all-cause readmission in the US of approximately 1% [15].
Taking the above facts into consideration, studies of patient-
centred and clinically-oriented outcomes in regional
anaesthesia can be very challenging.
Changing practice
The next important question is whether the results of
this study will change clinical practice. The goal of
Editorial
reducing hemidiaphragmatic paralysis
priority for regional anaesthetists since Urmey et al.
reported the 100% incidence of this effect following
interscalene brachial plexus block in 1991 [16]. Since
then, many modifications of and alternatives to the
interscalene brachial plexus block have been described
[13, 17] with the shoulder block (a combination of distal
suprascapular and axillary nerve blocks) introduced in
2007 being one of them [8]. A randomised clinical trial
by Dhir et al. comparing the shoulder block with the
interscalene brachial plexus block for
shoulder surgery, and designed to test for equivalence,
demonstrated that interscalene brachial plexus block
provides more effective analgesia in the immediate
postoperative period [18]. In addition, even in the
hands of regional anaesthesia experts who performed
the procedures, the shoulder block on average took
approximately twice as long to perform than the
interscalene brachial plexus block, which makes sense
since two individual nerves are blocked [18]. This
suggests that a high level of expertise is required as
well as sufficient time. In the US, few patients having
outpatient surgery tend to experience peri-operative
nerve blockade, approximately 3% among all surgical
subtypes, and the majority of patients who specifically
undergo outpatient shoulder surgery (59%) do not
receive a block [19]. System- and training-related factors
play a role, but every regional anaesthetist knows that
the only block that a patient will get is the one that
best combines speed and efficacy. Distal blocks for
shoulder surgery likely have an indication in specific
cases (e.g. patient with severe chronic pulmonary
disease causing respiratory failure), but
continue to be tools for experts who already practice
regional anaesthesia. It is our opinion
technique will not be regularly adopted by novices as
the go-to block technique for shoulder surgery and
therefore will not expand patient access to regional
anaesthesia [20].
American entrepreneur Katrina Cole advised us to
“focus on things that are small enough to change but big
enough to matter”. This could be the rallying cry for
outcomes research in regional anaesthesia.
innovations in block techniques may never result in the
same public health impact as the three-point seatbelt, they
can aspire to address patient-centred outcomes, the things
that matter to real people and lead to changes in clinical
practice that ideally improve quality and access to regional
anaesthetic techniques.
© 2020 Association of Anaesthetists
Anaesthesia 2020
had been a Acknowledgements
This work was supported with resources and the use of
facilities at the Veterans Affairs Palo Alto Health Care System
(Palo Alto, CA, USA). The contents do not represent the
views of the Department of Veterans Affairs or the United
States Government. KE is an Editor of Anaesthesia. No
external funding or competing interests declared.
References
1. Hodson-Walker NJ. The value of safety belts: a review.
arthroscopic Canadian Medical Association Journal 1970; 102: 391–3.
2. Banerjee A. Seat belts and injury patterns: evolution and
present perspectives. Postgraduate Medicine Journal 1989;
65: 199–204.
3. Stoklosa A. The Three-Point Seatbelt Turns 60, and It’s a Damn
Hero Car and Driver. 2019. https://www.caranddriver.com/ne
ws/a28775593/three-point-seatbelt-history/: Hearst Communi
cations, 2019 (accessed 18/12/2019).
4. Al-Ozaibi L, Adnan J, Hassan B, Al-Mazroui A, Al-Badri F. Seat
belt syndrome: delayed or missed intestinal injuries, a case
report and review of literature. Interntional Journal of Surgery
Case Reports 2016; 20: 74–6.
5. Song CT, Teo I, Song C. Systematic review of seat-belt trauma to
the female breast: a new diagnosis and management
classification. Journal of Plastic Reconstructive and Aesthetic
Surgery 2015; 68: 382–9.
6. Crandall CS, Olson LM, Sklar DP. Mortality reduction with air
bag and seat belt use in head-on passenger car collisions.
American Journal of Epidemiology 2001; 153: 219–24.
7. Ferre F, Pommier M, Laumonerie P, et al. Hemidiaphragmatic
paralysis following ultrasound-guided suprascapular nerve
block: a randomized, double-blind study evaluating an anterior
versus posterior needle approach. Anaesthesia 2020; https://
doi.org/10.1111/anae.14978.
8. Price DJ. The shoulder block: a new alternative to interscalene
brachial plexus blockade for the control of postoperative
shoulder pain. Anaesthesia and Intensive Care 2007; 35: 575–
81.
9. Frank L, Basch E, Selby JV. The PCORI perspective on patient-
centered outcomes research. Journal of the American Medical
they will Association 2014; 312: 1513–14.
10. Shah A, Bailey CR. Outcomes following surgery: are we
measuring what really matters? Anaesthesia 2019; 74: 696–
that this 9.
11. Wiesmann T, Feldmann C, Muller HH, et al. Phrenic palsy and
analgesic quality of continuous supraclavicular vs. interscalene
plexus blocks after shoulder surgery. Acta Anaesthesiologica
Scandinavica 2016; 60: 1142–51.
12. Kim DH, Lin Y, Beathe JC, et al. Superior trunk block: a phrenic-
sparing alternative to the interscalene block: a randomized
controlled trial. Anesthesiology 2019; 131: 521–33.
13. El-Boghdadly K, Chin KJ, Chan VWS. Phrenic nerve palsy and
regional anesthesia for shoulder surgery: anatomical,
physiologic, and clinical considerations. Anesthesiology 2017;
While 127: 173–91.
14. Petrar SD, Seltenrich ME, Head SJ, Schwarz SK.
Hemidiaphragmatic paralysis following ultrasound-guided
supraclavicular versus infraclavicular brachial plexus blockade:
a randomized clinical trial. Regional Anesthesia and Pain
Medicine 2015; 40: 133–8.
15. Hill JR, McKnight B, Pannell WC, et al. Risk factors for 30-day
readmission following shoulder arthroscopy. Arthroscopy
2017; 33: 55–61.
3
Anaesthesia 2020
16. Urmey WF, Talts KH, Sharrock NE. One hundred percent
incidence of hemidiaphragmatic paresis associated with
interscalene brachial plexus anesthesia as diagnosed by
ultrasonography. Anesthesia and Analgesia 1991; 72: 498–
503.
17. Tran QH, Elgueta MF, Aliste J, Finlayson RJ. Diaphragm-sparing
nerve blocks for shoulder surgery. Regional Anesthesia and
Pain Medicine 2017; 42: 32–8.
18. Dhir S, Sondekoppam RV, Sharma R, Ganapathy S, Athwal GS.
A comparison of combined suprascapular and axillary nerve
4 © 2020 Association of Anaesthetists
Editorial
blocks to interscalene nerve block for analgesia in arthroscopic
shoulder surgery: an equivalence study. Regional Anesthesia
and Pain Medicine 2016; 41: 564–71.
19. Gabriel RA, Ilfeld BM. Use of regional anesthesia for outpatient
surgery within the United States: a prevalence study using a
nationwide database. Anesthesia and Analgesia 2018; 126:
2078–84.
20. Turbitt LR, Mariano ER, El-Boghdadly K. Future directions in
regional anaesthesia: not just for the cognoscenti. Anaesthesia
2019. https://doi.org/10.1111/anae.14768.
Anaesthesia 2019 doi:10.1111/anae.14853

Editorial

Abandoning inhalational anaesthesia

S. M. White1 and C. L. Shelton2

1 Consultant, Department of Anaesthesia, Sussex University Hospitals NHS Trust, Brighton, East Sussex, UK
2 Consultant, Department of Anaesthesia, Wythenshawe Hospital, Manchester University NHS Foundation Trust,
Manchester, UK
............................................................................................................................................................................................................................................................................................................
Correspondence to: S. White
Email: stuart.white6@nhs.net
Accepted: 20 August 2019
Keywords: anesthesia, inhalation; adverse effects; anesthesia, toxicity; air pollutants, environmental
Twitter: @DrCliffShelton

The 2018 Intergovernmental Panel on Climate Change
(IPCC) Special Report Global Warming of 1.5°C (SR15)
makes sobering reading [1]. At the current rate, global
warming is likely to reach 1.5°C above pre-industrial levels
between 2030 and 2052, and this is likely to result in far-
reaching effects, including: climate change; sea level rise;
loss of biodiversity; global health problems; food and
water shortage; mass
insecurity. Adaptation to
possible below a rise of 1.5°C, but would become
increasingly difficult towards 2°C and beyond, as ‘tipping
points’ are exceeded.
Limiting global warming to 1.5°C will require global net
emissions of carbon dioxide to fall by approximately 45%
from 2010 levels by 2030, reaching ‘net zero’ around 2050
[1]. These targets are much more stringent than those
agreed by the 1998 Kyoto Protocol [2], exceed the scope of
the 2015 Paris Agreement [3] and will require ‘rapid and far
reaching’ co-operative global socio-economic changes,
including: shifts to renewable energy; changes to food
systems; transport electrification; urban restructuring; and
carbon capture. Independent consensus supports the
immediacy of the changes needed [4].
Inhalational anaesthetic agents are fluorocarbons that
contribute to man-made climate change primarily through
‘radiative forcing’ (i.e. acting like greenhouse gases), but
also through the carbon cost of their life cycle (manufacture,
storage, transport, use and recovery). Their global warming
potentials can be compared with that of carbon dioxide
over a given time period (usually 20 years or 100 years).
Desflurane has the highest global warming potential (one
240-ml bottle of vaporised desflurane being equivalent to
the global warming effect of 1296 kg CO2 over 20 years),
much higher than that of isoflurane and sevoflurane
1
(521 kg and 132 kg CO2 equivalents (CO2e).250ml
bottle, respectively) [5–7].
Currently, the global market for inhalational
anaesthetic agents is ~12.5 million bottles (worth US$1.12
migration; and geopolitical billion), a figure projected to rise to ~14.9 million bottles by
these challenges remains 2025 (worth $1.34 billion) [8]. Sevoflurane is the market
leader (sevoflurane 70%, desflurane 20%, isoflurane 10%).
From these figures, and assuming negligible recovery or
metabolism, the estimated annual contribution of
inhalational anaesthetic agents represents ~ 0.01% of global
CO2 production (0.005/37.1 Gt CO2e) [9]. Although a
relatively small proportion, this is equivalent to flying an
average commercial airliner ~418 times around the world.
The NHS Sustainable Development Unit has estimated that
anaesthetic gases (including nitrous oxide) contribute ~5%
of the carbon footprint of UK hospitals (~0.56/10.4 million
tons CO2e; ~0.08 mt (~0.8%) inhalational anaesthetic agents
1
only) [10]. An anaesthetist using low-flow (< 0.5 l.min ) 6%
desflurane in 50/50 air/O2 after intravenous induction in
theatres without scavenging, uses ~60 ml desflurane during
a 7-h operating list [11]. If the gases are vented to the
atmosphere (as is invariably the case), this is equivalent to
~ 325 kg CO2 (GWP20). Assuming three lists.week 1
, 45
1
weeks.year , over a 40-year career, that anaesthetist would
have been responsible for releasing 1755 tons CO2e; this is
equivalent to ~880 return passenger flights between
London Heathrow and JFK airports, ~ 22.year 1
or one
trans-Atlantic flight for every week of work.

© 2019 Association of Anaesthetists 1

Anaesthesia 2019
The Association of Anaesthetists has done much to
raise professional awareness about the environmental
impacts of using inhalational anaesthetic agents and how to
mitigate these [https://anaesthetists.org/Home/Resources-
publications/Environment], and recently supported the
NHS ‘Clean Air’ day [https://anaesthetists.org/Home/
Resources-publications/Environment/Clean-Air-Day-2019].
Similarly, the Royal College of Anaesthetists has recently
committed to minimising the environmental effects of
anaesthesia, but without specifically addressing the impact
of inhalational anaesthetic agents [12]. Other authors have
called for enhanced awareness of the environmental harms
of inhalational agents and reductions in their use through
low-flow anaesthesia [13] and the capture and reprocessing
of exhaled vapours [14]. However, even if
technologies were universally available and affordable,
atmospheric escape of inhalational agents would continue
to occur.
Instead, given the urgency required to avoid the
imminent climate catastrophe identified by the IPCC, we
argue that it is ‘too late to mitigate’: the profession of
anaesthesia must consider specifically abandoning the use
of inhalational anaesthetic agents by 2030.
Are inhalational agents indispensable?
Given their environmental impact, it is perhaps surprising
that the use of inhalational anaesthetic agents has not been
addressed by either international climate agreements or
national carbon reduction strategies, to date. Neither the
Montreal Protocol and Kigali Amendment (dealing with
ozone depletion) nor the Kyoto Protocol and Paris
Agreement mention inhalational anaesthetic
Likewise, they are not referred to in the Climate Change Act
2008, its healthcare application in the NHS Carbon
Reduction Strategy 2009, or in the more recent UK Carbon
Plan 2017, even though the UK Government has recently
committed to net zero carbon emissions by 2050.
This may be oversight on the part of the authorities
involved, but it may also represent a form of healthcare
exceptionalism, in effect ignoring the environmental harm
caused by inhalational anaesthetic agents because they are
considered indispensable for healthcare. But are they?
Locoregional anaesthesia (LRA) and total intravenous
anaesthesia (TIVA) can be equally effective as, and in some
cases superior to, inhalational anaesthesia [15]. We accept
that these options are not without their own environmental
costs, but even accounting for the increased production of
(often plastic) waste involved in these modalities, their
carbon impact is many times less than that of the
inhalational alternative.
2 © 2019 Association of Anaesthetists
Editorial
Furthermore, we suggest that there is no absolute
indication for inhalational anaesthesia use that could not be
substituted using LRA/TIVA, including: the management of
complex airways; unfasted patients; those with poor
vascular access; and anaesthesia for bariatric and paediatric
surgery. Barriers to the use of LRA/TIVA in these contexts
appear to relate more to operator confidence than to
evidence. Although we fully accept that additional research
would be valuable, it appears that better training should be
the primary focus in aiding their adoption [16].
Perhaps the main reason that the profession has been
sluggish in committing to using alternatives to inhalational
anaesthesia is cultural. Inhalational anaesthesia remains the
default method of providing general anaesthesia, globally.
these Delivered by ‘gas(wo)men’, its effectiveness has been
confirmed through many patient-years of experience. From
the outset of their specialist education, anaesthetists are
taught that inhalational maintenance is the ‘standard’ option
for providing general anaesthesia. In contrast, TIVA does
not appear as a training objective until page 66 of the Royal
College of Anaesthetist’s Intermediate Training Curriculum,
and is framed as an ‘alternative’ to inhalational anaesthesia,
against which its comparative risks and benefits should be
assessed. Why shouldn’t the converse apply in future
iterations of the document?
Bowker and Star have pointed out that superior
marketing and the favour of a community of gatekeepers
may maintain the use of established standards ahead of
those which may be technological improvements, noting
“there is no natural law that the best standard shall win” [17].
We suggest that the continued pre-eminence of inhalational
agents. anaesthesia within the orthodoxy of our profession occurs
not due to any clinical superiority but due to a clinical
tradition and an educational model that makes such
acceptance almost inevitable.
Critics of our position may point out that anaesthetists
have a duty to provide safe and effective anaesthesia to the
patient in front of them, and that for many clinicians this
means using an inhalational technique. However, we
suggest that we also have a duty of care to a person whose
health is, or will become, affected by climate change, which
will include people in the UK if global warming continues at
its current rate [18].
Change management
We recognise that change is difficult. There are challenges
for those who choose to adopt TIVA as their primary method
of administering general anaesthesia, for example,
although many sources of information and training aids
exist [16, 19]. Professionally, however, we have always been
Editorial
able to accommodate similar considerations (e.g. when
propofol was introduced replacing thiopentone, and when
halothane/enflurane use was discontinued). In addition, we
need to recognise that patients may increasingly express a
shared decision-making interest in the carbon footprint of
their healthcare. The National Institute for Health and Care
Excellence (NICE), for example, has recently published a
decision aid to help patients and their doctors choose an
appropriate metered-dose inhaler for asthma management
based not only on clinical factors but also on the carbon
footprint of each dose of medication [20]. If patients are
questioning the carbon impact of their inhalers, we do not
think that it will not be long before patients ask similar
questions about their anaesthetic.
How might cultural obstacles to change be overcome?
Probably the most important and effective option is for
individual anaesthetists to take personal responsibility and
commit to abandoning inhalational anaesthetic agent
administration, as well as advocating for changes within
their department and nationally.
Professionally, bodies such as the Association of
Anaesthetists, Royal College of Anaesthetists, American
Society of Anesthesiologists and European Society of
Anaesthesiology need to: publicly and explicitly support the
rapid phasing out of inhalational anaesthetic agents
through improved education and endorsement of
alternative methods of anaesthesia; re-allocate research
funding away from studies involving fluorocarbon
inhalational agents; and reduce sponsorship income from
inhalational anaesthetic agent manufacturers, but work
(perhaps preferentially) with these companies to find
environmentally neutral alternatives through research).
Nationally, and internationally, governments, institutions
and non-governmental organisations need to: formally
recognise the contribution to global warming made by
inhalational anaesthetic agents; and commit to phasing
out their use within carbon reduction strategies, as well as
supporting research by the manufacturers into carbon-
neutral alternatives.
In order that anaesthetists and patients can make
informed choices about anaesthetic technique, manufacturers
should be compelled to make accurate data available
regarding the life-cycle carbon impact of volatile agents and
their alternatives, as well as the volume and destination of
annual inhalational anaesthetic agent sales.
It has been postulated that it takes ~17 years to change
practice in healthcare systems [21], and given that the IPCC
has forecast that global warming will exceed the 1.5°C
critical threshold between 11 and 33 years time unless its
recommendations are implemented, it seems incumbent on
© 2019 Association of Anaesthetists
Anaesthesia 2019
our profession to reduce its carbon footprint as soon as
possible. Foremost among the ‘rapid and far-reaching’
options available to us to achieve this is to abandon
inhalational anaesthesia agent use. If they haven’t done so
already, we strongly urge colleagues to play their part in
setting a new standard of care which benefits not only the
patient in front of them but also the health of future
generations, by changing their practice accordingly.
Acknowledgements
CS is a co-opted member of the Association of
Anaesthetist’s Environment and Sustainability Committee.
References
1. Masson-Delmotte V, Zhai P, P€ ortner HO, et al. Global Warming
of 1.5°C. An IPCC Special Report on the impacts of global
warming of 1.5°C above pre-industrial levels and related global
greenhouse gas emission pathways, in the context of
strengthening the global response to the threat of climate
change, sustainable development, and efforts to eradicate
poverty. 2018. https://www.ipcc.ch/sr15/ (accessed 01/07/2019).
2. United Nations. Kyoto Protocol to the United Nations
Framework Convention on Climate Change. 1998. http://
unfccc.int/resource/docs/convkp/kpeng.pdf (accessed 01/07/
2019).
3. United Nations Framework Convention on Climate
Change (UNFCCC). Adoption of the Paris Agreement. 2015.
http://unfccc.int/resource/docs/2015/cop21/eng/l09r01.pdf
(accessed 01/07/2019).
4. Watts N, Amann M, Arnell N, et al. The 2018 report of
the Lancet Countdown on health and climate change: shaping
the health of nations for centuries to come. Lancet 2018; 392:
2479–514.
5. Sherman J, Le C, Lamers V, Eckelman M. Life cycle greenhouse
gas emissions of anesthetic drugs. Anesthesia and Analgesia
2012; 114: 1086–90.
€
6. Ozelsel TJ, Sondekoppam RV, Buro K. The future is now-it’s
time to rethink the application of the Global Warming Potential
to anesthesia. Canadian Journal of Anesthesia 2019. https://
doi.org/10.1007/s12630-019-01385-w.
7. Campbell M, Pierce JMT. Atmospheric science, anaesthesia,
and the environment. Continuing Education in Anaesthesia
Critical Care and Pain 2015; 15: 173–9.
8. Grand View Research, Inc. Inhalation anesthesia market size
share & trends analysis report by application (induction,
maintenance), by product (sevoflurane, isoflurane, desflurane),
by region, and segment forecasts, 2019-2025. https://www.gra
ndviewresearch.com/industry-analysis/inhalation-anesthesia-
market (accessed 01/07/2019).
9. Ritchie H, Roser M. CO2 and other greenhouse gas emissions.
https://ourworldindata.org/co2-and-other-greenhouse-gas-
emissions (accessed 01/07/2019).
10. NHS Sustainable Development Unit. Carbon footprint from
anaesthetic gas use. 2013. https://www.sduhealth.org.uk/d
ocuments/publications/Anaesthetic_gases_research_v1.pdf
(accessed 01/07/2019).
11. Pierce T. Anaesthetic gases calculator. https://anaesthetists.
org/Portals/0/PDFs/Environment/Real_time_calculator_v10_%
20AAGBI.xlsx?ver=2019-05-12-081454-227 (accessed 01/07/
2019).
12. Royal College of Anaesthetists. Sustainability strategy 2019-
2022. 2019. https://www.rcoa.ac.uk/system/files/Sustainab
ilityStrategy2019-2022.pdf (accessed 01/07/2019).
3
Anaesthesia 2019
13. Feldman JM. Managing fresh gas flow to reduce environmental
contamination. Anesthesia and Analgesia 2012; 114: 1093–
101.
14. Charlesworth M, Swinton F. Anaesthetic gases, climate change,
and sustainable practice. Lancet Planet Health 2017; 1: e216-7.
15. Schraag S, Pradelli L, Alsaleh AJO, et al. Propofol
vs. inhalational agents to maintain general anaesthesia in
ambulatory and in-patient surgery: a systematic review and
meta-analysis. BMC Anesthesiology 2018; 18: 162.
16. Nimmo AF, Absalom AR, Bagshaw O, et al. Guidelines for the
safe practice of total intravenous anaesthesia (TIVA): Joint
Guidelines from the Association of Anaesthetists and the Society
for Intravenous Anaesthesia. Anaesthesia 2019; 74: 211–24.
17. Bowker G, Star SL. Sorting Things Out: Classification and its
Consequences. Cambridge, Massachusetts: The MIT Press,
1999: 1–32.
4 © 2019 Association of Anaesthetists
Editorial
18. Fritz Z, Cox C. Conflicting demands on a modern healthcare
service: can Rawlsian justice provide a guiding philosophy for
the NHS and other socialized health services? Bioethics 2019;
33: 609–16.
19. Al-Rifai Z, Mulvey D. Principles of total intravenous anaesthesia:
practical aspects of using total intravenous anaesthesia.
Continuing Education in Anaesthesia Critical Care and Pain
2016; 16: 276–80.
20. National Institute for Health and Care Excellence. Patient
decision aid: inhalers for asthma. 2019. https://www.nice.org.
uk/guidance/ng80/resources/inhalers-for-asthma-patient-dec
ision-aid-pdf-6727144573 (accessed 01/07/2019).
21. Balas EA, Boren SA. Managing clinical knowledge for health
care improvement. In: Bemmel J, McCray AT, eds. Yearbook of
medical informatics 2000: patient-centered systems. Stuttgart,
Germany: Schattauer Verlagsgesellschaft mbH, 2000: 65–70.
Anaesthesia 2019 doi:10.1111/anae.14900

Original Article

Red blood cell transfusion in surgery: an observational

study of the trends in the USA from 2011 to 2016*
A. T. Nordestgaard,1,2 L. S. Rasmussen,3 M. Sillesen,4 J. Steinmetz,5 A. I. Eid,1 K. Meier,1
H. M. A. Kaafarani6 and G. C. Velmahos7

1 Research Fellow, 6 Associate Professor of Surgery, 7 Professor of Surgery, Division of Trauma, Emergency Surgery and
Surgical Critical Care, Department of Surgery, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
2 Research Fellow, 3 Professor, 5 Chief Physician, Department of Anaesthesia, Centre of Head and Orthopaedics 4231,
Rigshospitalet, University of Copenhagen, Denmark
4 Chief Resident, Department of Surgical Gastroenterology and Institute for Inflammation Research, Rigshospitalet,
University of Copenhagen, Denmark

Summary
Guidelines recommend restrictive red blood cell transfusion strategies. We conducted an observational study
to examine whether the rate of peri-operative red blood cell transfusion in the USA had declined during the
period from 01 January 2011 to 31 December 2016. We included 4,273,168 patients from all surgical
subspecialties. We examined parallel trends in rates of the following: pre-operative transfusion; prevalence of
bleeding disorders and coagulopathy; and minimally invasive procedures. To account for changes in
population and procedure characteristics, we performed multivariable logistic regression to assess whether the
risk of receiving a transfusion had declined over the study period. Clinical outcomes included peri-operative
myocardial infarction, stroke and all-cause mortality at 30 days. Peri-operative red blood cell transfusion rates
declined from 37,040/441,255 (8.4%) in 2011 to 46,845/1,000,195 (4.6%) in 2016 (p < 0.001) across all
subspecialties. Compared with 2011, the corresponding adjusted OR (95%CI) for red blood cell transfusion
decreased gradually from 0.88 (0.86–0.90) in 2012 to 0.51 (0.50–0.51) in 2016 (p < 0.001). Pre-operative red
blood cell transfusion rates and the prevalence of bleeding disorders decreased, whereas haematocrit levels
and the proportion of minimally invasive procedures increased. Compared with 2011, the adjusted hazard
ratios (95%CI) in 2012 and 2016 were 0.96 (0.90–1.02) and 1.05 (0.99–1.11) for myocardial infarction, 0.91
(0.83–0.99) and 0.99 (0.92–1.07) for stroke and 0.98 (0.94–1.02) and 0.99 (0.96–1.03) for all-cause mortality. Use
of peri-operative red blood cell transfusion declined from 2011 to 2016. This was not associated with an
increase in adverse clinical outcomes.

.................................................................................................................................................................
Correspondence to: A. T. Nordestgaard
Email: ask.tybjaerg.nordestgaard@regionh.dk
Accepted: 25 September 2019
Keywords: anaemia and coagulation; peri-operative risk of MI; transfusion
*Presented in part at the Danish Society of Anaesthesiology and Intensive Care Annual Meeting, Copenhagen, Denmark,
November 2018.
Twitter: @hayfarani

Introduction superior, to liberal transfusion strategies in various patient

Multiple randomised controlled trials and systematic populations, including patients who are critically ill,
reviews have shown that restrictive transfusion is equal, or undergoing cardiac surgery, orthopaedic surgery and

© 2019 Association of Anaesthetists 1

Anaesthesia 2019
following major burns [1–8] (Supporting Information,
Fig. S1). Transfusion practices have changed over the last
two decades in accordance with this evidence, and
guidelines published after 2011 do not recommend red
blood cell transfusion for non-acute, haemodynamically
1
stable patients with haemoglobin levels above 60–80 g.l
[9–12]. Few studies have described the trends of red blood
cell transfusion in patients undergoing surgery. A decrease
in transfusion has been observed in abdominal surgery for
malignancy, but data on red blood cell transfusion are
lacking for most other surgical subspecialties [13].
Our primary objective was to evaluate the changes in
rates of peri-operative red blood cell transfusion over a
period of 6 years. Our secondary objectives were to
determine the prevalence and effect of factors that could
influence any changes in the rates of peri-operative red
blood cell transfusion, including pre-operative red cell
transfusion, prevalence of bleeding disorders and
coagulopathy, and minimally invasive procedures (as
opposed to open procedures). Additionally, we wished to
evaluate the risk of peri-operative myocardial infarction,
stroke and all-cause 30-day mortality over the study period.
We hypothesised that a reduction in red blood cell
transfusion rates would not be associated with an increase
in adverse clinical outcomes.
Methods
This study was approved by the Massachusetts General
Hospital Institutional Review Board and we have reported
Figure 1 Flow diagram of the patient cases from the American College of Surgeons National Surgical Quality Improvement
Program that were included in the study. All patients with available information on sex, age and ethnicity from 2011 to 2016 were
included.
2 © 2019 Association of Anaesthetists
Nordestgaard et al. | Transfusion trends in the USA
our findings in accordance with strengthening the reporting
of observational studies in epidemiology (STROBE)
guidelines. We included patients with available data on age,
sex and surgical specialty from the database of the
American College of Surgeons National Surgical Quality
Improvement Program (ACS NSQIP) for the period from 01
January 2011 to 31 December 2016 (Fig. 1). The database
contains information on surgical cases from up to 680
participating hospital sites across the US collected by
certified reviewers [14]. Both elective and emergency cases
were included in the analyses. We obtained estimates of all
surgical procedures in the US using the publicly available
discharge data from the National Inpatient Sample, the
Healthcare Cost and Utilization Project and the Agency for
Healthcare and Research and Quality [15].
We defined a peri-operative red blood cell transfusion
as a transfusion of at least one packed or whole red blood
cell unit either intra-operatively or up to 72 h
postoperatively from the index operation. We collected
data on pre-operative red blood cell transfusion, bleeding
disorders, coagulopathy and minimally invasive procedures
(as opposed to open procedures) over the study period.
Bleeding disorders included vitamin K deficiency,
haemophilia, thrombocytopenia and anticoagulation
therapy that was not stopped before surgery. Coagulopathy
was assessed using pre-operative haematocrit, international
normalised ratio (INR) and partial thromboplastin time. To
examine the use of open vs. minimally invasive techniques,
we focused on three common types of operation: open vs.
Nordestgaard et al. | Transfusion trends in the USA
endovascular repair of abdominal aortic aneurysm; open vs.
laparoscopic colectomy; and open vs. laparoscopic/robotic
assisted prostatectomy. The procedures were captured
using Current Procedural Terminology coding (Supporting
Information, Table S1). We also collected data on ethnicity,
comorbidities (Supporting Information, Table S2) and type
of anaesthesia: general; monitored anaesthesia care;
epidural/spinal/regional/local; or other/none.
We chose myocardial infarction, stroke and all-cause
30-day mortality as our clinical outcomes of interest to
reflect the potential adverse effects of a restrictive red blood
cell transfusion strategy. These outcomes were registered
by certified ACS NSQIP reviewers [16]. Myocardial infarction
was defined as a condition with either ECG changes
indicative of acute infarction or a rise in troponin levels
greater than three times the upper level of the institutional
reference value in the setting of suspected myocardial
ischemia. Stroke was defined as an embolic or thrombotic
cerebrovascular accident or a stroke with sudden motor,
sensory or cognitive dysfunction that persisted for more
than 24 h. Outcome data were recorded until 30 days after
the index operation for all patients and were registered in
ACS NSQIP. We used Work Relative Value Unit based on
each Current Procedural Terminology code (Centers for
Medicare and Medical Services Resource Based Relative
Value Scale) as an indicator of operative complexity as done
previously [17, 18].
We assumed that missing continuous variables were
missing at random and imputed them using linear
regression analyses with age and sex as predictors (body
mass index (BMI): 88,038/4,273,168; 2%). Most patients did
not have all laboratory tests performed (15–45% according
to the specific tests) [14]. To calculate mean values, only
patients with information on the specific test were included.
When laboratory tests were included as covariates, missing
values were imputed to complete the multivariable
analyses. The imputation was based on the assumption that
missing laboratory tests were not performed as they were
not clinically-indicated, and thus were most likely to be in
the normal range. We coded missing comorbidity variables
as ‘no comorbidity’ (5/4,273,168; < 0.1%). Other missing
categorical variables were grouped separately in the
analyses (ethnicity: 510,984/4,273,618 (12%); ASA score:
12,153/4,273,618 (0.3%); anaesthesia type: 1222/4,273,168
(< 0.1%). We grouped BMI in deciles when entered into the
multivariable regression analyses due to the large data set
and the infinite number of possible BMI values.
We first calculated peri-operative red blood cell
transfusion rates over the study period in all patients and
subsequently for each subspecialty. Secondly, we
© 2019 Association of Anaesthetists
Anaesthesia 2019
calculated pre-operative red blood cell transfusion rates,
prevalence of bleeding disorders and coagulopathy and
the proportion of minimally invasive procedures for each
year in all patients and in each surgical subspecialty.
Thirdly, we calculated the OR (95%CI) for receiving a
peri-operative red blood cell transfusion for the period
2012–2016 compared with 2011, using multivariable
logistic regression analyses adjusted for the following
variables: age; sex; BMI; ethnicity; smoking status;
diabetes; comorbidites; ASA physical status, pre-
operative red blood cell transfusion; bleeding disorders;
pre-operative haematocrit; pre-operative INR; pre-
operative partial thromboplastin time; Work Relative
Value unit; type of anaesthesia; duration of surgery; and
surgical specialty. This analysis was done in all patients
and in sub-groups of patients who received a pre-
operative red blood cell transfusion; those with bleeding
disorders; according to haematocrit levels; and in
patients undergoing open vs. minimally invasive
procedures.
Finally, we evaluated the trends in peri-operative
myocardial infarction, stroke and all-cause 30-day
mortality during the study period. We used Fine and Gray
competing risk regressions with postoperative days as
time scale (up to 30 days) and adjusted as above to
calculate the hazard ratios (95%CI) for ischaemic events in
2012–2016 compared with 2011 [19]. This regression
model was used to account for time since the index
operation and the competing risk of death. Patients who
died within 30 days from the index operation could not
experience an ischaemic event during the rest of the
follow-up period, making death a competing event for
ischaemic outcomes. Similar Cox regressions were used
to calculate hazard ratios for all-cause mortality.
We performed all analyses using STATA/SE 14.2
(StataCorp LP, College Station, TX, USA). We considered a
two-sided p value of less than 0.05 to be statistically
significant. Test for trend was by Cuzick’s extension of the
Wilcoxon rank sum test. To test for trend, we included the
year of the procedure in the regression analysis as a
continuous variable.
Results
A total of 4,273,168 surgical patients with 257,235 peri-
operative red blood cell transfusions were included in the
study (Table 1) (Fig. 1). Peri-operative red blood cell
transfusion rates declined from 37,040/441,255 (8.4%) in
2011 to 46,845/1,000,195 (4.6%) in 2016 (p < 0.001). The
same pattern was observed across all surgical
subspecialties (p = 0.027 for all specialties) (Figs. 2 and 3).
3
Anaesthesia 2019 Nordestgaard et al. | Transfusion trends in the USA

Table 1 Characteristics of the 4,273,168 surgical cases from the American College of Surgeons National Surgical Quality
Improvement Program 2011–2016 by procedure year. Values are mean (SD), median (IQR [range]) or number (proportion).
Procedure year 2011 2012 2013 2014 2015 2016 All p value

Individuals 441,255 543,814 651,764 750,791 885,349 1,000,195 4,273,168

Age; years 56 (17) 57 (17) 57 (17) 56 (17) 56 (17) 57 (17) 57 (17) < 0.001
Sex; male 189,026 232,163 276,710 324,125 384,005 434,783 1,840,812 < 0.001
(42.8%) (42.7%) (43.5%) (43.2%) (43.4%) (43.5%) (43.1%)
Body mass 30 (8) 30 (8) 30 (8) 30 (8) 30 (8) 30 (8) 30 (8) < 0.001
2
index; kg.m
Ethnicity < 0.001
White 332,345 403,718 483,948 551,803 640,517 705,368 3,117,699
(84.0%) (83.9%) (83.3%) (82.5%) (82.5%) (82.1%) (82.9%)
American Indian 2512 (0.6%) 2639 (0.6%) 4205 (0.7%) 5390 (0.8%) 5563 (0.7%) 5408 (0.6%) 25,726 (0.7%)
or Alaska Native
Asian or Pacific 13,923 (3.5%) 18,759 (3.9%) 21,567 (3.7%) 24,183 (3.6%) 27,471 (3.5%) 30,272 (3.5%) 136,175 (3.6%)
Islander
Black 41,878 (10.6%) 50,557 (10.5%) 62,767 (10.8%) 75,876 (11.4%) 88,936 (11.5%) 100,507 (11.7%) 420,521 (11.2%)
Hispanic 5093 (1.3%) 5688 (1.2%) 8436 (1.5%) 11258 (1.7%) 13635 (1.8%) 17953 (2.1%) 62,063 (1.7%)
Diabetes 67,446 (15.3%) 82,348 (15.1%) 99,985 (15.3%) 114,838 (15.3%) 136,865 (15.5%) 156,468 (15.6%) 657,950 (15.4%) < 0.001
Chronic obstructive 21,124 (4.8%) 25,054 (4.6%) 30,229 (5%) 34,304 (4.6%) 40,980 (4.6%) 43,711 (4.4%) 195,402 (4.6%) < 0.001
pulmonary disease
Congestive 3847 (0.9%) 4119 (0.8%) 5421 (0.8%) 6626 (0.9%) 8221 (0.9%) 8800 (0.9%) 37,034 (0.9%) < 0.001
heart failure
Ascites 2506 (0.6%) 2495 (0.5%) 2616 (0.4%) 2832 (0.4%) 3182 (0.4%) 3016 (0.3%) 16,647 (0.4%) < 0.001
Dialysis 7106 (1.6%) 7631 (1.4%) 9062 (1.4%) 10,100 (1.4%) 11,985 (1.4%) 13,070 (1.3%) 58,954 (1.4%) < 0.001
Disseminated cancer 9335 (2.1%) 11,287 (2.1%) 14,846 (2.3%) 17,846 (2.4%) 20,472 (2.3%) 22,355 (2.3%) 96,141 (2.3%) < 0.001
Pre-operative INR 1.0 (1.0–1.1 1.0 (1.0–1.1 1.0 (1.0–1.1 1.0 (1.0–1.1 1.0 (1.0–1.1 1.0 (1.0–1.1 1.0 (1.0–1.1 < 0.001
[0.1–10.0]) [0.1–10.0]) [0.1–10.0]) [0.1–10.0]) [0.1–10.0]) [0.1–10.0]) [0.1–10.0])
Pre-operative PTT; s 29 (27–32 29 (27–32 29 (27–32 29 (27–32 29 (27–33 29 (27–33 29 (27–32 < 0.001
[5–222]) [5–120]) [5–120]) [5–120]) [5–120]) [5–120]) [5–222])
ASA physical status < 0.001
1 40,452 (9.2%) 51,837 (9.6%) 59,687 (9.2%) 69,682 (9.3%) 79,480 (9.0%) 86,033 (8.6%) 387,171 (9.1%)
2 199,211 (45.3%) 248,051 (45.8%) 294,570 (45.3%) 337,651 (45.1%) 396, 530 (44.9%) 444,636 (44.6%) 1,920,649 (45.1%)
3 179,226 (38.7%) 208,595 (38.5%) 255,546 (39.3%) 295,947 (39.5%) 352,654 (39.9%) 404,725 (40.6%) 1,687,693 (39.6%)
4 28,855 (6.6%) 32,639 (6.0%) 39,007 (6.0%) 44,171 (5.9%) 52,638 (6.0%) 60,258 (6.0%) 257,568 (6.0%)
5 1005 (0.2%) 1020 (0.2%) 1165 (0.2%) 1343 (0.2%) 1591 (0.2%) 1810 (0.2%) 7934 (0.2%)
Type of anaesthesia < 0.001
General anaesthesia 399,474 (90.6%) 491,039 (90.3%) 587,349 (90.1%) 675,128 (89.9%) 788,688 (89.1%) 888,527 (88.9%) 3,830,205 (89.6%)
Monitored 17,188 (3.9%) 21,341 (3.9%) 28,278 (4.3%) 33,330 (4.4%) 41,315 (4.7%) 47,345 (4.7%) 188,806 (4.4%)
anaesthesia care
Epidural/spinal/ 23,927 (5.4%) 30,934 (5.7%) 35,464 (5.4%) 41,392 (5.5%) 54,367 (6.1%) 62,931 (6.3%) 249,015 (5.8%)
regional/local
Other/none 356 (0.1%) 468 (0.1%) 579 (0.1%) 776 (0.1%) 639 (0.1%) 1102 (0.1%) 3,920 (0.1%)
Duration of 88 (51–148 86 (50–144 86 (50–143 86 (50–142 84 (50–139 84 (50–138 85 (50–141 < 0.001
surgery; min [0–1434]) [0–1440]) [0–1440]) [0–1440]) [0–1440]) [0–1429]) [0–1440])
Work relative 16 (10–23 15 (10–23 15 (10–23 15 (9–21 15 (9–21 15 (10–21 15 (10–21 < 0.001
value unit [0–109]) [0–93]) [0–109]) [0–93]) [0–109]) [0–93]) [0–109])
Surgical specialty < 0.001
Cardiac 3472 (0.8%) 4028 (0.7%) 3093 (0.5%) 3693 (0.5%) 4033 (0.5%) 4281 (0.4%) 22,600 (0.5%)
General 239,962 (54.4%) 277,925 (51.1%) 322,056 (49.4%) 360,397 (48.0%) 409,230 (46.2%) 445,639 (44.6%) 2,055,211 (48.1%)
Gynaecological 27,211 (6.2%) 36,941 (6.8%) 46,644 (7.2%) 55,339 (7.4%) 65,653 (7.4%) 77,744 (7.8%) 309,533 (7.2%)
Neurological 17,910 (4.1%) 23,585 (4.3%) 32,640 (5.0%) 37,442 (5.0%) 46,665 (5.3%) 53,127 (5.3%) 211,369 (5.0%)
Orthopaedic 63,103 (14.3%) 92,591 (17.0%) 122,033 (18.7%) 153,320 (20.4%) 197,868 (22.4%) 235,029 (23.5%) 863,944 (20.2%)

(continued)

4 © 2019 Association of Anaesthetists

Nordestgaard et al. | Transfusion trends in the USA Anaesthesia 2019

Table 1 (continued)

Procedure year 2011 2012 2013 2014 2015 2016 All p value

Ear, nose and 10,891 (2.5%) 14,621 (2.7%) 17,001 (2.6%) 20,859 (2.8%) 24,658 (2.8%) 28,147 (2.9%) 116,177 (2.8%)
throat
Plastic 10,669 (2.4%) 14,777 (2.7%) 18,440 (2.8%) 21,108 (2.8%) 24,596 (2.8%) 29,382 (2.9%) 118,972 (2.8%)
Thoracic 5552 (1.3%) 7397 (1.4%) 8375 (1.3%) 9420 (1.4%) 9788 (1.1%) 11,323 (1.1%) 51,855 (1.2%)
Urologic 22,234 (5.0%) 29,044 (5.3%) 34,013 (5.2%) 39,632 (5.3%) 49,184 (5.6%) 57,963 (5.8%) 232,070 (5.4%)
Vascular 40,251 (9.1%) 42,904 (7.9%) 47,467 (7.3%) 49,581 (6.6%) 53,674 (6.1%) 57,560 (5.8%) 291,437 (6.8%)

INR, international normalised ratio; PTT, partial thromboplastin time.

Pre-operative red blood cell transfusion rates declined
from 6657/441,255 (1.5%) in 2011 to 8515/1,000,195 (0.9%)
in 2016 (p < 0.001) (Fig. 4, upper left panel). The
prevalence of bleeding disorders decreased from 22,926/
441,255 (5.2%) in 2011 to 40,386/1,000,195 (4.0%) in 2016
(p < 0.001) (Fig. 4, upper right panel). Mean (SD) pre-
operative haematocrit increased from 38.9 (5.3) % in 2011
to 39.7 (5.3) % in 2016 (p < 0.001) (Fig. 4, lower left panel).
We observed no major changes in median INR and partial
thromboplastin time over the study period (Supporting
Information, Fig. S2). In abdominal aortic aneurysm repair,
colectomy and prostatectomy, the proportion of minimally
invasive techniques increased from 2873/4310 (67%) in
2011 to 3484/4840 (72%) in 2016, from 11,639/27,964
(42%) to 30,957/57,887 (53%) and from 4394/5721 (77%) to
8573/10,719 (80%), respectively (p < 0.001) (Fig. 4, lower
right panel). Trends in pre-operative red blood cell
transfusions, bleeding disorders and haematocrit levels in
each subspecialty are presented in the Supporting
Information (Supporting Information, Figs. S3–S5).
Compared with 2011, the corresponding adjusted OR
(95%CI) for red blood cell transfusion decreased gradually
Figure 2 Peri-operative red blood cell transfusion rates
from 2011 to 2016. Dotted lines represent 95%CI.
from 0.88 (0.86–0.90) in 2012 to 0.51 (0.50–0.51) in 2016
(p < 0.001) (Fig. 5). Similar patterns were observed in all
sub-groups when the patients were stratified according to
whether they required pre-operative red blood cell
transfusions, suffered from bleeding disorders, had low or
normal haematocrit levels or underwent minimally invasive
or open procedures (Supporting Information, Figs. S6 and
S7).
There were 19,466 peri-operative myocardial
infarction, 11,573 stroke and 52,977 all-cause mortality
events within 30 days of the index operation. Compared
with 2011, the adjusted hazard ratios (95%CI) in 2012 and
2016 were 0.96 (0.90–1.02) and 1.05 (0.99–1.11) for
myocardial infarction (Fig. 6, upper panel); 0.91 (0.83–0.99)
and 0.99 (0.92–1.07) for stroke (Fig. 6, middle panel); and
0.98 (0.94–1.02) and 0.99. (0.96–1.03) for mortality (Fig. 6,
lower panel). Hazard ratios for the outcomes in each surgical
subspecialty are presented in the Supporting Information
(Figs. S8–S10).
Discussion
In 4,273,168 surgical patients from across the USA, peri-
operative red blood cell transfusion rates declined from
37,040/441,255 (8.4%) to 46,845/1,000,195 (4.6%) between
2011 and 2016. This was consistent across all surgical
subspecialties, and even after we accounted for variation in
population and procedure characteristics. The lower peri-
operative red blood cell transfusion rates were not
associated with increasing risk of peri-operative myocardial
infarction, stroke or all-cause mortality.
Three studies on specific patient populations have
reported contradictory trends in peri-operative red blood
cell transfusion. Among patients who underwent
abdominal surgery for malignancy, Ecker et al. reported
an approximately 20–35% decrease in peri-operative red
blood cell transfusion between 2005 and 2013 [13]. In a
cohort of hospitalised patients, Roubinian et al. reported
a combined 22% decrease in red blood cell transfusion
between 2009 and 2013 [20]. On the other hand, Shehata
et al. observed increasing red blood cell transfusions

© 2019 Association of Anaesthetists 5

Anaesthesia 2019 Nordestgaard et al. | Transfusion trends in the USA

between 2006 and 2012 in a similar patient group, but a
decrease in the total number of red blood cell units
transfused [21]. We observed an approximately 45%
decrease in peri-operative red blood cell transfusions
over a 6-year period, which represents a greater change
than previously reported. This difference could be related
to the inclusion of patients undergoing procedures with
higher transfusion rates compared with those undergoing
general surgery (e.g. a 64% decrease in orthopaedic
surgery). When we restricted the analyses to general
surgery patients, the decrease in peri-operative red blood
cell transfusions was 36%, which is similar to the trend
reported by Ecker et al. although within a shorter time
span [13].
There are many potential reasons for the decrease in
peri-operative red blood cell transfusions, including
adoption of restrictive transfusion practices by adhering to
current guidelines. We found that the proportion of patients
suffering from bleeding disorders decreased over the study
period, and that minimally invasive procedures were more
commonly observed in patients undergoing abdominal
aortic aneurysm repair, colectomy and prostatectomy.
Increasing pre-operative haematocrit levels, through
optimisation of pre-operative anaemia, may also have
contributed to less transfusion. After adjusting for variations
in baseline characteristics, comorbidities, laboratory tests
and procedure type and complexity over the study period,
we still observed a trend towards a lower risk of requiring a
peri-operative red blood cell transfusion compared with
2011. Further support for restrictive transfusion practices
includes evidence that haematocrit levels among patients
who received either pre-, intra- or postoperative red cell

Figure 3 Peri-operative red blood cell transfusion rates in each surgical subspecialty from 2011 to 2016. Dotted lines represent
95%CI. (a) Cardiac surgery (p = 0.027); (b) general surgery (p < 0.001); (c), gynaecological surgery (p < 0.001); (d) neurological
surgery (p < 0.001); (e) orthopaedic surgery (p < 0.001); (f) ear, nose and throat surgery (p < 0.001); (g) plastic surgery
(p < 0.001); (h) thoracic surgery (p < 0.001); (i) urologic surgery (p < 0.001); (j) vascular surgery (p < 0.001).

6 © 2019 Association of Anaesthetists

Nordestgaard et al. | Transfusion trends in the USA Anaesthesia 2019

Figure 4 Rates of pre-operative red blood cell transfusion, prevalence of bleeding disorders, pre-operative haematocrit levels
in all surgical subspecialties and rates of minimally invasive procedures from 2011 to 2016. Dotted lines represent 95%CI.
(a) pre-operative red blood cell transfusion (p < 0.001); (b) bleeding disorders (p < 0.001); (c) pre-operative haematocrit levels
(solid line, all patients; small dots, no transfusion; long dots, any transfusion) (p all groups < 0.001); (d) minimally invasive
procedures (solid line, abdominal aortic aneurysm repair; small dots, colectomy; long dots, prostatectomy) (p for all groups
< 0.001).

had not changed since 2011 (8.4%), the corresponding

Figure 5 Odds ratios for peri-operative red blood cell
transfusions for 2012–2016 vs. 2011. Error bars represent
95%CI.
transfusion (not including patients who did not receive
transfusions) declined slightly over the study period (Fig. 4,
lower left panel).
In 2011, there were approximately 10,701,342 surgical
discharges in the USA reported from the National Inpatient
Sample. If the transfusion rates reported in this study are
reliable estimates of national practice, 898,396 peri-operative
red cell transfusions took place nationwide in 2011. If the rate
number of transfusions would have been 807,768 in 2016 (of
9,609,921 estimated discharges). The actual rate of 4.6%
equals approximately 450,089 peri-operative red cell
transfusions or 356,679 fewer transfusions than if the rate had
remained unchanged since 2011. Even if we assume that each
transfusion consisted of only one red blood cell unit, the
estimated decline in red cell transfusion corresponds to 5945
fewer febrile reactions in 2016, given a rate of 1:60 reactions
per red blood cell unit; 3567 fewer circulatory overloads, given
a rate of 1:100; and 1427 fewer allergic reactions, given a rate
of 1:250 [10, 22–29]. Assuming a total cost of £602 ($761) per
red blood cell unit, the decline would have resulted in a
decreased nationwide healthcare expenditure of
approximately £214,775,759 ($271,433,412) [30]. For the
combined period from 2012 to 2016, the decline in
transfusions corresponds to 20,184 fewer febrile reactions;
12,110 fewer circulatory overloads; 4844 fewer allergic
reactions; and a decreased expenditure of approximately
£729,231,076 ($921,601,582). Since a mean number of 2.2 red
blood cell units were given per transfusion episode from 2007
to 2009 in ACS NSQIP, these figures are likely to be
underestimates.
The limitations of our study include the need to impute
missing variables. This may be counterbalanced by the fact
that this represents the largest sample size to date, and the

© 2019 Association of Anaesthetists 7

Anaesthesia 2019
Figure 6 Hazard ratios for peri-operative myocardial
infarction, stroke and all-cause mortality within 30 days of
the index operation for 2012–2016 vs. 2011. Error bars
represent 95%CI. (a) myocardial infarction (p < 0.001);
(b) stroke (p = 0.262); (c) mortality (p = 0.751).
high quality of data that are available from ACS NSQIP.
Since ACS NSQIP is based on data from engaged
institutions that actively contribute data, it is possible that the
results cannot be generalised to other institutions in the USA
or to other countries. We were unable to estimate the mean
number of red blood cell units given per transfusion, since
this information was not consistently collected after 2009.
Between 2007 and 2009, the median (IQR [range]) number
of units given per transfusion declined somewhat from 2 (1–
3 [1–41]) in 2007 to 2 1–3 [1–42]) in 2009; p < 0.001), which
has also been observed elsewhere [21]. Assuming that this
trend has continued since, the decrease in the actual
number of red blood cell units transfused is likely to be
higher than we have estimated using red blood cell
transfusion rates. Although information on peri-operative
red blood cell transfusion between 2007 and 2010 is
available in ACS NSQIP, we did not include these years due
8 © 2019 Association of Anaesthetists
Nordestgaard et al. | Transfusion trends in the USA
to the lack of a standardised definition of a peri-operative
red blood cell transfusion before 2011. After 2011, the
definition was revised to include any red blood transfusions
for up to 72 h after the index operation, whereas only intra-
operative red blood cell transfusions were registered in the
earlier years. The coagulation tests that we used are
recognised to be poor predictors of coagulopathy, and we
were not able to estimate pre-operative haemoglobin levels
over the study period. Data on the use of tranexamic acid,
cell savage or other methods that might reduce the need for
red blood cell transfusions were also not available.
In conclusion, peri-operative red blood cell
transfusions in the USA declined between 2011 and 2016.
This was consistent across all surgical specialties and after
adjustment for multiple confounders. The decrease was not
associated with an increased risk of peri-operative
myocardial infarction, stroke or death from any cause.
Acknowledgements
Data were provided by the American College of Surgeons
National Surgical Quality Improvement Program and the
participating hospitals; they have not verified and are not
responsible for the statistical validity of the data analyses or the
conclusions derived by the authors. The study was supported
by Massachusetts General Hospital, Harvard Medical School
and Rigshospitalet, University of Copenhagen. The funding
sources were not involved in the design of the study; in the
collection, analysis and interpretation of data; in the writing of
the manuscript; or in the decision to submit the article for
publication. No competing interests declared.
References
1. Hill SR, Carless PA, Henry DA, et al. Transfusion thresholds and
other strategies for guiding allogeneic red blood cell transfusion.
Cochrane Database of Systematic Reviews 2002; 4: CD002042.
2. Carless PA, Henry DA, Carson JL, Hebert PP, McClelland B, Ker
K. Transfusion thresholds and other strategies for guiding
allogeneic red blood cell transfusion. Cochrane Database of
Systematic Reviews 2010; 10: CD002042.
3. Carson JL, Carless PA, Hebert PC. Transfusion thresholds and
other strategies for guiding allogeneic red blood cell transfusion.
Cochrane Database of Systematic Reviews 2012; 4: CD002042.
4. Carson JL, Stanworth SJ, Roubinian N, et al. Transfusion
thresholds and other strategies for guiding allogeneic red
blood cell transfusion. Cochrane Database of Systematic
Reviews 2016; 10: CD002042.
5. Hebert PC, Wells G, Blajchman MA, et al. A multicenter,
randomized, controlled clinical trial of transfusion requirements
in critical care. Transfusion Requirements in Critical Care
Investigators, Canadian Critical Care Trials Group. New
England Journal of Medicine 1999; 340: 409–17.
6. Carson JL, Terrin ML, Noveck H, et al. Liberal or restrictive
transfusion in high-risk patients after hip surgery. New England
Journal of Medicine 2011; 365: 2453–62.
7. Mazer CD, Whitlock RP, Fergusson DA, et al. Restrictive or
liberal red-cell transfusion for cardiac surgery. New England
Journal of Medicine 2017; 377: 2133–44.
Nordestgaard et al. | Transfusion trends in the USA
8. Palmieri TL, Holmes JH, Arnoldo B, et al. Transfusion
Requirement in Burn Care Evaluation (TRIBE): a multicenter
randomized prospective trial of blood transfusion in major burn
injury. Annals of Surgery 2017; 266: 595–602.
9. American Society of Anesthesiologists Task Force on
Perioperative Blood M. Practice guidelines for perioperative
blood management: an updated report by the American
Society of Anesthesiologists Task Force on Perioperative Blood
Management. Anesthesiology 2015; 122: 241–75.
10. Carson JL, Guyatt G, Heddle NM, et al. Clinical practice
guidelines from the AABB: red blood cell transfusion
thresholds and storage. Journal of the American Medical
Association 2016; 316: 2025–35.
11. Kozek-Langenecker SA, Ahmed AB, Afshari A, et al.
Management of severe perioperative bleeding: guidelines
from the European Society of Anaesthesiology: First update
2016. European Journal Anaesthesiology 2017; 34: 332–95.
12. Klein AA, Arnold P, Bingham RM, et al. AAGBI guidelines: the
use of blood components and their alternatives 2016.
Anaesthesia 2016; 71: 829–42.
13. Ecker BL, Simmons KD, Zaheer S, et al. Blood transfusion in major
abdominal surgery for malignant tumors: a trend analysis using the
national surgical quality improvement program. Journal of the
American Medical Association Surgery 2016; 151: 518–25.
14. American College of Surgery. User guide for the ACS NSQIP
participant use data file. 2016. https://www.facs.org/quality-progra
ms/acs-nsqip/participant-use (accessed 12/12/2018).
15. HCUPnet. Healthcare Cost and Utilization Project (HCUP). Agency
for Healthcare Research and Quality, Rockville, MD. 2006-2009.
https://hcupnet.ahrq.gov/#setup (accessed 12/12/2018).
16. Murphy GJ, Reeves BC, Rogers CA, Rizvi SI, Culliford L, Angelini
GD. Increased mortality, postoperative morbidity, and cost
after red blood cell transfusion in patients having cardiac
surgery. Circulation 2007; 116: 2544–52.
17. Bohnen JD, Mavros MN, Ramly EP, et al. Intraoperative
Adverse events in abdominal surgery: what happens in the
operating room does not stay in the operating room. Annals of
Surgery 2017; 265: 1119–25.
18. Davenport DL, Henderson WG, Khuri SF, Mentzer RM Jr.
Preoperative risk factors and surgical complexity are more
predictive of costs than postoperative complications: a case
study using the National Surgical Quality Improvement Program
(NSQIP) database. Annals of Surgery 2005; 242: 463–8.
19. Fine JP, Gray RJ. A proportional hazards model for the
subdistribution of a competing risk. Journal of the American
Statistical Association 1999; 94: 496–509.
20. Roubinian NH, Escobar GJ, Liu V, et al. Trends in red blood cell
transfusion and 30-day mortality among hospitalized patients.
Transfusion 2014; 54: 2678–86.
21. Shehata N, Forster A, Lawrence N, et al. Changing trends in
blood transfusion: an analysis of 244,013 hospitalizations.
Transfusion 2014; 54: 2631–9.
22. US Food and Drug Administration. Transfusion/donation fatalitis:
notification process for transfusion related fatalities and donation
related deaths. 2015. https://www.fda.gov/vaccines-blood-
biologics/report-problem-center-biologics-evaluation-researc
h/transfusiondonation-fatalities (accessed 09/10/2019).
23. Federowicz I, Barrett BB, Andersen JW, Urashia M, Popovsky
MA, Anderson KC. Characterization of reactions after
transfusion of cellular blood components that are white cell
reduced before storage. Transfusion 1996; 36: 21–8.
24. Popovsky MA, Audet AM, Andrzejewski C Jr. Transfusion-
associated circulatory overload in orthopedic surgery patients: a
multi-institutional study. Immunohematology 1996; 12: 87–9.
25. Clifford L, Jia Q, Subramanian A, et al. Characterizing the
epidemiology of postoperative transfusion-related acute lung
injury. Anesthesiology 2015; 122: 12–20.
© 2019 Association of Anaesthetists
Anaesthesia 2019
26. DeBaun MR, Gordon M, McKinstry RC, et al. Controlled trial of
transfusions for silent cerebral infarcts in sickle cell anemia.
New England Journal of Medicine 2014; 371: 699–710.
27. The Sanguis Study Group. Use of blood products for elective
surgery in 43 European hospitals. Transfusion Medicine 1994; 4:
251–68.
28. Zou S, Dorsey KA, Notari EP, et al. Prevalence, incidence, and
residual risk of human immunodeficiency virus and hepatitis C virus
infections among United States blood donors since the introduction
of nucleic acid testing. Transfusion 2010; 50: 1495–504.
29. Stramer SL, Notari EP, Krysztof DE, Dodd RY. Hepatitis B virus
testing by minipool nucleic acid testing: does it improve blood
safety? Transfusion 2013; 53: 2449–58.
30. Shander A, Hofmann A, Ozawa S, Theusinger OM, Gombotz H,
Spahn DR. Activity-based costs of blood transfusions in surgical
patients at four hospitals. Transfusion 2010; 50: 753–65.
Supporting Information
Additional supporting information may be found online in
the Supporting Information section at the end of the article.
Table S1. Procedural terminology codes used to
capture abdominal aortic aneurysm repair, colectomy and
prostatectomy procedures.
Table S2. Detailed diagnostic criteria for the
comorbidities adjusted for in the multivariable analyses.
Figure S1. Timeline of publication years of selected
major clinical trials and guidelines on red blood cell
transfusion strategies.
Figure S2. Median (IQR) pre-operative international
normalised ratio and partial thromboplastin time from 2011
to 2016.
Figure S3. Pre-operative red blood cell transfusion
rates in each surgical subspecialty.
Figure S4. Prevalence of bleeding disorders in each
surgical subspecialty.
Figure S5. Mean pre-operative haematocrit in each
surgical subspecialty.
Figure S6. Odds Ratios for peri-operative red cell
transfusions for 2012–2016 vs. 2011 in sub-groups of pre-
operative red blood cell transfusion, bleeding disorders
and haematocrit levels.
Figure S7. Odds Ratios for peri-operative red blood
cell transfusion for 2012–2016 vs. 2011 in minimally invasive
and open abdominal aortic aneurysm repair, colectomy and
prostatectomy procedures.
Figure S8. Hazard ratios for intra- and postoperative
myocardial infarction in 2012–2016 vs. 2011 in each surgical
subspecialty.
Figure S9. Hazard ratios for intra- and postoperative
stroke in 2012–2016 vs. 2011 in each surgical subspecialty.
Figure S10. Hazard ratios for intra- and post-operative
all-cause mortality in 2012–2016 vs. 2011 in each surgical
subspecialty.
9
Anaesthesia 2019 doi:10.1111/anae.14862

Original Article

The association of allogeneic blood transfusion and the

recurrence of hepatic cancer after surgical resection
Y. H. Tai,1 H. L. Wu,2 M. S. Mandell,3 M. Y. Tsou4 and K. Y. Chang5

1 Lecturer, Department of Anaesthesiology, School of Medicine, College of Medicine, Taipei Medical University, Taipei,
Taiwan
2 Lecturer, School of Medicine, National Yang-Ming University, Taipei, Taiwan
3 Professor, Department of Anaesthesiology, University of Colorado Health Sciences Center, Aurora, Colorado, USA
4 Professor, School of Medicine, National Yang-Ming University, Taipei, Taiwan
5 Associate Professor, School of Medicine, National Yang-Ming University, Taipei, Taiwan

Summary
There is conflicting evidence whether allogeneic blood transfusion influences survival or cancer recurrence
after resection of hepatocellular cancer. We followed up 1469 patients who had undergone hepatocellular
resection for a median (IQR [range]) of 45 (21–78 [0–162]) months, of whom 626 (43%) had had blood
transfusion within 7 days of surgery. Both disease-free survival and patient survival were measured using a
proportional hazards regression model and inverse probability of treatment weighting. We used restricted
cubic splines for the association of the number of packed red blood cell units transfused with cancer recurrence
and survival. We found that peri-operative blood transfusion was independently associated with survival and
cancer recurrence after resection of hepatocellular carcinoma. Adjusted hazard ratios (95%CI) for the
association of blood transfusion with cancer recurrence and all-cause mortality were 1.3 (1.1–1.4) and 1.9 (1.6–
2.3), p < 0.001 for both. With more units transfused cancer recurrence was more likely and survival was shorter.
The association of the number of transfused units was non-linear for cancer recurrence and linear response for
all-cause mortality.

.................................................................................................................................................................
Correspondence to: K. Y. Chang
Email: kychang@vghtpe.gov.tw
Accepted: 5 September 2019
Keywords: cancer recurrence; hepatectomy; hepatocellular carcinoma

Introduction Data from the American College of Surgeons National

Hepatocellular carcinoma is the fourth most common cause Surgical Quality Improvement Program suggests
of death due to cancer globally, leading to approximately transfusion is an independent risk factor for higher
800,000 deaths each year [1]. Surgical resection is the unadjusted and adjusted mortality in patients undergoing
definitive treatment for patients with primary hepatocellular trans-abdominal oncological surgery [7]. However, it is
carcinoma and in cirrhotic patients without complications of difficult to separate effects of allogeneic blood
portal hypertension [2]. However recurrence occurs in up to administration on tumour recurrence from additional
54% of patients within 2 years of resection [3]. patient, surgical and tumour-modifying characteristics [8].
Factors associated with recurrence are mainly related We used a propensity scoring model to adjust our data
to tumour characteristics [4, 5]. However patients with for imbalances in baseline characteristics that could
hepatocellular carcinoma often have pre-operative anaemia influence outcomes. Our study was specifically designed to
and surgical blood loss, which may require transfusion [6]. isolate the effects of allogeneic blood transfusion on

© 2019 Association of Anaesthetists 1

Anaesthesia 2019
recurrence and survival rates in patients who underwent
liver resection for hepatocellular carcinoma and determine
if dose influences outcomes. We applied robust statistics to
a detailed list of clinicopathological features associated with
cancer recurrence to assess postoperative disease-free
duration and survival. Based on positive associations
observed in previous studies [9–11], we hypothesised that
blood transfusion increases the risk of hepatocellular
carcinoma recurrence and death. Also, we thought that the
amount of blood given could modify risk. The purpose of
our analysis is to provide healthcare providers with detailed
information about risk of transfusion in order to guide
clinical decision-making in the peri-operative period.
Methods
The Institutional Review Board of the Taipei Veterans
General Hospital approved this study and written informed
consent was waived. We anonymised data before analysis.
We reviewed the medical records of 2215 consecutive
patients who had undergone resection of hepatocellular
carcinoma from January 2005 to December 2016. We
excluded patients who had: repeat surgery; liver
transplantation; lymph node involvement or metastases; or
liver resection for benign lesions. We defined peri-operative
transfusion as allogeneic transfusion of packed red blood
cells during surgery or within 7 days after operation.
We recorded: patient characteristics; the severity of liver
cirrhosis and viral serology [12]; and pre-operative serum
aspartate aminotransferase, alanine aminotransferase and
alpha-fetoprotein [13]. We also recorded the size and
number of tumours [14], cellular differentiation,
microvascular invasion [14] and extracapsular invasion [15].
We recorded whether patients had pre-operative
radiofrequency ablation, transarterial chemoembolisation
and percutaneous ethanol injection. We recorded the extent
of resection (> 2 Couinaud liver segments or not), the
surgical margin, intra-operative blood loss, the use of
laparoscopic or robot-assisted techniques and anaesthesia
time. Transfusion and related laboratory data were collected,
including haemoglobin and platelet concentrations and the
international normalised ratio. A specialist anaesthetist not
involved in data analysis recorded these data. We randomly
sampled data to verify values.
Hepatic resections were performed by experienced
general surgeons. A Pringle’s manoeuvre was routinely
used to control blood loss. The liver parenchyma was
transected using a clamp-crushing technique, with argon
beam coagulation for haemostasis. Laparoscopic or robot-
assisted techniques were used in selected patients after
July 2011.
2 © 2019 Association of Anaesthetists
Tai et al. | Blood transfusion and recurrence of liver cancer
Postoperatively, serum alpha-fetoprotein was
measured every 4 months for 2–3 years, then every
6 months as part of routine surveillance. Imaging studies
(ultrasonography, computed tomography or magnetic
resonance imaging) were performed every 4 months for 2–
3 years, then every 6 months. Positron emission
tomography or bone scan was performed if recurrence or
distant metastases were suspected. The presence of
recurrent cancer was confirmed by biopsy if the diagnosis
was equivocal. Patients typically underwent postoperative
surveillance for recurrence for at least 5 years.
Data were collected up to the end of September 2018.
The primary outcome was disease-free survival; defined as
the time from the date of surgery to the date of first cancer
recurrence. The secondary outcome was overall survival,
defined as the time from the date of surgery to the date of
death. Survival times were the corresponding censored
observations in patients without recurrence or death.
The Kolmogorov–Smirnov and Shapiro–Wilk tests were
employed as a measure of normality. We used logarithms to
transform variables that were not normally distributed. We
used restricted cubic splines with three knots (placed at 1, 5
and 9 units of packed red blood cells) to assess non-linear
associations of transfusion with outcome [16]. We used
inverse probability treatment weighting for propensity
score to detect pseudopopulations and reduce
confounding from imbalances in variables [17]. We first
used logistic regression to estimate the probability of
having blood transfusion based on patient characteristics
(Table S2). Regression analysis was then weighted with the
inverse of estimated probability and the distribution was
truncated at 99% to reduce the impact of large weights. We
assessed the balance of covariates with standardised
differences and graphical comparisons [18]. We used Cox
regression to evaluate the association of blood transfusion
and other covariates with disease-free survival and overall
survival. For sensitivity analysis, multivariable Cox
regression analysis was implemented using a stepwise
model selection strategy with the entry and removal
significance criteria of 0.1 and 0.05, respectively, to
determine the factors associated with disease-free and
overall survival from significant predictors in the univariate
analysis. We considered p < 0.05 statistically significant. All
the statistical analyses were performed using SAS software,
version 9.4 (SAS Institute Inc., Cary, NC, USA). We used
Schoenfeld’s formula to estimate the sample size needed
for a proportional hazards model [19]. A total of 812 events
were required to attain a power of 0.8 assuming a type-1
error rate of 0.05, relative hazard of cancer recurrence
1.22 [11].
Tai et al. | Blood transfusion and recurrence of liver cancer Anaesthesia 2019

Results 5-year overall survival rates without and with peri-operative

We followed up 1469 patients who underwent transfusion were 83% (80–86%) and 67% (63–72%),
hepatocellular resection for a median (IQR [range]) of 45 respectively, p < 0.001. Unadjusted and adjusted mortality
(21–78 [0–162]) months, of whom 626 (43%) had had peri- hazards (95%CI) after transfusion were 2.3 (1.8–2.8) and 1.9
operative blood transfusion (Fig. 1). Differences in (1.6–2.3), respectively, p < 0.001 for both, consistent with
characteristics between patients who had transfusions and multivariable analysis, hazard ratio (95%CI) 1.4 (1.0–2.0),
those who did not were reduced by inverse probability p = 0.030. Tables 2 and 3 list univariate and multivariate
weighting (Table 1 and Fig. S1). associations of characteristics with cancer recurrence and
The mean (95%CI) unadjusted 5-year disease-free mortality. Variables independently associated with
survival rates without and with peri-operative transfusion peri-operative blood transfusion are listed in Table 4.
were 43% (39–46%) and 29% (25–33%), respectively,
p < 0.001. The adjusted hazard ratio (95%CI) for cancer Discussion
recurrence after blood transfusion was 1.3 (1.1–1.4), We found that peri-operative blood transfusion was
p < 0.001, consistent with multivariable analysis, hazard independently associated with survival and cancer
ratio (95%CI) 1.3 (1.1–1.6), p = 0.009. The relationship recurrence after resection of hepatocellular carcinoma.
between blood transfusion and cancer recurrence was non- With more units transfused cancer recurrence was more
linear (Fig. 2 and Table S3). The mean (95%CI) unadjusted likely and survival was shorter [20].

Figure 1 Flow diagram for patient selection.

© 2019 Association of Anaesthetists 3

Anaesthesia 2019 Tai et al. | Blood transfusion and recurrence of liver cancer

Table 1 Patient characteristics and their standardised differences between the transfused and non-transfused patients before
and after generation of pseudopopulations with inverse probability treatment weighting. Value are mean (SD) or number
(proportion).
Original populations Pseudopopulations after IPTW
Peri-operative transfusion Peri-operative transfusion
No Yes No Yes
Characteristic (n = 843) (n = 626) SDiff (n = 1014) (n = 1125) SDiff
Age; y 60 (13) 63 (13) 26.4 62 (12) 62 (13) 5.7
Sex; male 652 (77%) 473 (76%) 4.2 774 (76%) 848 (75%) 2.3
ASA ≥ 3 203 (24%) 233 (37%) 28.8 290 (29%) 408 (36%) 16.3
Aetiology of cancer
HBsAg positive 606 (72%) 376 (60%) 25.1 696 (69%) 751 (67%) 4.1
Anti-HCV Ab positive 173 (21%) 152 (24%) 9.0 238 (24%) 243 (22%) 4.5
Alcoholism 62 (7%) 43 (7%) 1.9 73 (7%) 57 (5%) 8.7
Liver cirrhosis 347 (41%) 292 (47%) 11.1 438 (43%) 510 (45%) 4.2
Child-Pugh class; B or C 17 (2%) 37 (6%) 20.1 18 (2%) 35 (3%) 8.8
Diabetes mellitus 188 (22%) 168 (27%) 10.6 234 (23%) 291 (26%) 6.6
Chronic kidney disease 65 (8%) 75 (12%) 14.4 79 (8%) 86 (8%) 0.3
Pre-operative blood tests
1
Haemoglobin; g.dl 13.8 (1.5) 12.8 (1.9) 55.7 13.4 (1.8) 13.1 (1.8) 17.3
Platelet count; 103 ll 1
177 (73) 183 (91) 7.9 172 (68) 171 (87) 0.2
International normalised ratio 1.08 (0.97) 1.06 (0.41) 2.8 1.05 (0.09) 1.06 (0.32) 3.9
1
Total bilirubin ≥ 1.0 mg.dl 188 (22%) 144 (23%) 1.7 240 (24%) 292 (26%) 5.2
1
AST > 40 IU.l 337 (40%) 341 (55%) 29.4 445 (44%) 559 (50%) 11.8
ALT > 40 IU.l 1
400 (47%) 300 (48%) 0.9 481 (48%) 530 (47%) 0.7
Alpha-fetoprotein > 20 ng.ml 1
404 (49%) 323 (53%) 7.1 523 (52%) 599 (53%) 3.3
1
Albumin ≤ 3.5 g.dl 41 (5%) 72 (12%) 24.5 55 (5%) 100 (9%) 13.3
1
Serum creatinine; lmol.l 87.5 (74.3) 95.5 (80.4) 10.1 87.9 (74.7) 88.4 (65.6) 0.7
Cancer characteristics
Tumour diameter > 5 cm 215 (26) 319 (51%) 54.3 317 (31%) 446 (40%) 17.6
Multiple nodules 169 (20%) 151 (24%) 9.8 213 (21%) 234 (21%) 0.4
Poor or undifferentiated histology 282 (34%) 227 (36%) 5.9 339 (34%) 372 (33%) 0.8
Microvascular invasion 547 (65%) 460 (74%) 18.7 712 (70%) 825 (73%) 7.1
Extracapsular invasion 345 (41%) 274 (44%) 5.8 401 (40%) 475 (42%) 5.3
Positive surgical margin 27 (3%) 63 (10%) 27.8 32 (3%) 64 (6%) 12.6
Pre-operative TACE/RFA/PEI 63 (8%) 68 (11%) 11.8 77 (8%) 98 (9%) 4.3
Operative variables
> two segments resected 253 (30%) 286 (46%) 32.7 324 (32%) 428 (38%) 12.8
Laparoscopic or robotic surgery 64 (8%) 26 (4%) 14.7 77 (8%) 104 (9%) 6.1
Blood loss; ml 453 (322) 1725 (1807) 98.0 576 (379) 1046 (1420) 45.2
Epidural analgesia 325 (39%) 236 (38%) 1.8 365 (36%) 389 (35%) 3.0
Anaesthesia time; min 328 (97) 420 (132) 79.7 346 (104) 385 (129) 32.9
Date of surgery; 2011–2016 454 (54%) 311 (50%) 8.4 595 (59%) 673 (60%) 2.4
ALT, alanine aminotransferase; Anti-HCV Ab, hepatitis C antibody; AST, aspartate aminotransferase; HBsAg, hepatitis B surface antigen;
HCC, hepatocellular carcinoma; IPTW, inverse probability treatment weighting; pRBC, packed red blood cell; PEI, percutaneous ethanol
injection; RFA, radiofrequency ablation; SDiff, standardised difference (imbalance > 20); TACE, transarterial chemoembolisation.

The effects of blood transfusion on outcomes after control for confounding variables may explain discrepant
hepatocellular cancer resection are contentious [9, 10, 21, 22]. findings. Our study design controlled for the effects of
Small sample sizes, incomplete data collection and failure to variables that are known to interact with blood transfusion

4 © 2019 Association of Anaesthetists

Tai et al. | Blood transfusion and recurrence of liver cancer Anaesthesia 2019

Figure 2 Dose–response association between the number of packed red blood cell units transfused and the hazard ratio (solid
line) with 95% confidence intervals (dotted line) for cancer recurrence, (a) unadjusted, (b) inverse probability of treatment-
weighted, (c) multivariable analysis; all-cause mortality, (d) unadjusted, (e) inverse probability of treatment-weighted,
(f) multivariable analysis. The restricted cubic spline function used three knots located at 1, 5 and 9 units.

© 2019 Association of Anaesthetists 5

Anaesthesia 2019 Tai et al. | Blood transfusion and recurrence of liver cancer

Table 2 Univariate analysis of cancer recurrence and all-cause mortality.

Cancer recurrence All-cause mortality
HR 95%CI p value HR 95%CI p value
Red cell transfusion 1.62 1.42–1.86 <0.001 2.28 1.83–2.82 <0.001
Age; y 1.00 1.00–1.01 0.223 1.01 1.00–1.02 0.023
Sex; male 1.08 0.92–1.27 0.365 1.08 0.83–1.40 0.559
ASA ≥ 3 1.07 0.92–1.24 0.365 1.25 0.99–1.57 0.059
HBsAg positive 1.10 0.95–1.28 0.187 0.97 0.77–1.22 0.813
Anti-HCV Ab positive 1.16 0.99–1.36 0.062 1.11 0.86–1.43 0.413
Alcoholism 0.88 0.68–1.15 0.359 0.84 0.54–1.32 0.453
Liver cirrhosis 1.38 1.21–1.58 <0.001 1.59 1.29–1.97 <0.001
Child–Pugh class; B or C 1.74 1.25–2.42 0.001 2.21 1.46–3.35 <0.001
Diabetes mellitus 1.11 0.95–1.29 0.199 1.44 1.14–1.83 0.002
Chronic kidney disease 0.95 0.75–1.20 0.650 1.64 1.21–2.22 0.001
1
Haemoglobin; g.dl 0.96 0.93–1.00 0.050 0.88 0.83–0.94 <0.001
3 1
Platelet count; 10 .ll 1.00 1.00–1.00 0.874 1.00 1.00–1.00 0.999
International normalised ratio 1.04 0.98–1.11 0.215 1.09 1.02–1.16 0.012
Total bilirubin ≥ 1.0 mg.dl 1
1.09 0.93–1.28 0.300 1.34 1.06–1.70 0.016
1
AST > 40 IU.l 1.71 1.50–1.96 <0.001 1.97 1.59–2.45 <0.001
1
ALT > 40 IU.l 1.24 1.09–1.42 0.001 1.21 0.98–1.50 0.080
Alpha-fetoprotein > 20 ng.ml 1a
1.62 1.42–1.86 <0.001 1.59 1.28–1.98 <0.001
Albumin ≤ 3.5 g.dl 1
1.48 1.16–1.88 0.001 1.76 1.22–2.54 0.002
Serum creatinine; lmol.l 1a
1.00 1.00–1.00 0.488 1.00 1.00–1.01 0.006
Tumour diameter > 5 cm 1.56 1.36–1.79 <0.001 2.08 1.68–2.57 <0.001
Multiple nodules 1.80 1.55–2.09 <0.001 1.57 1.24–1.99 <0.001
Poor or undifferentiated histology 1.43 1.24–1.64 <0.001 1.33 1.07–1.66 0.011
Microvascular invasion 1.71 1.47–1.99 <0.001 1.91 1.49–2.44 <0.001
Extracapsular invasion 1.29 1.13–1.48 <0.001 1.30 1.05–1.61 0.015
Positive surgical margin 2.13 1.65–2.76 <0.001 2.71 1.90–3.84 <0.001
Pre-operative TACE/RFA/PEI 1.23 0.98–1.54 0.081 1.55 1.11–2.16 0.009
Extent of hepatectomy > two segments resected 1.27 1.10–1.45 0.001 1.59 1.28–1.96 <0.001
Laparoscopic or robotic surgery 0.77 0.56–1.06 0.109 0.55 0.28–1.07 0.078
Blood loss; mla 1.23 1.17–1.29 <0.001 1.38 1.27–1.50 <0.001
Epidural analgesia 1.04 0.91–1.20 0.537 1.02 0.82–1.26 0.892
a
Anaesthesia time; min 1.59 1.36–1.85 <0.001 2.14 1.67–2.72 <0.001
Date of surgery; 2011–16 vs. 2005–10 0.82 0.72–0.94 0.005 0.70 0.56–0.89 0.003
ALT, alanine aminotransferase; Anti-HCV Ab, hepatitis C antibody; AST, aspartate aminotransferase; HBsAg, hepatitis B surface antigen;
HR, hazard ratio; PEI, percutaneous ethanol injection; RFA, radiofrequency ablation; TACE, transarterial chemoembolisation.
a
On base-2 logarithmic scale.

and our sample size was relatively large. We confined
operations to those performed after 2005 to reduce the
effects of known and unknown time-varying covariates. We
think that time-varying effects might have been largely
controlled by the inverse probability of treatment weighting
and multivariable analyses, both of which incorporated time
as a variable. Propensity scoring has been used in previous
studies, but inverse probability of treatment weighting has
the benefit of preserving the original sample size, increasing
the power of propensity score analysis [21–23]. The
association of allogeneic blood transfusion with cancer
recurrence and death after hepatocellular resection is
consistent with meta-analysis [11].
If transfusion causes tumour recurrence the effect might
be mediated by modulation of the immune system.
Potential causes of transfusion-induced immune
dysfunction include reduced cytotoxic cell activity and
inhibition of interleukin-2 production, along with increased

6 © 2019 Association of Anaesthetists

Tai et al. | Blood transfusion and recurrence of liver cancer Anaesthesia 2019

Table 3 Multivariable analysis of cancer recurrence and all-cause mortality.

Cancer recurrence HR 95%CI p value All-cause mortality HR 95%CI p value
Red cell transfusion 1.29 1.07–1.57 0.009 pRBC transfusion 1.42 1.04–1.96 0.030
Liver cirrhosis 1.23 1.05–1.43 0.008 Liver cirrhosis 1.55 1.20–2.02 0.001
1
AST > 40 IU.l 1.43 1.22–1.66 <0.001 Child-Pugh grade; B or C vs. A 2.08 1.31–3.30 0.002
Alpha-fetoprotein > 20 ng.ml 1
1.49 1.27–1.75 <0.001 Diabetes mellitus 1.69 1.29–2.22 <0.001
Multiple nodules 1.45 1.22–1.73 <0.001 Chronic kidney disease 1.47 1.02–2.11 0.039
1
Poor or undifferentiated histology 1.24 1.05–1.45 0.009 Total bilirubin ≥ 1.0 mg.dl 1.46 1.10–1.94 0.008
1
Microvascular invasion 1.29 1.07–1.55 0.007 AST > 40 IU.l 1.55 1.14–2.11 0.005
Extracapsular invasion 1.30 1.11–1.51 0.001 ALT > 40 IU.l 1
0.73 0.54–0.98 0.037
Positive surgical margin 1.50 1.11–2.03 0.008 Alpha-fetoprotein > 20 ng.ml 1
1.52 1.17–1.97 0.002
Blood lossa 1.09 1.02–1.16 0.015 Tumour diameter > 5 cm 1.49 1.13–1.97 0.005
Microvascular invasion 1.66 1.20–2.30 0.002
Positive surgical margin 1.93 1.25–2.96 0.003
Blood lossa 1.13 1.01–1.27 0.029
Date of surgery; 2011–16 vs. 2005–10 0.66 0.50–0.87 0.003

ALT, alanine aminotransferase; AST, aspartate aminotransferase.

a
On base-2 logarithmic scale.

suppressor T-cell activity and immunosuppressive the effect of other blood products (e.g. fresh frozen plasma
prostaglandins [24]. The cytotoxic mediators, natural killer and platelet concentrates) which might alter immune
cell activity and lymphocyte count are decreased after responses and affect oncological outcomes after surgery [24].
transfusions of allogeneic red blood cells [24–26]. These In conclusion, allogeneic blood transfusion was a
findings have been associated with increased tumour significant and independent variable associated with
proliferation, apoptosis and metastasis spread [27]. Studies increased cancer recurrence and all-cause mortality in
in patients with colorectal cancer show a similar disruption patients undergoing liver resection with intent to cure
of the immune response associated with blood transfusion hepatocellular carcinoma. Well-designed prospective studies
that is thought to control cancer spread [28, 29]. are needed to test whether the association is the causal.
There are limitations to our study. Our analysis was
limited to a list of variables with evidence supporting an effect Acknowledgements
on cancer recurrence and spread. There may be other We are indebted to Dr. Chen-Yang Hsu and Professor Amy
influential variables we did not measure. We did not evaluate Ming-Fang Yen (School of Oral Hygiene, College of Oral
Medicine, Taipei Medical University, Taipei, Taiwan) for
statistical consultation and providing SAS Macro for
Table 4 Independent predictors of peri-operative blood
restricted cubic spline analyses. This work was supported by
transfusion.
grants from Taipei Veterans General Hospital (V104C-096),
OR 95%CI p value
Yen Tjing Ling Medical Foundation, Taipei, Taiwan
Age; y 1.02 1.00–1.03 0.012 (CI-108-27), Anesthesiology Research and Development
Sex; male 0.59 0.41–0.85 0.004 Foundation, Taipei, Taiwan (ARDF10701) and Ministry of
ASA ≥ 3 1.39 1.01–1.92 0.045 Science and Technology, Taipei, Taiwan, R.O.C. (MOST104-
Child-Pugh class; B or C vs. A 2.15 1.08–4.31 0.030 2314-B-075-015). No competing interests declared.
Haemoglobin level; g.dl 1
0.63 0.57–0.69 <0.001
Total bilirubin ≥ 1.0 mg.dl 1
1.59 1.12–2.27 0.010 References
ALT > 40 IU.l 1
0.73 0.54–0.98 0.037 1. World Health Organization International Agency for Research
on Cancer. GLOBOCAN: estimated cancer incidence, mortality
Blood loss; mla 6.48 5.33–7.89 <0.001
and prevalence worldwide in 2018. http://gco.iarc.fr/today/fac
Epidural analgesia 0.71 0.52–0.96 0.026 t-sheets-cancers (accessed 22/04/2019).
Anaesthesia time; mina 2.41 1.64–3.54 <0.001 2. Chan AWH, Zhong J, Berhane S, et al. Development of pre and
post-operative models to predict early recurrence of
ALT, alanine aminotransferase. hepatocellular carcinoma after surgical resection. Journal of
a
On base-2 logarithmic scale. Hepatology 2018; 69: 1284–93.

© 2019 Association of Anaesthetists 7

Anaesthesia 2019
3. Tabrizian P, Jibara G, Shrager B, Schwartz M, Roayaie S. Recurrence
of hepatocellular cancer after resection: patterns, treatments, and
prognosis. Annals of Surgery 2015; 261: 947–55.
4. Xu XF, Xing H, Han J, et al. Risk factors, patterns, and outcomes
of late recurrence after liver resection for hepatocellular
carcinoma: a multicenter study from China. Journal of the
American Medical Association Surgery 2018; 154: 209–17.
5. Zou H, Zhu CZ, Wang C, et al. Recurrence of Barcelona clinic
liver cancer stage A hepatocellular carcinoma after
hepatectomy. American Journal of the Medical Sciences 2017;
354: 262–7.
6. Cescon M, Vetrone G, Grazi GL, et al. Trends in perioperative
outcome after hepatic resection: analysis of 1500 consecutive
unselected cases over 20 years. Annals of Surgery 2009; 249:
995–1002.
7. Ferraris VA, Davenport DL, Saha SP, Austin PC, Zwischenberger
JB. Surgical outcomes and transfusion of minimal amounts of
blood in the operating room. Archives of Surgery 2012; 147:
49–55.
8. Odell DD, Bilimoria KY. Evaluating appropriate blood
transfusion in cancer surgery. Journal of the American Medical
Association Surgery 2016; 151: 525–6.
9. Matsumata T, Ikeda Y, Hayashi H, Kamakura T, Taketomi A,
Sugimachi K. The association between transfusion and cancer-
free survival after curative resection for hepatocellular
carcinoma. Cancer 1993; 72: 1866–71.
10. Makino Y, Yamanoi A, Kimoto T, El-Assal ON, Kohno H,
Nagasue N. The influence of perioperative blood transfusion
on intrahepatic recurrence after curative resection of
hepatocellular carcinoma. American Journal of Gastro-
enterology 2000; 95: 1294–300.
11. Liu L, Wang Z, Jiang S, et al. Perioperative allogenenic blood
transfusion is associated with worse clinical outcomes for
hepatocellular carcinoma: a meta-analysis. PLoS ONE 2013; 8:
e64261.
12. Poon RT, Fan ST, Lo CM, Liu CL, Ng IO, Wong J. Long-term
prognosis after resection of hepatocellular carcinoma
associated with hepatitis B-related cirrhosis. Journal of Clinical
Oncology 2000; 18: 1094–101.
13. Tangkijvanich P, Anukulkarnkusol N, Suwangool P, et al.
Clinical characteristics and prognosis of hepatocellular
carcinoma: analysis based on serum alpha-fetoprotein levels.
Journal of Clinical Gastroenterology 2000; 31: 302–8.
14. Abou-Alfa GK, Pawlik SJ, Vauthey JN. Liver. In: Amin MB, ed.
AJCC Cancer Staging Manual. Chicago, IL: AJCC, 2017: 191–
200.
15. Iguchi T, Aishima S, Taketomi A, et al. Extracapsular penetration
is a new prognostic factor in human hepatocellular carcinoma.
American Journal of Surgical Pathology 2008; 32: 1675–82.
16. Desquilbet L, Mariotti F. Dose-response analyses using
restricted cubic spline functions in public health research.
Statistics in Medicine 2010; 29: 1037–57.
17. Thoemmes F, Ong AD. A primer on inverse probability of
treatment weighting and marginal structural models. Emerging
Adulthood 2016; 4: 40–59.
18. Schoenfeld DA. Sample-size formula for the proportional-
hazards regression model. Biometrics 1983; 39: 499–503.
19. Austin PC, Stuart EA. Moving towards best practice when using
inverse probability of treatment weighting (IPTW) using the
propensity score to estimate causal treatment effects in
8 © 2019 Association of Anaesthetists
Tai et al. | Blood transfusion and recurrence of liver cancer
observational studies. Statistics in Medicine 2015; 34: 3661–
79.
20. Hill AB. The environment and disease: association or
causation? Proceedings of the Royal Society of Medicine 1965;
58: 295–300.
21. Yang T, Lu JH, Lau WY, et al. Perioperative blood transfusion
does not influence recurrence-free and overall survivals after
curative resection for hepatocellular carcinoma: a propensity
score matching analysis. Journal of Hepatology 2016; 64: 583–
93.
22. Kuroda S, Tashiro H, Kobayashi T, Oshita A, Amano H, Ohdan
H. No impact of perioperative blood transfusion on recurrence
of hepatocellular carcinoma after hepatectomy. World Journal
of Surgery 2012; 36: 651–8.
23. Pirracchio R, Resche-Rigon M, Chevret S. Evaluation of the
propensity score methods for estimating marginal odds ratios
in case of small sample size. BMC Medical Research
Methodology 2012; 12: 70.
24. Cata JP, Wang H, Gottumukkala V, Reuben J, Sessler DI.
Inflammatory response, immunosuppression, and cancer
recurrence after perioperative blood transfusions. British
Journal of Anaesthesia 2013; 110: 690–701.
25. Kwon AH, Matsui Y, Kamiyama Y. Perioperative blood
transfusion in hepatocellular carcinomas. Cancer 2001; 91:
771–8.
26. Sugita S, Sasaki A, Iwaki K, et al. Prognosis and postoperative
lymphocyte count in patients with hepatocellular carcinoma
who received intraoperative allogenic blood transfusion: a
retrospective study. European Journal of Surgical Oncology
2008; 34: 339–45.
27. Ydy LR, Slhessarenko N, de Aguilar-Nascimento JE. Effect of
perioperative allogeneic red blood cell transfusion on the
immune-inflammatory response after colorectal cancer
resection. World Journal of Surgery 2007; 31: 2044–51.
28. Acheson AG, Brookes MJ, Spahn DR. Effects of allogeneic red
blood cell transfusions on clinical outcomes in patients
undergoing colorectal cancer surgery: a systematic review and
meta-analysis. Annals of Surgery 2012; 256: 235–44.
29. Wu HL, Tai YH, Lin SP, Chan MY, Chen HH, Chang KY.
The impact of blood transfusion on recurrence and
mortality following colorectal cancer resection: a
propensity score analysis of 4,030 patients. Scientific
Reports 2018; 8: 13345.
Supporting Information
Additional supporting information may be found online in
the Supporting Information section at the end of the article.
Table S1. The frequency table describing the number
of units transfused in the study cohort.
Table S2. The result of logistic regression analysis for
inverse probability of treatment weighting.
Table S3. Dose–response effects of blood transfusion
on cancer recurrence and all-cause mortality.
Figure S1. Standardised differences in observed
covariates in the original and weighted samples.
Anaesthesia 2020, 75, 472–478 doi:10.1111/anae.14918

Original Article

The association of pre-operative anaemia with survival after

orthotopic liver transplantation
P. Lichtenegger,1 J. Schiefer,2 A. Graf,3 G. Berlakovich,4 P. Faybik,5 D. M. Baron5 and
J. Baron-Stefaniak2

1 Specialty Registrar, 2 Consultant, 5 Associate Professor, Department of Anaesthesia, Intensive Care Medicine and Pain
Medicine, 3 Associate Professor, Section for Medical Statistics, Centre for Medical Statistics, Informatics, and Intelligent
Systems, 4 Head, Division of Transplantation, Department of Surgery, Medical University of Vienna, Vienna, Austria

Summary
Anaemia is common in patients with end-stage liver disease. Pre-operative anaemia is associated with greater
mortality after major surgery. We analysed the association of pre-operative anaemia (World Health
Organization classification) with survival and complications after orthotopic liver transplantation using Cox and
logistic regression models. We included patients undergoing their first orthotopic liver transplantation between
2004 and 2016. Out of 599 included patients, 455 (76%) were anaemic before transplantation. Pre-operative
anaemia was not associated with the survival of 485/599 (81%) patients to 1 year after liver transplantation, OR
(95%CI) 1.04 (0.64–1.68), p = 0.88. Pre-operative anaemia was associated with higher rates of intra-operative
blood transfusions and acute postoperative kidney injury on multivariable analysis, OR (95%CI) 1.70 (0.82–2.59)
and 1.72 (1.11–2.67), respectively, p < 0.001 for both. Postoperative renal replacement therapy was associated
with pre-operative anaemia on univariate analysis, OR (95%CI) 1.87 (1.11–3.15), p = 0.018.

.................................................................................................................................................................
Correspondence to: D. M. Baron
Email: david.baron@meduniwien.ac.at
Accepted: 15 October 2019
Keywords: end-stage liver disease; morbidity; mortality; orthotopic liver transplantation; outcome; pre-operative
anaemia; transfusion

Introduction after cardiac, orthopaedic and abdominal surgery [11–16].

Orthotopic liver transplantation can cure patients suffering
from acute or chronic liver failure, the outcome of which is
associated with several patient variables, such as age,
diabetes mellitus, postoperative acute kidney injury and
peritonitis [1–5]. Survival to one postoperative year is about
80% [6].
Pre-operative anaemia is present in about one-third of
patients who undergo major surgery and it is associated
with subsequent morbidity and mortality [7, 8]. The
incidence of anaemia before orthotopic
transplantation is as high as 75%, but it is uncertain whether
it is associated with outcome [6, 9, 10].
Packed red blood cells are often transfused to correct
anaemia, yet transfusion is associated with adverse
outcomes and may cause increased morbidity and mortality
Transfusion is also associated with adverse outcomes after
liver transplantation [2, 17].
Our primary aim was to evaluate the association of
anaemia before orthotopic liver transplantation with
survival to 1 year and secondarily survival to the end of
follow-up. We were also interested in the associations of
peri-operative variables, including blood transfusion,
with these outcomes and their interactions with pre-
operative anaemia.
liver
Methods
The ethics committee of the Medical University of Vienna
approved this study. We studied patients admitted to our
hospital for orthotopic liver transplantation between
January 2004 and December 2016. We did not include

472 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use,
distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.
Lichtenegger et al. | Pre-operative anaemia and outcome after liver transplantation Anaesthesia 2020, 75, 472–478

children (< 18 years), patients having concurrent lung or We determined the association of WHO-defined pre-
kidney transplantation, and patients with missing operative anaemia (< 13 g.dl 1
for men and < 12 g.dl 1
for
haemoglobin values within 24 pre-operative hours. women) with 1-year survival. We performed secondary
Orthotopic liver transplantation was performed under analyses of the association of the severity of anaemia in men
1
general anaesthesia, with caval replacement without the and women with survival, categorised as mild (11–13 g.dl
1 1
routine of veno-venous bypass. Blood group and and 10–12 g.dl , respectively), moderate (8–11 g.dl and
1 1
physiological body-to-weight ratio were matched between 8–10 g.dl , respectively) or severe (< 8 g.dl ) [18]. The
donor grafts and transplant recipients. All patients were secondary outcome was survival to the end of 2017.
admitted to an intensive care unit (ICU) postoperatively. We assessed the associations of pre-operative
Immunosuppressive therapy was started with intravenous variables other than anaemia with outcomes such as age;
dexamethasone 40 mg before graft reperfusion. For the first sex; body mass index; model for end-stage liver disease
five postoperative days patients received a reduced (MELD) score; coronary artery disease; chronic
dexamethasone dose every 24 h, followed by a daily dose of obstructive pulmonary disease; diabetes mellitus; and
4 mg for 3 months (and longer for patients with autoimmune pre-operative hospitalisation. We defined coronary artery
1
disease). Anti-thymocyte globulin 2.5 mg.kg was started disease as the presence of coronary atherosclerosis
on arrival in the ICU and continued for 3 days, after which (> 50% stenosis of a coronary artery), or previous
immunosuppressive therapy was continued with low-dose percutaneous coronary intervention. We defined pre-
tacrolimus or cyclosporin A at target concentrations of 6– operative hospitalisation as inpatient treatment for at
1 1
8 ng.ml and 130–150 ng.ml , respectively. least 3 days before liver transplantation, or repeated

Table 1 Characteristics of 599 patients who had orthotopic liver transplantation, categorised by pre-operative anaemia. Values
are mean (SD), median (IQR [range]) or number (proportion).
Anaemic
All patients No Yes
Characteristic (n = 599) (n = 144) (n = 455) p value
Pre-operative
Age; years 53 (10) 55 (8) 53 (11) 0.001
Sex; male 434 (72%) 101 (70%) 333 (73%) 0.54
2
Body mass index; kg m 26.1 (4.4) 26.4 (4.2) 26 (4.5) 0.42
MELD score 16 (7) 13 (6) 17 (7) < 0.001
Coronary artery disease 20 (3%) 6 (4%) 14 (3%) 0.52
COPD 38 (6%) 10 (7%) 28 (6%) 0.89
Diabetes mellitus 135 (23%) 42 (29%) 93 (20%) 0.038
Hospitalisation 60 (10%) 6 (4%) 54 (12%) 0.011
Intra-operative
Cold ischaemia time; min 471 (139) 482 (147) 468 (136) 0.32
Warm ischaemia time; min 79 (19) 77 (18) 79 (20) 0.24
Transfusion
Packed red blood cells 2 (0–5 [0–40]) 0 (0–2 [0–19]) 3 (1–6 [0–40]) < 0.001
Fresh frozen plasma 7 (4–12 [0–62]) 5 (0–9 [0–30]) 8 (4–12 [0–62]) < 0.001
Platelets 0 (0–1 [0–9]) 0 (0–0 [0–3]) 0 (0–1 [0–9]) 0.020
Postoperative
Surgical complications 113 (19%) 21 (15%) 92 (20%) 0.18
Early allograft dysfunction 138 (23%) 27 (19%) 111 (24%) 0.20
Acute kidney injury 412 (69%) 82 (57%) 330 (73%) < 0.001
Renal replacement therapy 126 (21%) 20 (14%) 106 (23%) 0.023
Length of ICU stay; days 6 (4–11 [0–98]) 5 (4–9 [1–90]) 7 (4–12 [0–98]) < 0.001
Length of hospital stay; days 19 (14–28 [6–196]) 15 (13–25 [6–118]) 20 (14–30 [9–196]) 0.002
COPD, chronic obstructive pulmonary disease; ICU, intensive care unit; MELD, model for end-stage liver disease.

© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists 473
Anaesthesia 2020, 75, 472–478 Lichtenegger et al. | Pre-operative anaemia and outcome after liver transplantation

hospital admissions within 1 month before liver
transplantation. We also analysed intra-operative and
postoperative variables: cold ischaemia time; warm
ischaemia time; and transfused blood products – packed
red blood cells, fresh frozen plasma and platelets;
surgical revision; early allograft dysfunction; acute kidney
injury; and renal replacement therapy. We defined
surgical revision as surgery for bleeding, vascular or
biliary duct stenosis, biliary duct leak or intra-abdominal
haematoma and peritonitis within one postoperative
month. We defined early allograft dysfunction as at least
1
one of serum bilirubin concentration ≥ 10 mg.dl or
international normalised ratio ≥ 1.6 7 days after surgery;
or serum aminotransferase concentration > 2000 IU.ml 1
within one postoperative week [19]. We defined acute
kidney injury with standard criteria [20]. We categorised
patients by whether they had renal replacement therapy
within one postoperative week.
We did not calculate sample size for this retrospective
study. We used two-sample t-test or Wilcoxon rank-sum test
for continuous variables and Chi-square test or Fisher’s
exact test for categorical variables to compare patients with
and without anaemia. We analysed 1-year survival using
univariable and multivariable logistic regression and overall
survival using univariable and multivariable Cox
proportional hazard regression models. We used
Schoenfeld residuals to test for violation of proportional
hazards. We used accelerated failure time models for
sensitivity analysis. We used cubic splines to test for non-
linear associations of continuous variables with outcome.
We tested for interactions between variables. We used a
stepwise selection and retention method for multivariable
model selection, adding and removing variables at
p < 0.05, so that the final model only included variables with
p < 0.05. We used univariable and multivariable logistic
regression models for the association of anaemia with
postoperative acute kidney injury, renal replacement
therapy, early allograft dysfunction and surgical
complications and linear regression models for the
association of anaemia and packed red blood cell
transfusion. We used R, release 3.3.3 [21] and SAS 9.4 for
the statistical analyses. We considered p < 0.05 to be
statistically significant.
Results
We studied 599/659 patients, most of whom were
anaemic before orthotopic liver transplantation (Table 1
and Table S1). The mean (SD) pre-operative
1
haemoglobin concentration was 13.8 (1.1) g.dl for 144
1
patients who were not anaemic and 10.2 (1.5) g.dl for
455 anaemic patients. Anaemic patients were younger
but with worse liver disease, for which they were more

Figure 1 Survival after orthotopic liver transplantation of 599 patients, categorised by pre-operative haemoglobin
concentration as not anaemic in 144 patients (green line); or anaemic in 455 patients (red line). Corresponding 95%CIs are
shown as green or red areas, respectively. There were no statistically significant differences during the first year or for overall
survival, p = 0.88 and p = 0.20, respectively.

474 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Lichtenegger et al. | Pre-operative anaemia and outcome after liver transplantation Anaesthesia 2020, 75, 472–478

often hospitalised, and they were more likely to have
diabetes (Table 1).
There were 114/599 (19%) deaths in the first
postoperative year from infection in 36 (32%); bleeding in
15 (13%); graft failure in 11 (10%); multi-organ failure in 9
(8%); cardiovascular events in 6 (5%); cancer in 5 (4%);
surgical complications in 5 (4%); disease recurrence in 3
(3%); respiratory failure in 3 (3%); and unknown in 19 (17%).
The median (IQR [range]) follow-up was 4 (1–9 [0–15]) years,
and the median survival (95%CI) was 11 (8–13) years. There
were 235 deaths during follow-up from infection in 51
(22%); bleeding in 20 (9%); graft failure in 18 (8%); multi-
organ failure in 15 (6%); cardiovascular events in 12 (5%);
cancer in 11 (5%); surgical complications in 5 (2%); disease
recurrence in 13 (6%); respiratory failure in 5 (2%); and
unknown in 84 (36%).
Survival to one postoperative year and to the end of
follow-up was not associated with anaemia (Fig. 1), but
was associated with a number of variables (Tables 2 and
3). Pre-operative anaemia was associated with increased
peri-operative events, OR (95%CI): 1.70 (0.82–2.59) for
intra-operative red blood cell transfusion, p < 0.001; and
OR (95%CI) 1.72 (1.11–2.67) for acute post-operative
kidney injury, p < 0.001. Pre-operative anaemia showed
association with renal replacement therapy; OR (95%CI)
1.87 (1.11–3.15), p = 0.018 in univariate analysis;
however, when correcting for confounding factors, for
instance the MELD score, this association did not remain
significant. Anaemia was not associated with early
allograft dysfunction and surgical complications, OR (95%
CI) 1.40 (0.87–2.24) and 1.47 (0.88–2.47), p = 0.16 and
p = 0.14, respectively.

Table 2 The association of variables with survival to 1 year after orthotopic liver transplantation for 599 patients.
Univariable model Multivariable model
Variables OR 95%CI p value OR 95%CI p value
Pre-operative
Anaemia grade
Any 1.04 0.64–1.68 0.88
Mild 1.09 0.62–1.94 0.76
Moderate 1.00 0.59–1.70 0.99
Severe 1.00 0.39–2.54 > 0.99
Haemoglobin 0.99 0.90–1.10 0.87
Age; years 0.98 0.96–1.00 0.087
Sex; male 0.97 0.61–1.54 0.90
1 2
Body mass index; kg .m 0.99 0.94–1.04 0.61
MELD score 0.95 0.93–0.98 0.001
Coronary artery disease 1.50 0.43–5.24 0.52
COPD 1.62 0.61–4.26 0.33
Diabetes mellitus 0.89 0.54–1.45 0.63
Hospitalisation 0.52 0.28–0.98 0.043
Intra-operative
1
Cold ischaemia time; min 1.00 1.00–1.00 0.43
1
Warm ischaemia time; min 1.00 0.98–1.01 0.38
Transfusion
Packed red blood cells 0.91 0.87–0.95 < 0.001 0.95 0.91–0.99 0.027
Fresh frozen plasma 0.96 0.94–0.99 0.002
Platelets 0.72 0.58–0.88 0.001
Postoperative
Surgical complications 0.26 0.16–0.41 < 0.001 0.42 0.25–0.70 0.001
Early allograft dysfunction 0.48 0.30–0.75 0.001
Acute kidney injury 0.63 0.39–1.01 0.055
Renal replacement therapy 0.22 0.14–0.34 < 0.001 0.35 0.21–0.58 < 0.001
Year liver transplantation 0.96 0.91–1.02 0.16
COPD, chronic obstructive pulmonary disease; MELD, model for end-stage liver disease; OR, odds ratio; CI, confidence interval.

© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists 475
Anaesthesia 2020, 75, 472–478 Lichtenegger et al. | Pre-operative anaemia and outcome after liver transplantation

Table 3 The association of variables with survival to the end of follow-up after orthotopic liver transplantation for 599 patients.
Univariable model Multivariable model
Variables HR 95%CI p value HR 95%CI p value
Pre-operative
Anaemia grade
Any 1.21 0.91–1.62 0.20
Mild 1.15 0.81–1.62 0.43
Moderate 1.30 0.94–1.80 0.11
Severe 0.96 0.54–1.68 0.88
Haemoglobin 0.96 0.90–1.02 0.22
Age; years 0.97 0.96–0.99 < 0.001 0.98 0.96–0.99 0.001
Sex; male 0.78 0.58–1.05 0.10
1
Body mass index; kg .m2 0.99 0.96–1.02 0.56
MELD score 0.98 0.97–1.00 0.071
Coronary artery disease 0.79 0.43–1.45 0.52
COPD 0.81 0.49–1.32 0.40
Diabetes mellitus 0.70 0.53–0.94 0.017 0.71 0.53–0.96 0.025
Hospitalisation 0.91 0.59–1.42 0.69
Intra-operative
1
Cold ischaemia time; min 1.00 1.00–1.00 0.17
1
Warm ischaemia time; min 1.00 0.99–1.01 0.48
Transfusion
Packed red blood cells 0.95 0.92–0.97 < 0.001
Fresh frozen plasma 0.98 0.96–0.99 0.003
Platelets 0.84 0.76–0.93 0.001
Postoperative
Surgical complications 0.49 0.37–0.66 < 0.001 0.62 0.45–0.86 0.004
Early allograft dysfunction 0.72 0.54–0.96 0.025
Acute kidney injury 0.84 0.64–1.12 0.23
Renal replacement therapy 0.44 0.33–0.58 < 0.001 0.50 0.37–0.68 < 0.001
Year liver transplantation 0.98 0.94–1.01 0.23
COPD, chronic obstructive pulmonary disease; MELD, model for end-stage liver disease; HR, hazard ratio; CI, Confidence interval.

Discussion of the MELD score with survival to one postoperative year in

We found that pre-operative anaemia was not associated our cohort is consistent with its association with coagulation.
with survival after liver transplantation. Pre-operative Pre-operative anaemia was associated with acute post-
anaemia was associated with peri-operative blood operative kidney injury in our cohort. We assume that pre-
transfusion and postoperative acute kidney injury. operative anaemia was also associated with renal
Pre-operative anaemia is associated with worse replacement therapy, with which anaemia was associated
outcome after major cardiac and non-cardiac surgery [7, 14, in isolation from other variables. Other studies have
22]. Some previous studies have reported an reported contradictory evidence for association of
association of pre-operative anaemia with survival after liver anaemia with renal replacement therapy after liver
transplantation [10], while others have not [6]. Pre-operative transplantation, which in turn is associated with worse
anaemia may be indirectly associated with survival after liver survival [2, 6, 24].
transplantation, as survival is worse after blood transfusion Acute and chronic gastro-intestinal haemorrhage,
[16, 17, 23]. The decision to transfuse blood is affected by malnutrition, iron deficiency, folate or vitamin B12
variables such as the MELD score or cold ischaemia time, deficiency and suppression of bone marrow are mainly
which are associated with impaired coagulation, as well as responsible for anaemia in end-stage liver disease patients
haemoglobin concentration. Thus, the univariate association [9, 25–28]. Increasing pre-operative haemoglobin

476 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Lichtenegger et al. | Pre-operative anaemia and outcome after liver transplantation
concentrations reduces peri-operative blood transfusion
after gynaecological, obstetric, oncologic, orthopaedic and
abdominal surgery [29–32]. We think that patients awaiting
liver transplantation might benefit from pre-operative iron
supplementation, which needs to be investigated in
prospective randomised studies.
Unfortunately, due to insufficient diagnostic data, the
cause of anaemia remains unknown in most of our patients.
Iron status was only available for approximately half the
study population. Further limitations of our study are the
retrospective and monocentric design, which might limit
the generalisability of our results. We analysed survival after
operations that spanned 12 years, during which medical
interventions have changed – such as the use of
hydroxyethyl starch – as has general population survival, but
date of surgery was not associated with observed survival in
our cohort [33, 34].
In conclusion, pre-operative anaemia was not
associated with survival after liver transplantation. However,
anaemia was associated with blood transfusion,
postoperative acute kidney injury and renal replacement
therapy.
Acknowledgements
DB has received honoraria and funding for an investigator-
initiated trial from Vifor Pharma. No other external funding
or competing interests declared.
References
1. Hoehn RS, Singhal A, Wima K, et al. Effect of pretransplant
diabetes on short-term outcomes after liver transplantation: a
national cohort study. Liver International 2015; 35: 1902–9.
2. Hilmi IA, Damian D, Al-Khafaji A, et al. Acute kidney injury
following orthotopic liver transplantation: incidence, risk
factors, and effects on patient and graft outcomes. British
Journal of Anaesthesia 2015; 114: 919–26.
3. Barri YM, Sanchez EQ, Jennings LW, et al. Acute kidney injury
following liver transplantation: definition and outcome. Liver
Transplantation 2009; 15: 475–83.
4. Zand MS, Orloff MS, Abt P, et al. High mortality in orthotopic
liver transplant recipients who require hemodialysis. Clinical
Transplantation 2011; 25: 213–21.
5. Pungpapong S, Alvarez S, Hellinger WC, et al. Peritonitis after
liver transplantation: incidence, risk factors, microbiology
profiles, and outcome. Liver Transplantation 2006; 12: 1244–
52.
6. Collas O, Robertson FP, Fuller BJ, Davidson BR. Anaemia in
patients with chronic liver disease and its association with
morbidity and mortality following liver transplantation.
International Journal of Surgery 2018; 53: 48–52.
7. Baron DM, Hochrieser H, Posch M, et al. Preoperative anaemia
is associated with poor clinical outcome in non-cardiac surgery
patients. British Journal of Anaesthesia 2014; 113: 416–23.
8. Musallam KM, Tamim HM, Richards T, et al. Preoperative
anaemia and postoperative outcomes in non-cardiac surgery: a
retrospective cohort study. Lancet 2011; 378: 1396–407.
© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Anaesthesia 2020, 75, 472–478
9. Gonzalez-Casas R, Jones EA, Moreno-Otero R. Spectrum of
anemia associated with chronic liver disease. World Journal of
Gastroenterology 2009; 15: 4653–8.
10. Badawy A, Kaido T, Hammad A, et al. The impact of
preoperative hemoglobin level on the short-term outcomes
after living donor liver transplantation. World Journal of
Surgery 2018; 42:4081–9.
11. Gombotz H, Hofmann A. Patient blood management: three
pillar strategy to improve outcome through avoidance of
allogeneic blood products. Der Anaesthesist 2013; 62: 519–
27.
12. Mu~ noz M, Gomez-Ramırez S, Kozek-Langeneker S, et al. ‘Fit to
fly’: overcoming barriers to preoperative haemoglobin
optimization in surgical patients. British Journal of Anaesthesia
2015; 115: 15–24.
13. Haase M, Bellomo R, Story D, et al. Effect of mean
arterial pressure, haemoglobin and blood transfusion
during cardiopulmonary bypass on post-operative acute
kidney injury. Nephrology Dialysis Transplantation 2012;
27: 153–60.
14. LaPar DJ, Hawkins RB, McMurry TL, et al. Preoperative anemia
versus blood transfusion: which is the culprit for worse
outcomes in cardiac surgery? Journal of Thoracic and
Cardiovascular Surgery 2018; 156: 66–74.
15. Bernard AC, Davenport DL, Chang PK, Vaughan TB,
Zwischenberger JB. Intraoperative transfusion of 1 U to 2 U
packed red blood cells is associated with increased 30-day
mortality, surgical-site infection, pneumonia, and sepsis in
general surgery patients. Journal of the American College of
Surgeons 2009; 208: 931–7.
16. Glance LG, Dick AW, Mukamel DB, et al. Association between
intraoperative blood transfusion and mortality and morbidity in
patients undergoing noncardiac surgery. Anesthesiology
2011; 114: 283–92.
17. de Boer MT, Christensen MC, Asmussen M, et al. The impact of
intraoperative transfusion of platelets and red blood cells on
survival after liver transplantation. Anesthesia and Analgesia
2008; 106: 32–44.
18. World Health Organization. Haemoglobin concentrations for
the diagnosis of anaemia and assessment of severity. 2011.
http://www.who.int/vmnis/indicators/haemoglobin.pdf. http://
www.who.int/vmnis/indicators/haemoglobin.pdf (accessed
17/10/2018).
19. Olthoff KM, Kulik L, Samstein B, et al. Validation of a current
definition of early allograft dysfunction in liver transplant
recipients and analysis of risk factors. Liver Transplantation
2010; 16: 943–9.
20. Fliser D, Laville M, Covic A, et al. A European Renal Best
Practice (ERBP) position statement on the Kidney Disease
Improving Global Outcomes (KDIGO) Clinical Practice
Guidelines on Acute Kidney Injury: part 1: definitions,
conservative management and contrast-induced nephropathy.
Nephrology Dialysis Transplantation 2012; 27: 4263–72.
21. Core R, Team R. A language and environment for statistical
computing. Vienna, Austria: R Foundation For Statistical
Computing, 2019.
22. Hung M, Besser M, Sharples LD, Nair SK, Klein AA. The
prevalence and association with transfusion, intensive care unit
stay and mortality of pre-operative anaemia in a cohort of
cardiac surgery patients. Anaesthesia 2011; 66: 812–18.
23. Carson JL, Terrin ML, Noveck H, et al. Liberal or restrictive
transfusion in high-risk patients after hip surgery. New England
Journal of Medicine 2011; 365: 2453–62.
24. Biteker M, Dayan A, Tekkesßin A_I, et al. Incidence, risk factors,
and outcomes of perioperative acute kidney injury in
noncardiac and nonvascular surgery. American Journal of
Surgery 2014; 207: 53–9.
477
Anaesthesia 2020, 75, 472–478 Lichtenegger et al. | Pre-operative anaemia and outcome after liver transplantation
25. Stein J, Connor S, Virgin G, Ong DE, Pereyra L. Anemia and iron
deficiency in gastrointestinal and liver conditions. World
Journal of Gastroenterology 2016; 22: 7908–25.
26. Mahamid M, Mahroum N, Bragazzi N, et al. Folate and B12
levels correlate with histological severity in NASH patients.
Nutrients 2018; 10: E440.
27. Villanueva C, Aracil C, Colomo A, et al. Acute hemodynamic
response to b-blockers and prediction of long-term outcome in
primary prophylaxis of variceal bleeding. Gastroenterology
2009; 137: 119–28.
28. Bihari C, Anand L, Rooge S, et al. Bone marrow stem cells and
their niche components are adversely affected in advanced
cirrhosis of the liver. Hepatology 2016; 64: 1273–88.
29. Ellermann I, Bueckmann A, Eveslage M, et al. Treating anemia in
the preanesthesia assessment clinic: results of a retrospective
evaluation. Anesthesia and Analgesia 2018; 127: 1202–10.
30. Keding V, Zacharowski K, Bechstein WO, Meybohm P,
Schnitzbauer AA. Patient blood management improves
outcome in oncologic surgery. World Journal of Surgical
Oncology 2018; 16: 159.
31. Gupta PB, DeMario VM, Amin RM, et al. Patient blood
management program improves blood use and clinical
478 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
outcomes in orthopedic surgery. Anesthesiology 2018; 129:
1082–91.
32. Froessler B, Palm P, Weber I, Hodyl NA, Singh R, Murphy EM.
The important role for intravenous iron in perioperative patient
blood management in major abdominal surgery: a randomized
controlled trial. Annals of Surgery 2016; 264: 41–6.
33. Massicotte L, Lenis S, Thibeault L, Sassine M-P, Seal RF, Roy A.
Effect of low central venous pressure and phlebotomy on blood
product transfusion requirements during liver transplantations.
Liver Transplantation 2006; 12: 117–23.
34. Schroeder RA, Collins BH, Tuttle-Newhall E, et al.
Intraoperative fluid management during orthotopic liver
transplantation. Journal of Cardiothoracic and Vascular
Anesthesia 2004; 18: 438–41.
Supporting Information
Additional supporting information may be found online via
the journal website.
Table S1. Prevalence of anaemia according to the
aetiology of liver disease.
Anaesthesia 2020 doi:10.1111/anae.14989

Original Article

The role of intra-operative cell salvage in patient blood

management for revision hip arthroplasty: a prospective
cohort study
A. J. R. Palmer,1 T. D. Lloyd,2 V. N. Gibbs,3 A. Shah,4 P. Dhiman,5 R. Booth,6 M. F. Murphy,7
A. H. Taylor,8 B. J. L. Kendrick,9 and collaborators#

1 NIHR Academic Clinical Lecturer in Trauma and Orthopaedics, 2 NIHR Academic Clinical Fellow in Trauma and
Orthopaedics, Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford,
Oxford, UK
3 NIHR Clinical Research Fellow and Specialty Trainee in Trauma and Orthopaedics, NHS Blood and Transplant, Oxford, UK
4 NIHR Doctoral Research Fellow and Specialty Trainee in Anaesthesia, Radcliffe Department of Medicine, University of
Oxford, Oxford, UK
5 Senior Statistician, Centre for Statistics in Medicine, University of Oxford, Oxford, UK
6 Transfusion Practitioner, Nuffield Orthopaedic Centre, Oxford University Hospitals NHS Foundation Trust, Oxford, UK
7 Professor of Transfusion Medicine and Consultant Haematologist, NHS Blood and Transplant and NIHR Biomedical
Research Centre, Oxford University Hospitals NHS Foundation Trust, Oxford, UK
8 Consultant Orthopaedic Surgeon, Oxford University Hospitals NHS Foundation Trust, Oxford, UK
9 Honorary Senior Lecturer and Consultant Orthopaedic Surgeon, Nuffield Department of Orthopaedics, Rheumatology and
Musculoskeletal Sciences, University of Oxford, Oxford, UK

Summary
Cell salvage is an important component of blood management in patients undergoing revision hip arthroplasty
surgery. However concerns regarding efficacy and patient selection remain. The aims of this study were to describe
intra-operative blood loss, cell salvage re-infusion volumes and red blood cell transfusion rates for revision hip
procedures and to identify factors associated with the ability to salvage sufficient blood intra-operatively to permit
processing and re-infusion. Data were collected from a prospective cohort of 664 consecutive patients undergoing
revision hip surgery at a single tertiary centre from 31 March 2015 to 1 April 2018. Indications for revision surgery were
aseptic (n = 393 (59%)) fracture (n = 160 (24%)) and infection (n = 111 (17%)). Salvaged blood was processed and re-
infused when blood loss exceeded 500 ml. Mean (SD) intra-operative blood loss was 1038 (778) ml across all
procedures. Salvaged blood was re-infused in 505 of 664 (76%) patients. Mean (SD) re-infusion volume was 253 (169)
ml. In total, 246 of 664 (37%) patients received an allogeneic red blood cell transfusion within 72 h of surgery. Patients
undergoing femoral component revision only (OR (95%CI) 0.41 (0.23–0.73)) or acetabular component revision only
(0.53 (0.32–0.87)) were less likely to generate sufficient blood salvage volume for re-infusion compared with revision
of both components. Compared with aseptic indications, patients undergoing revision surgery for infection (1.87
(1.04–3.36)) or fracture (4.43 (2.30–8.55)) were more likely to generate sufficient blood salvage volume for re-infusion.
Our data suggest that cell salvage is efficacious in this population. Cases where the indication is infection or fracture
and where both femoral and acetabular components are to be revised should be prioritised.

.................................................................................................................................................................
Correspondence to: A. J. R. Palmer
Email: antony.palmer@ndorms.ox.ac.uk
Accepted: 12 December 2019
Keywords: anaemia; arthroplasty; blood management; cell salvage; transfusion
#
For collaborators, see Appendix 1
Twitter: @ajrpalmer, @DocAShah

© 2020 Association of Anaesthetists 1

Anaesthesia 2020
Introduction
Revision hip arthroplasty is associated with significant peri-
operative blood loss and high allogeneic red blood cell
transfusion requirements [1]. Blood loss and red blood cell
transfusion are both independently associated with
increased morbidity and mortality [2, 3]. Strategies to
reduce peri-operative blood loss and allogeneic red blood
cell transfusion requirements include optimisation of pre-
operative anaemia [4], antifibrinolytic therapy [5] and intra-
operative cell salvage [6–8].
Whilst pre-operative anaemia optimisation and
antifibrinolytic therapy are indicated for all patients
undergoing revision hip and knee surgery, the role of intra-
operative cell salvage is less well defined [8, 9].
Interpretation of currently available data is limited by
heterogeneity of surgical procedures and patient chara-
cteristics, as well as uncertainties regarding estimates of
blood loss and re-infusion volumes [10, 11]. A previous
study of aseptic elective hip revision arthroplasty found
that adequate blood for re-infusion was collected in only
half of all patients [12].
Current guidelines vary in their recommendations for
cell salvage. The National Institute for Health and Care
Excellence recommends considering intra-operative cell
salvage in patients who are “expected to lose a very high
volume of blood, for example in cardiac and complex
vascular surgery, major obstetric procedures, and pelvis
reconstruction and scoliosis surgery” [9], whereas the
Association of Anaesthetists recommends its use when “it
can be expected to reduce the likelihood of allogeneic
(donor) red cell transfusion and/or severe postoperative
anaemia” [13]. Importantly the use of cell salvage may also
be limited by costs and resources [8]. Therefore, it is
essential to target patients who are most likely to benefit
from this intervention.
The aims of this study were to determine intra-operative
blood loss, cell salvage re-infusion volumes and allogeneic
transfusion rates for revision hip arthroplasty and to identify
factors associated with the ability to salvage sufficient blood
intra-operatively to permit processing and re-infusion.
Methods
We performed a prospective cohort study of all patients
undergoing revision hip arthroplasty who received intra-
operative cell salvage at the Nuffield Orthopaedic Centre,
Oxford University Hospitals NHS Foundation Trust, Oxford,
UK, between 31 March 2015 and 01 April 2018. Revision
was defined as removal or exchange of any hip prosthesis
component. Indications for revision hip surgery were
categorised as either aseptic, fracture or infection. Aseptic
2 © 2020 Association of Anaesthetists
Palmer et al. | Intra-operative cell salvage in revision hip surgery
indications were primarily implant wear or loosening that
was causing pain or mechanical symptoms. There were no
exclusion criteria. We report our findings in accordance with
the Strengthening the Reporting of Observational Studies in
Epidemiology (STROBE) statement. Research Ethics
Committee approval was not required as per the Health
Research Authority assessment tool. The study received
Caldicott Guardian approval.
Pre-operative assessment was performed within
6 weeks of surgery for elective cases and on admission to
hospital for emergency cases. Routine pre-operative
investigations included haemoglobin and haematocrit
measurements. Haemoglobin and haematocrit were
measured intra-operatively using blood gas analysis with
a frequency of measurement determined by the rate of
blood loss and patient haemodynamic status. A formal
laboratory measurement of haemoglobin and haematocrit
was also performed on the morning following surgery. All
patients routinely received general anaesthesia with or
without neuraxial anaesthesia followed by 1 g intravenous
(i.v.) tranexamic acid before skin incision. This was
followed by a 2 g topical dose of tranexamic acid at the
discretion of the operating surgeon before skin closure.
Drains were not routinely inserted. Routine postoperative
thromboprophylaxis consisted of low-molecular weight
heparin commenced 6 h after skin closure and thrombo-
embolic deterrent stockings. There was no change in
routine surgical or anaesthetic technique during the
period of study.
As per our local institution’s policy, we employed intra-
operative cell salvage collection for all revision hip
arthroplasty procedures except during first stage revision
and debridement and implant retention for infection, when
it was used at the discretion of the senior surgeon and
anaesthetist. We did not consider infection as a contra-
indication to cell salvage, but any fluid visibly contaminated
by infection, cement or metallic debris was not collected.
Wound irrigation was performed using saline, and if
chlorhexidine solution was used, it was reserved until the
end of the procedure once cell salvage was complete. A
leucocyte reduction filter (40 lm) was used for infected
cases. We commenced all procedures with the cell salvage
system in ‘collect only’ mode. Salvaged blood was
processed to re-suspend red blood cells with a haematocrit
of 50–60% once estimated blood loss exceeded 500 ml.
Cell salvage was performed using Sorin XtraTM (LivaNova,
London, UK) machines. Bowl size was selected based on the
volume of salvaged blood (range 55–225 ml). In keeping
with national guidelines [9], our haemoglobin red blood cell
transfusion threshold was 70 g.l
1, except for patients with
Palmer et al. | Intra-operative cell salvage in revision hip surgery
acute coronary syndrome where the threshold was 80 g.l
1.
Red blood cell transfusions on the day of surgery were
defined as those taking place before midnight.
Data were collected from local electronic patient
records and cell salvage databases. Data included: patient
demographics (age, sex, ASA, BMI); procedure performed
and indication; cell salvage records (intra-operative blood
loss, re-infused blood volume and haematocrit);
haemoglobin and haematocrit measurements (pre-
operative, intra-operative and postoperative); and
allogeneic red blood cell transfusion records.
Estimated intra-operative blood loss was calculated
according to the following formula [14]:
Blood loss (ml) ¼ Estimated blood volume (ml)
 
Hct procedure start ð%Þ
Hct pretransfusion ð%Þ
 :
(Hct procedure start ð%Þ þ Hct pretransfusion ð%ÞÞ=2
where estimated blood volume is 65 ml.kg
1 for females

1
and 75 ml.kg for males using ideal body weight, the
proportion of estimated blood loss re-infused was
calculated as:
Cell salvage efficiencyð%Þ
 
re-infusion volume (ml)  re-infusion Hct (%)
¼ :
Hct procedure start ð%Þ  estimated blood loss (ml)
Estimated increase in haemoglobin concentration after
re-infusing salvaged blood was calculated using the
formula:
Change in Hb ¼ Hb pre - transfusion (g/L)
 
re- infusion volume (ml)  reinfusion Hct (%)
 1þ : generate sufficient blood salvage for re-infusion when
estimated blood volume (ml)  Hct pre-transfusion (%)
The volume of salvaged blood was standardised to
haematocrit 60%. Two outcomes of interest were assessed,
namely whether a sufficient volume of blood was salvaged
intra-operatively to allow for processing and re-infusion and
standardised volume of re-infused blood (haematocrit
60%). We performed logistic and univariate linear
regression analyses to explore the associations between
patient demographics, pre-operative anaemia, procedure
performed and indication for surgery with each outcome of
interest, respectively. Factors that demonstrated a indication for surgery was not statistically associated with re-
statistically significant relationship in univariate analysis
were included in multivariable regression models. Total
femoral replacement was included as revision of the femoral
and acetabular component. A value of p < 0.05 was
considered statistically significant.
© 2020 Association of Anaesthetists
Anaesthesia 2020
Results
A total of 664 patients were included in our analysis. Mean
intra-operative blood loss exceeded 500 ml for all revision
hip procedures and 1000 ml for procedures where the
femoral and acetabular components were revised (Tables 1
and 2). Thirtyseven percent (244/664) of patients required
an intra-operative or postoperative allogeneic red blood
cell transfusion within 72 h of surgery (Fig. 1). A median
(IQR [range]) of 2 (2–3 [0–12]) units were transfused per
patient. Eight patients (1.2%) required more than four units
of red blood cells on the day of surgery (range 0–12 units).
There were no changes in pre-operative and postoperative
haemoglobin concentration or allogeneic red blood cell
transfusion rates over the course of the study (see also
Supporting Information, Figure S1).
A sufficient volume of blood was salvaged for processing
and re-infusion in 76% (505/664) of procedures. There was no
difference in patient characteristics for patients where
sufficient blood was salvaged for re-infusion compared with
patients where it was not (see also Supporting Information,
Table S1). Mean (SD) re-infusion volume was 253 (169) ml
(standardised haematocrit 60%) (Fig. 2), equating to an
estimated mean (SD) increase in haemoglobin of 10.2 (6.8)
g.l
1 per patient receiving autologous blood. An estimated
one third of blood volume lost intra-operatively was returned
to the patient (Tables 1 and 2). Patient factors age, sex, BMI
and haemoglobin concentration at the start of the procedure
were not statistically associated with the ability to salvage
sufficient blood for re-infusion.
Patients undergoing femoral component revision only
(OR (95%CI) 0.41 (0.23–0.73)) or acetabular component
revision only (OR 0.53 (0.32–0.87)) were less likely to
compared with when both the femoral and acetabular
components were revised. Compared with aseptic
indications, patients undergoing revision surgery for
infection (OR 1.87 (1.04–3.36)) or fracture (OR 4.43 (2.30–
8.55)) were more likely to generate sufficient blood salvage
for re-infusion (Table 3).
Smaller blood volumes were re-infused with increasing
age (coefficient (95%CI)
1.9 (
2.9–0.8)) and higher
volumes were re-infused in males (coefficient (95%CI) 52.0
(25.0–79.0)). A smaller volume of blood was re-infused
when only revising a single component, whereas the
infusion volume (Table 4).
Discussion
In this study, the mean intra-operative blood loss for each
revision hip procedure exceeded 500 ml; intra-operative
3
Anaesthesia 2020 Palmer et al. | Intra-operative cell salvage in revision hip surgery

Table 1 Patient characteristics for each procedure. Values are mean (SD) or number (proportion).
Revise Revise Revise Exploration
femoral and femoral acetabular First Second and
acetabular component component Revision to stage stage modular All
component only only total femur revision revision exchange procedures

Number of 216 169 104 30 46 68 31 664

patients
Age; years 70.4 (13.0) 75.1 (12.1) 68.7 (13.9) 64.6 (14.5) 67.1 (12.6) 66.2 (12.3) 68.5 (12.4) 70.3 (13.2)
Pre-operative 125.3 (17.8) 123.2 (19.5) 123.1 (18.2) 125.9 (19.3) 115.2 (17.2) 117.5 (17.9) 120.3 (21.0) 122.6 (18.8)
haemoglobin;
g.l
1
Intra- 1069.5 (672.3) 955.6 (828.0) 762.9 (502.0) 1545.3 (1002.3) 1479.3 (1031.5) 1247.5 (819.0) 585.3 (436.9) 1037.6 (778.3)
operative
blood loss;
ml
Patients with 170 (79%) 134 (78%) 68 (65%) 29 (100%) 36 (78%) 58 (87%) 12 (39%) 507 (76%)
sufficient
collection for
re-infusion
Volume of 260.3 (163.4) 225.1 (150.9) 217.7 (135.2) 259.1 (163.1) 300.9 (227.5) 314.7 (208.3) 190.3 (64.6) 253.0 (168.8)
re-infused
blood; mla
Proportion 34.7 (11.5) 34.5 (10.5) 38.2 (12.6) 28.6 (10.8) 33.3 (12.7) 41.7 (13.7) 37.9 (11.6) 35.5 (12.0)
of blood
loss salvaged;
%
Patients 75 (35%) 59 (35%) 33 (32%) 10 (35%) 20 (43%) 31 (46%) 16 (52%) 244 (37%)
receiving
peri-operative
red blood
cell transfusion
Day 1 95.9 (16.0) 96.2 (14.8) 96.3 (14.3) 95.4 (18.0) 90.0 (13.1) 94.8 (12.6) 96.3 (16.3) 95.5 (15.0)
postoperative
haemoglobin;
g.l
1
a
Excludes cases where inadequate blood for re-infusion. Standardised to haematocrit 60%.

cell salvage provided sufficient autologous blood for re-
infusion in over three-quarters of these patients. Where
sufficient blood was salvaged for processing and re-
infusion, this equated to nearly one unit of packed red
blood cells per patient, yet still more than one third of
patients required a peri-operative allogeneic red blood cell
transfusion. Patients undergoing revision hip arthroplasty
for infection or fracture and those where both femoral and
acetabular components were revised, were more likely to
generate sufficient blood salvage for re-infusion. We found
no association between patient characteristics and pre-
operative haemoglobin concentration, and the utility of cell
salvage.
Our estimated intra-operative blood
comparable with other studies with an average of
1000 ml across all procedures [15], as was the average
volume of re-infusion of 250 ml, although
frequently do not specify the haematocrit of the re-
infused blood [12, 15]. The proportion of patients with
sufficient blood salvage for re-infusion in our study is
comparable or higher than in other studies [12, 15]. In a
cohort of 298 patients undergoing aseptic revision hip
arthroplasty, re-infusion was only possible in 54% of
patients [12]. In a cohort of 210 patients undergoing
revision hip arthroplasty for a range of indications, cell
salvage was used for 88 patients of whom 68 had
sufficient blood salvage for re-infusion [15]. Potential
reasons for an increased salvage yield in our study
include more fastidious suctioning, salvage from swabs or
possibly a different set up and performance of cell
salvage machines.
loss was Our red blood cell transfusion rates are lower than
those reported in other studies, where rates of 57% [12] and
58% [15] have been described. The mean pre-operative
studies haemoglobin in our cohort was 124 g.l
1, which is similar to
other studies [15]. Our lower allogeneic red blood cell

4 © 2020 Association of Anaesthetists

Palmer et al. | Intra-operative cell salvage in revision hip surgery Anaesthesia 2020

Table 2 Patient characteristics according to indication for surgery. Values are mean (SD) or number (proportion).
Aseptic Fracture Infection
Number of patients 393 160 111
Age; years 68.7 (12.7) 77.2 (11.5) 66.4 (13.8)
Pre-operative haemoglobin; g.l
1 123.4 (18.8) 123.5 (18.5) 118.7 (18.5)
Intra-operative blood loss; ml 945.0 (779.2) 945.9 (563.8) 1488.8 (882.8)
Patients with sufficient collection for re-infusion 271 (69%) 142 (88%) 94 (84%)
Volume of re-infused blood; mla 267.6 (179.2) 206.9 (127.0) 279.2 (180.8)
Proportion of blood loss salvaged; % 38.4 (11.9) 33.2 (10.9) 31.1 (11.9)
Patients receiving peri-operative red blood cell transfusion 137 (35%) 57 (36%) 50 (45%)
Day 1 Postoperative haemoglobin; g.l
1 96.8 (14.5) 95.3 (15.1) 91.3 (16.3)
a
Excludes cases where inadequate blood for re-infusion. Standardised to haematocrit 60%.

emphasis on patient blood management including use of

Figure 1 Proportion of patient cohort receiving allogeneic
blood transfusions (n = 644). Blue, intra-operative; red,
postoperative.
Figure 2 Histogram of re-infusion volumes (standardised
to haematocrit 60%) (n = 507).
transfusion rates may reflect: more efficient cell salvage; a
smaller volume of blood loss secondary to surgical
technique; preference for combined general and spinal
anaesthesia together with the effects of the current
antifibrinolytic agents (such as tranexamic acid); and
adherence to restrictive transfusion thresholds. The highest
rate of red blood cell transfusion was for patients
undergoing exploration and modular exchange. This
procedure had the lowest intra-operative blood loss of all
procedures and as such, fewer patients had sufficient blood
collection to allow re-infusion.
An estimated 35% of intra-operative blood loss was
salvaged and re-infused during surgery. The greatest
efficiency was achieved during second stage revision,
perhaps due to the absence of contaminated blood as the
implant has previously been removed and infection
eradicated. The lowest efficiency was during total femoral
replacement, likely due to difficulty containing blood loss
within an extensive surgical field. The ability to salvage
sufficient blood loss for re-infusion is dependent on the
volume of intra-operative blood loss; however, the
decision to employ cell salvage should be based on
information available pre-operatively. Data from our study
suggest that patient factors are not important
determinants of the ability to collect sufficient blood for re-
infusion. Pre-operative anaemia was not associated with
the ability to collect sufficient blood for re-infusion or the
volume of re-infused blood. As observed previously,
smaller volumes of blood were salvaged with increasing
age [12] and re-infusion volumes were greater in males
than females, which are thought to reflect greater
circulating volumes in males.
The most important determinants of whether sufficient
blood is salvaged for re-infusion was the procedure
performed and the indication for surgery. The odds of
collecting sufficient blood were halved when revising a
single component compared with revising the acetabular
and femoral components. The odds of collecting sufficient
blood were almost four times greater when the indication

© 2020 Association of Anaesthetists 5

Anaesthesia 2020 Palmer et al. | Intra-operative cell salvage in revision hip surgery

Table 3 Factors associated with salvaging sufficient blood for re-infusion.

Univariate logistic regression Multivariable logistic regression
OR 95%CI p value OR 95%CI p value
Age 0.99 0.97–1.00 0.050
Sex:
Female Reference Reference
Male 0.99 0.98–1.01 0.945
BMI 0.99 0.98–1.01 0.583
Pre-operative haemoglobin concentration 0.99 0.98–1.00 0.122
Procedure:
Revise femur and acetabulum Reference Reference Reference Reference Reference Reference
Revise femur only 0.79 0.51–1.25 0.320 0.41 0.23–0.73 0.002
Revise acetabulum only 0.42 0.26–0.68 < 0.001 0.53 0.32–0.87 0.013
Exploration and modular exchange 0.14 0.06–0.30 < 0.001 0.16 0.07–0.34 < 0.001
Indication:
Aseptic Reference Reference Reference Reference Reference Reference
Infection 2.28 1.32–3.94 0.003 1.87 1.04–3.36 0.037
Fracture 3.30 1.95–5.57 < 0.001 4.43 2.30–8.55 < 0.001
BMI, Body mass index.

Table 4 Factors associated with volume of re-infusion.

Univariate logistic regression Multivariable logistic regression
OR 95%CI p value OR 95%CI p value
Age
1.9
3.0 to
0.9 < 0.001
1.9
2.9 to
0.8 0.001
Sex:
Female Reference Reference Reference Reference Reference Reference
Male 53.1 25.3–81.0 < 0.001 52.0 25.0 to 79.0 < 0.001
BMI 1.3
0.2 to 2.8 0.093
Pre-operative haemoglobin concentration
0.2
0.6 to 1.0 0.617
Procedure:
Revise femur and acetabulum Reference Reference Reference Reference Reference Reference
Revise femur only
52.6
85.2 to
19.9 0.002
39.1
79.4 to 1.1 0.056
Revise acetabulum only
84.4
123.5 to
45.4 < 0.001
69.2
109.2 to
29.3 < 0.001
Exploration and modular exchange
144.6
210.2 to
79.0 < 0.001
143.6
208.2 to
79.1 < 0.001
Indication:
Aseptic Reference Reference Reference Reference Reference Reference
Infection 63.0 24.7–101.4 0.001 38.1
0.4 to 76.7 0.053
Fracture 0.3
33.0 to 33.7 0.985 15.1
26.5 to 56.7 0.477
BMI, Body mass index.

was an emergency revision periprosthetic fracture employing cell salvage whenever anticipated blood loss
compared with an elective revision for an aseptic indication. exceeds 500 ml [8]. The results of our study provide surgical
These findings can help allocate cell salvage when there is teams with an estimation of intra-operative blood loss for
limited resource, which is an important barrier to different revision hip arthroplasty procedures, although we
widespread utilisation [8]. recognise there may be considerable variation between
Recent Association of Anaesthetists guidelines on cell patients and centres. The proposed threshold of 500 ml for
salvage for peri-operative blood conservation recommend using cell salvage is lower than suggested in existing

6 © 2020 Association of Anaesthetists

Palmer et al. | Intra-operative cell salvage in revision hip surgery
guidelines [9, 16]. As a result, there will be an increased
number of patients when insufficient blood is salvaged for
re-infusion, reducing the cost effectiveness of this
intervention. Costs can be reduced by setting up the cell
salvage machine in ‘collect only’ mode [8], as practised in
our institution, although consumable and staff costs will
remain. There is a need for an updated analysis of the cost
effectiveness of cell salvage when used alongside other
blood conservation strategies.
A revision hip arthroplasty procedure where cell
salvage may not be warranted is prosthesis exploration
and modular exchange, when only 35% of patients
received salvaged blood. The odds of collecting sufficient
blood for re-infusion during this procedure were one-
seventh of those when revising femoral and acetabular
components. The presence of tumour or infection is not a
contra-indication for cell salvage at our institution and this
is supported by an increasing number of studies [17, 18].
Instead, revision hip procedures performed for infection
are associated with higher blood loss and cell salvage re-
infusion volumes than for aseptic indications. During our
study period, cell salvage was employed for an increasing
proportion of infected patients, hence the smaller number
of first stage revisions than second stage revisions in the
cohort.
One strength of this study is that it describes one
of the largest cohort of patients undergoing revision hip
arthroplasty for a broad range of indications with the
use of intra-operative cell salvage. Revision hip
procedures are heterogenous in nature and include
cement-in-cement femoral revisions where the femoral
canal is not entered, through to extended trochanteric
osteotomies that increase the volume of blood loss [12].
However, the nature of the procedure performed
frequently depends on intra-operative findings and our
study focuses on pre-operative factors that can be used
to guide resource utilisation. Our study has limitations.
Calculated blood loss, cell salvage efficiency and
change in haemoglobin concentration after autologous
blood re-infusion are estimations that do not account for
i.v. fluid therapy and intra-operative red blood cell
transfusions. Anaesthetic technique may influence blood
loss, which was not explored in this study. We are
unable to determine whether intra-operative cell salvage
prevents red blood cell transfusion or its cost
effectiveness.
In conclusion, patients undergoing revision hip
arthroplasty for infection or fracture and those undergoing
revision of both femoral and acetabular components are
more likely to generate sufficient blood salvage for re-
© 2020 Association of Anaesthetists
Anaesthesia 2020
infusion. Our findings may prove valuable in a clinical
context of resource utilisation.
Acknowledgements
Support was received from the National Institute for Health
Research (NIHR) through individual awards and the NIHR
Oxford Biomedical Research Centre. No external funding or
competing interests declared.
Appendix 1
Collaborators: A. McGill (Consultant Anaesthetist Oxford
University Hospitals, Oxford NHS Foundation Trust, UK), A.
Alvand Consultant Orthopaedic Surgeon and Honorary Senior
Lecturer and A. J. Carr Nuffield Professor or Orthopaedics
(Nuffield Department of Orthopaedics, Rheumatology, and
Musculoskeletal Sciences, University of Oxford, UK).
References
1. Goel R, Buckley P, Sterbis E, Parvizi J. Patients with infected total
hip arthroplasty undergoing 2-stage exchange arthroplasty
experience massive blood loss. Journal of Arthroplasty 2018;
33: 3547–50.
2. Kim JL, Park JH, Han SB, Cho IY, Jang KM. Allogeneic blood
transfusion is a significant risk factor for surgical-site infection
following total hip and knee arthroplasty: a meta-analysis.
Journal of Arthroplasty 2017; 32: 320–5.
3. Fowler AJ, Ahmad T, Phull MK, Allard S, Gillies MA, Pearse RM.
Meta-analysis of the association between preoperative
anaemia and mortality after surgery. British Journal of Surgery
2015; 102: 1314–24.
4. Munoz M, Acheson AG, Auerbach M, et al. International
consensus statement on the peri-operative management of
anaemia and iron deficiency. Anaesthesia 2017; 72: 233–47.
5. Peck J, Kepecs DM, Mei B, et al. The effect of preoperative
administration of intravenous tranexamic acid during revision
hip arthroplasty: a retrospective study. Journal of Bone and
Joint Surgery 2018; 100: 1509–16.
6. Sullivan IJ, Ralph CJ. Obstetric intra-operative cell salvage: a
review of an established cell salvage service with 1170 re-
infused cases. Anaesthesia 2019; 74: 976–83.
7. Kelleher A, Davidson S, Gohil M, et al. A quality assurance
programme for cell salvage in cardiac surgery. Anaesthesia
2011; 66: 901–6.
8. Klein AA, Bailey CR, Charlton AJ, et al. Association of
Anaesthetists guidelines: cell salvage for peri-operative blood
conservation 2018. Anaesthesia 2018; 73: 1141–50.
9. National Institute for Health and Care Excellence. Blood
transfusion. NG24, 2015. https://www.nice.org.uk/guidance/
ng24 (accessed 16/12/2019).
10. van Bodegom-Vos L, Voorn VM, So-Osman C, et al. Cell
salvage in hip and knee arthroplasty: a meta-analysis of
randomized controlled trials. Journal of Bone and Joint Surgery
2015; 97: 1012–21.
11. Carless PA, Henry DA, Moxey AJ, O’Connell D, Brown T,
Fergusson DA. Cell salvage for minimising perioperative
allogeneic blood transfusion. Cochrane Database of
Systematic Reviews 2010; CD001888.
12. Greenky M, Shaner J, Rasouli MR, Han SB, Parvizi J, Hozack WJ.
Intraoperative blood salvage in revision total hip arthroplasty:
who benefits most? Journal of Arthroplasty 2014; 29: 1298–
300.
7
Anaesthesia 2020
13. Klein AA, Arnold P, Bingham RM, et al. AAGBI guidelines: the
use of blood components and their alternatives 2016.
Anaesthesia 2016; 71: 829–42.
14. Gross JB. Estimating allowable blood loss: corrected for
dilution. Anesthesiology 1983; 58: 277–80.
15. Walsh TS, Palmer J, Watson D, et al. Multicentre cohort study of
red blood cell use for revision hip arthroplasty and factors
associated with greater risk of allogeneic blood transfusion.
British Journal of Anaesthesia 2012; 108: 63–71.
16. Kozek-Langenecker SA, Afshari A, Albaladejo P, et al.
Management of severe perioperative bleeding: guidelines
from the European Society of Anaesthesiology. European
Journal of Anaesthesiology 2013; 30: 270–382.
17. Kumar N, Zaw AS, Kantharajanna SB, Khoo BL, Lim CT, Thiery
JP. Metastatic efficiency of tumour cells can be impaired by
intraoperative cell salvage process: truth or conjecture?
Transfusion Medicine 2017; 27(Suppl. 5): 327–34.
8 © 2020 Association of Anaesthetists
Palmer et al. | Intra-operative cell salvage in revision hip surgery
18. Esper SA, Waters JH. Intra-operative cell salvage: a fresh look at
the indications and contraindications. Blood Transfusion 2011;
9: 139–47.
Supporting Information
Additional supporting information may be found online via
the journal website.
Figure S1 Mean pre-operative (a) and postoperative
(b) haemoglobin concentration, and proportion of patients
receiving allogeneic blood transfusion within 72 h of
surgery (c) for each year of the study (with 95%CI).
Table S1 Cohort characteristics for individuals with and
without sufficient salvaged blood for re-infusion.
Anaesthesia 2019 doi:10.1111/anae.14950

Original Article

Prophylactic phenylephrine and fluid co-administration to

reduce spinal hypotension during elective caesarean
section in a resource-limited setting: a prospective
alternating intervention study
A. S. Buthelezi,1 D. G. Bishop,2 R. N. Rodseth3,4 and R. A. Dyer5

1 Registrar, 2 District Clinical Specialist, 3 Associate Professor, Metropolitan Department of Anaesthetics, Critical Care
and Pain Management, Pietermaritzburg, and School of Clinical Medicine, College of Health Sciences, Nelson R Mandela
School of Medicine, University of KwaZulu-Natal, Durban, South Africa
4 Outcomes Research Consortium, Cleveland Clinic, Cleveland, OH, USA
5 Emeritus Professor, Department of Anaesthesia and Peri-operative Medicine, Groote Schuur Hospital, Faculty of Health
Sciences, University of Cape Town, South Africa

Summary
Spinal hypotension is a common and clinically important problem during caesarean section. Current consensus
recommendations for resource-rich settings suggest the use of a titrated phenylephrine infusion, in
combination with fluid coloading, for prevention of maternal hypotension. In resource-limited settings, where
syringe drivers are unavailable, these recommendations advise the addition of 500 lg phenylephrine to the first
1 l of intravenous fluid given after initiation of spinal anaesthesia, with additional vasopressor boluses as
required. This prospective, alternating intervention study compared the use of a conventional phenylephrine
rescue bolus strategy for prevention of hypotension, defined as systolic arterial pressure < 90 mmHg, with a
phenylephrine infusion given according to the consensus recommendation. We studied 300 women having
elective caesarean section. There were 77 (51%) women who developed hypotension in the bolus group vs. 55
(37%) in the phenylephrine infusion group (p = 0.011). This represented a 29% reduction in hypotension, with a
number needed to treat of 6.8. The six highest systolic arterial pressure readings occurred in the phenylephrine
infusion group (range 166–188 mmHg), and there were four instances of bradycardia (heart rate < 50
beats.min 1) with preserved systolic arterial pressure in each group. There were no adverse clinical sequelae,
and no differences in neonatal Apgar scores in either group. The consensus recommendation for
phenylephrine and fluid co-administration in resource-limited settings appears effective in preventing maternal
hypotension, but at the cost of sporadic systolic hypertension.

.................................................................................................................................................................
Correspondence to: D. G. Bishop
Email: davidgbishop@gmail.com
Accepted: 14 November 2019
Keywords: anaesthesia, spinal; caesarean section; complications hypotension, obstetric, phenylephrine
Twitter: @DavidGray37

Introduction cardiac arrest and death [2, 3]. In South Africa, the majority
Hypotension after spinal anaesthesia is a common and of anaesthesia-related mortality is associated with spinal
important problem during caesarean section, which may be anaesthesia [2, 4]. The incidence of maternal hypotension
associated with nausea, vomiting [1], loss of consciousness, may approach 70%, depending on the definition [5, 6]. A

© 2019 Association of Anaesthetists 1

Anaesthesia 2019 Buthelezi et al. | Phenylephrine and fluid co-administration during caesarean section
recently published consensus statement recommends the
prevention of maternal hypotension with a phenylephrine
infusion titrated to between 25 and 50 lg.min 1
, in
combination with fluid coloading [3]. It further recommends
that, if a syringe driver is unavailable, it is acceptable to add
500 lg phenylephrine to the first litre of intravenous (i.v.)
fluid and commence rapid administration at the onset of
spinal anaesthesia. This statement is based on expert
opinion, and has not yet been tested in clinical practice.
Studies are required evaluating methods for the reliable
administration of phenylephrine in resource-limited settings
[3, 7, 8]. This study aimed to compare intermittent
phenylephrine boluses with prophylactic combined
phenylephrine and fluid co-administration for the
prevention of spinal hypotension in women undergoing
elective caesarean section in a resource-limited setting. We
hypothesised that the infusion would be a practical,
effective and safe alternative for the management of spinal
hypotension in this context.
Methods
We undertook a prospective, single-centre, alternating
intervention study in women undergoing elective caesarean
section under spinal anaesthesia in Edendale Hospital,
KwaZulu-Natal, South Africa. This hospital is a regional
referral centre that caters for both routine and complex
obstetric cases, with a caesarean section rate of
approximately 30%. The obstetric anaesthesia service is
provided primarily by medical officers, but with specialist
anaesthesia cover if required. Ethics approval for the study
was obtained from the University of KwaZulu-Natal, the
KwaZulu-Natal Department of Health Ethics Review
Committee and Edendale Hospital. Written informed
consent was obtained from all women on the day of surgery.
We included consecutive ASA physical status 1–2 women
having elective caesarean section under spinal anaesthesia.
Two protocols for the management of spinal hypotension
were used, alternating at 2-weekly intervals until the
required sample size in each group was reached. The
choice of the first protocol was decided by the flip of a coin.
The study was non-blinded.
All women were fasted overnight, and were given oral
sodium citrate 30 ml and 1 g cefazolin i.v. prophylactically.
Baseline heart rate and noninvasive blood pressure were
recorded in the operating theatre by the attending
anaesthetist using a Lifescope© BSM 3562 monitor (Nihon-
Koden, Tokyo, Japan), with the woman in the left lateral
position and the cuff on the right arm. An 18-G i.v. cannula
(B. Braun Introcan©; B. Braun Melsungen AG, Melsungen,
Germany) was inserted, and a 1-l Ringer’s lactate bag
2 © 2019 Association of Anaesthetists
1
connected using a 20 drop.ml giving set. In the
phenylephrine infusion group, we added 500 lg of
phenylephrine to the fluid bag. The infusion was not started
at this stage.
Single shot spinal anaesthesia using 9 mg hyperbaric
bupivacaine 0.5% with 10 lg fentanyl was performed under
aseptic conditions, with the woman in the sitting position,
using a 25-G pencil point spinal needle at the L3/4
interspace. The woman was then immediately placed
supine with a 20o obstetric wedge under the right hip to
provide uterine displacement.
In both groups, the i.v. fluid was then administered
rapidly as a co-load, without using a pressure bag. Non-
invasive blood pressure was measured at 1-min intervals
until delivery of the neonate.
To simplify vasopressor management in this practice
environment, hypotension was defined as systolic arterial
pressure < 90 mmHg. This was treated similarly in both
groups, using 50–100 lg boluses of phenylephrine if
maternal heart rate was ≥ 70 beats.min 1
, or 5–10 mg
ephedrine for heart rate < 70 beats.min 1
. If systolic
pressure increased to > 20% above baseline in the
phenylephrine infusion group, the infusion was stopped
and recommenced once the systolic pressure had returned
to within 10% of baseline. Bradycardia < 50 beats.min 1
was treated with 0.5 mg atropine if systolic pressure was
< 90 mmHg, and by stopping the phenylephrine infusion if
systolic pressure was ≥ 90 mmHg. The block level was
determined using loss of sensation to a cold stimulus.
Data collection was terminated on delivery of the
neonate. Following delivery, oxytocin 3 IU was
administered, and infusion of a further 1 l Ringer’s lactate
containing 7 IU oxytocin but no phenylephrine was started
in both groups.
We recorded neonatal Apgar scores and any intra-
operative adverse maternal outcomes including cardiac
arrest; death within 24 h of caesarean section; and
admission to the intensive care unit (ICU).
Data were entered into a Microsoft Excel© spreadsheet
(Microsoft Corporation, Redmond, WA, USA) by one of the
study investigators, and then checked by a separate study
investigator to ensure data fidelity. They were then exported
to a statistical analysis programme (Stata Release 15.1; Stata
Corp LP, College Station, TX, USA) for further analysis.
The primary outcome was the incidence of hypotension
from the commencement of spinal anaesthesia until delivery
of the neonate. The sample size calculation was based on
previous studies from our unit [9, 10]. The incidence of
hypotension was 39%, reducing to 24% with a fixed-rate
prophylactic phenylephrine infusion. To identify a similar
Buthelezi et al. | Phenylephrine and fluid co-administration during caesarean section Anaesthesia 2019

reduction in the incidence of hypotension and to achieve
80% power with an alpha value of 0.05, we required 150
women in each of the study groups. The Shapiro–Wilk test
was used for normality testing. We used Student’s t-test,
Wilcoxon–Mann–Whitney test and Chi-square test as
appropriate. For all analyses, a p value of < 0.05 defined
statistical significance.
Results
Enrolment for the study took place between May 2018 and
December 2018. Figure 1 shows the patient recruitment
diagram. The final analysis included 150 women each in the
bolus and the phenylephrine infusion groups. The
characteristics of the women are shown in Table 1.
The incidence of spinal hypotension was 77 (51%) in the
bolus group vs. 55 (37%) in the phenylephrine infusion
group (p = 0.01). This represented a 29% reduction in
hypotension, with a number needed to treat of 6.8. The
doses of phenylephrine are given in Table 2.
There were four instances of bradycardia
1
< 50 beats.min without hypotension in each group, with
no women requiring treatment with atropine. The six
highest systolic pressure readings (range 166–188 mmHg)
occurred in the phenylephrine infusion group. Rescue
ephedrine boluses were administered to 19 women in the
bolus group and 17 women in the phenylephrine infusion
group.
Maternal and neonatal outcomes are shown in Table 3.
There were no admissions to the ICU or serious morbidity.
Discussion
Maternal mortality in resource-limited environments
remains unacceptably high [11, 12]. Maternal mortality
directly related to anaesthesia constitutes approximately 3%
of maternal mortality in South Africa, although this may be
an underestimation due to the limitations of the national
data collection process [2, 4]. Spinal anaesthesia for
caesarean section has maternal and neonatal benefits;
however, 78% of deaths related to anaesthesia in South
Africa occur during spinal anaesthesia, and more than half
of these are associated with untreated spinal hypotension
(personal communication, National Anaesthetic
Co-ordinator, National Committee for Confidential
Enquiries into Maternal Deaths).
A recent consensus statement recommends the use of
routine prophylactic phenylephrine infusions in
combination with fluid loading and left lateral uterine
displacement to minimise the incidence of hypotension
after spinal anaesthesia [3]. Titrated phenylephrine infusions
require both appropriate syringe drivers, as well as skilled
anaesthetists trained to adjust the rate of infusion in
response to maternal haemodynamic changes. In resource-
limited settings, these elements may be unavailable, and
simpler methods of administration are desirable [8]. The
consensus statement suggested alternative approaches for
resource-limited environments. We have evaluated one of
these methods, and shown that prophylactic phenylephrine
co-administration with fluid coloading was effective in
reducing the incidence of maternal hypotension in a
resource-limited setting.
Maternal hypotension during spinal anaesthesia occurs
predominantly from a reduction in systemic vascular
resistance secondary to sympathetic blockade, and
phenylephrine remains the agent of choice in countering
this effect [13]. Dose-finding studies in resource-rich
environments have established that the optimal rate for

Paents recruited
(n = 305) Paents excluded (n = 5)
• Incorrect bupivacaine dose
(n = 2)
• Failed spinal anaesthesia (n = 2)
• Refused consent (n = 1)
Eligible paents (n = 300)

Phenylephrine Phenylephrine
bolus group infusion group
(n = 150) (n = 150)

Figure 1 CONSORT diagram of patient recruitment.

© 2019 Association of Anaesthetists 3

Anaesthesia 2019 Buthelezi et al. | Phenylephrine and fluid co-administration during caesarean section

Table 1 Physical characteristics of 300 women receiving environments where even syringe drivers are lacking, co-
phenylephrine boluses only vs. phenylephrine infusion. administering phenylephrine along with i.v. fluid co-load
Values are mean (SD) or median (IQR [range]). has theoretical benefit. We added 500 lg of phenylephrine
Bolus Infusion to 1 l of Ringers lactate, which approximates to 25–
n = 150 n = 150 p value
50 lg.min 1
if given over 10–20 min. We used an 18-G
Age; years 29.5 (5.4) 29.9 (5.3) 0.58 Introcan cannula, which allows fluid administration at up to
Weight; kg 88.7 (19.2) 87.8 (20.1) 0.70 100 ml.min 1
without a pressure bag. This would translate
Height; cm 160.4 (7.1) 160.2 (7.9) 0.81 to a 10-min period to infuse 1 l of fluid and 500 lg
Gestation; 38.8 (1.4) 38.7 (1.3) 0.47 phenylephrine (50 lg.min 1
) in laboratory conditions
weeks
without a pressure bag. If a 16-G cannula is used, it is
Parity 1 (1–2 [0–6]) 1 (1–2 [0–5]) 0.89
possible to administer the fluid twice as fast at
1
210 ml.min . The mean (SD) i.v. fluid administered in our
administration of phenylephrine is 25–50 lg.min 1
[14, 15]. study was 789 (389) ml, suggesting that 25 lg.min 1
of
A fixed, low-dose 25 lg.min 1
phenylephrine infusion is phenylephrine and 10 ml.kg 1
fluid co-load was achieved.
effective in reducing severe maternal hypotension in a The consensus statement further recommends that the
resource-limited setting [9]. Fixed-rate infusions are simpler anaesthetist should aim to maintain systolic pressure at
to use and have a low side-effect profile, due to the lower ≥ 90% of baseline using the infusion, but use vasopressor
dose administered. They are recommended for use where boluses if the systolic pressure decreases < 80% of the
syringe drivers are available, but the expertise to titrate the baseline. Rather than using the individual woman’s baseline
infusion may be lacking [3]. A further advantage of this pressure, we instead chose to use an absolute value of
method is that phenylephrine may be stopped systolic pressure 100 mmHg for therapy, targeting systolic
independently of fluid administration. However, in pressure ≥ 100 mmHg with the vasopressor infusion, and

Table 2 Dose of phenylephrine administered to 300 women receiving boluses only vs. phenylephrine infusion. Values are
median (IQR [range]).
Bolus Infusion
n = 150 n = 150 p value
Phenylephrine dose by infusion; lg – 384 (250–500 [50–1000])
Total rescue phenylephrine bolus; lg 349 (100–500 [0–1600]) 100 (0–300 [0–900]) < 0.01
Total phenylephrine dose; lg 349 (100–500 [0–1600]) 589 (350–800 [50–1800]) < 0.01

Table 3 Maternal and neonatal outcomes of 300 women receiving phenylephrine boluses only vs. infusion. Values are mean
(SD), number (proportion) or median (IQR [range]).
Bolus Infusion
n = 150 n = 150 p value
Highest systolic arterial pressure; mmHg 128.1 (15.0) 130.2 (16.6) 0.24
Lowest systolic arterial pressure; mmHg 89.0 (14.5) 94.6 (16.7) < 0.01
1
Highest heart rate; beats.min 109.3 (16.1) 106.8 (17.3) 0.19
1
Lowest heart rate; beats.min 70.3 (13.9) 69.7 (13.0) 0.71
Total fluid; ml 810.7 (382.6) 767.6 (394.4) 0.35
Nausea 12 (8%) 11 (7.3%) 0.83
Vomiting 8 (5.3%) 2 (1.3%) 0.05
Dizziness 17 (11.2%) 10 (6.7%) 0.16
Palpitations 1 (0.7%) 2 (1.3%) 0.56
Headache 1 (0.7%) 1 (0.7%) 1.00
Apgar score at 1 min 9 (8–9 [5–10]) 9 (8–9 [5–10]) 0.79
Apgar score at 5 min 9 (9–10 [5–10]) 9 (9–10 [5–10]) 0.87

4 © 2019 Association of Anaesthetists

Buthelezi et al. | Phenylephrine and fluid co-administration during caesarean section
using rescue vasopressor boluses for pressure < 90 mmHg.
We believe that this is a simpler method for a junior
anaesthetist to follow, and is easier to standardise in
resource-limited settings [5]. Surveys have shown that
specialist anaesthetists in a resource-rich setting also prefer
absolute haemodynamic targets [16]. Additionally, baseline
systolic pressure is often measured in the operating theatre
in our setting, potentially falsely elevating blood pressure
readings due to maternal anxiety before spinal anaesthesia.
Our pragmatic study design, using the administration
of phenylephrine as part of a fluid co-load technique, might
cause hypertension and bradycardia. While the average
systolic pressure in the two groups was not different, it is
noteworthy that the six highest systolic readings were all in
the phenylephrine infusion group. Earlier work using a high
phenylephrine infusion rate of 100 lg.min 1
revealed a
higher incidence of bradycardia, but lower absolute systolic
pressure measurements with no systolic pressure readings
> 150 mmHg, compared with our study; this was despite
using more phenylephrine per patient [17, 18]. These
studies, however, were conducted using high initial infusion
rates, followed by titration and cessation of the infusion if
systolic pressure increased to > 120% baseline systolic
pressure. This methodology may have prevented isolated
high readings, since it allowed for continuation of fluid
therapy while stopping the phenylephrine infusion. In our
study, it is possible that anaesthetists were reluctant to stop
fluid therapy during caesarean section, and thus continued
the phenylephrine co-administration in the face of high
systolic pressure readings. Although there were no clinical
sequelae of these isolated readings, it remains a limitation
of this technique. Limiting the administration of
phenylephrine infusion to women at higher risk of
hypotension might offer a solution, if the method of
identification of such women was reliable and easily applied
[19].
Our study does have some limitations. It is a single-
centre trial, and thus generalisability is limited. However,
Edendale Hospital, a South African regional referral centre
with a high case-load, is likely to be typical of many
resource-limited settings. We did not blind anaesthetists to
the group allocation. This was due to resource constraints,
and the potential need to discontinue phenylephrine
infusions in the event of hypertensive responses. We did not
use a pressure bag to assist with coloading; consequently
the average volume of fluid administered by the time of
1
delivery of the neonate was approximately 10 ml.kg . The
administration of a higher volume of fluid might have
produced different results. We also found a higher
incidence of hypotension than expected, possibly since we
© 2019 Association of Anaesthetists
Anaesthesia 2019
studied only women having elective caesarean section [3,
20]. We postulate that women undergoing emergency
caesarean section are often in labour, may have a higher
baseline concentration of circulating endogenous
catecholamines and raised cardiac output, and thus be less
likely to develop spinal hypotension. Women scheduled for
elective caesarean section are also often exposed to a
prolonged fasting period in limited-resource environments,
which may also contribute to a higher incidence of
hypotension.
This study has shown that a simple prophylactic
phenylephrine infusion, administered as part of a fluid co-
load, is associated with less spinal hypotension than the
administration of intermittent boluses of i.v. phenylephrine.
Better results may be achieved by adjusting the rate of fluid
administration, or by altering the dose of phenylephrine
administered. We have shown that this method appears
safe and effective in the hands of junior anaesthetic staff in a
resource-limited setting, and should be recommended in
settings where syringe drivers are unavailable. Further
research is required to clarify the effectiveness of this
method during emergency caesarean section.
Acknowledgements
Ethics approval was obtained from the Biomedical Research
Ethics Committee of the University of KwaZulu-Natal
(BE616/17), the KwaZulu-Natal Department of Health Ethics
Review Committee (HRKM Ref: 051/18) and Edendale
Hospital management. The trial was registered
retrospectively at ClinicalTrials.gov (NCT04005664). No
external funding or competing interests declared.
References
1. Balki M, Carvalho JCA. Intraoperative nausea and vomiting
during cesarean section under regional anesthesia.
International Journal of Obstetric Anesthesia 2005; 14: 230–41.
2. Pattinson RC. Saving Mothers 2011–2013: the Sixth Report of
the National Committee for Confidential Enquiries into Maternal
Deaths in South Africa. Pretoria: Government Printer, 2014.
3. Kinsella SM, Carvalho B, Dyer RA, et al. International consensus
statement on the management of hypotension with
vasopressors during caesarean section under spinal
anaesthesia. Anaesthesia 2018; 73: 71–92.
4. National Committee on Confidential Enquiries into Maternal
Deaths. Saving Mothers 2014–2016: seventh Triennial Report
on Confidential Enquiries into Maternal Deaths in South Africa:
executive Summary. Pretoria: National Department of Health,
2018.
5. Zwane SF, Bishop DG, Rodseth RN. Hypotension during spinal
anaesthesia for Caesarean section in a resource-limited setting:
towards a consensus definition. Southern African Journal of
Anaesthesia and Analgesia 2019; 25: 28–42.
6. Kl€ohr S, Roth R, Hofmann T, Rossaint R, Heesen M. Definitions of
hypotension after spinal anaesthesia for caesarean section:
literature search and application to parturients. Acta
Anaesthesiologica Scandinavica 2010; 54: 909–21.
5
Anaesthesia 2019 Buthelezi et al. | Phenylephrine and fluid co-administration during caesarean section
7. Reed A, Mumba JM, Dyer RA. spotlight on obstetric anesthesia
in the developing world: finally getting the attention it deserves.
Anesthesia and Analgesia 2015; 120: 1179–81.
8. Bishop DG, Rodseth RN, Dyer RA. Recipes for obstetric spinal
hypotension: the clinical context counts. South African Medical
Journal 2016; 106: 861–4.
9. Bishop DG, Cairns C, Grobbelaar M, Rodseth RN. Prophylactic
phenylephrine infusions to reduce severe spinal anesthesia
hypotension during cesarean delivery in a resource-
constrained environment. Anesthesia and Analgesia 2017;
125: 904–6.
10. Bishop DG, Cairns C, Grobbelaar M, Rodseth RN. Heart rate
variability as a predictor of hypotension following spinal for
elective caesarean section: a prospective observational study.
Anaesthesia 2017; 72: 603–8.
11. Bishop D, Dyer RA, Maswime S, et al. Maternal and neonatal
outcomes after caesarean delivery in the African Surgical
Outcomes Study: a 7-day prospective observational cohort
study. Lancet Global Health 2019; 7: e513–22.
12. Sobhy S, Arroyo-Manzano D, Murugesu N, et al. Maternal and
perinatal mortality and complications associated with
caesarean section in low-income and middle-income
countries: a systematic review and meta-analysis. Lancet 2019;
393: 1973–82.
13. Dyer RA, Reed AR, van Dyk D, et al. Hemodynamic effects of
ephedrine, phenylephrine, and the coadministration of
phenylephrine with oxytocin during spinal anesthesia for
elective cesarean delivery. Anesthesiology 2009; 111: 753–65.
14. Allen TK, George RB, White WD, Muir HA, Habib AS. A double-
blind, placebo-controlled trial of four fixed rate infusion
6 © 2019 Association of Anaesthetists
regimens of phenylephrine for hemodynamic support during
spinal anesthesia for cesarean delivery. Anesthesia and
Analgesia 2010; 111: 1221–9.
15. Heesen M, Kl€ ohr S, Rossaint R, Straube S. Prophylactic
phenylephrine for caesarean section under spinal anaesthesia:
systematic review and meta-analysis. Anaesthesia 2014; 69:
143–65.
16. Burns SM, Cowan CM, Wilkes RG. Prevention and management
of hypotension during spinal anaesthesia for elective
Caesarean section: a survey of practice. Anaesthesia 2001; 56:
794–8.
17. Ngan Kee WD, Khaw KS, Ng FF. Comparison of phenylephrine
infusion regimens for maintaining maternal blood pressure
during spinal anaesthesia for Caesarean section. British Journal
of Anaesthesia 2004; 92: 469–74.
18. Ngan Kee WD, Khaw KS, Ng FF. Prevention of
hypotension during spinal anesthesia for cesarean delivery:
an effective technique using combination phenylephrine
infusion and crystalloid cohydration. Anesthesiology 2005;
103: 744–50.
19. Bishop DG, Cairns C, Grobbelaar M, Rodseth RN. Obstetric
spinal hypotension: preoperative risk factors and the
development of a preliminary risk score - the PRAM score.
South African Medical Journal 2017; 107: 1127–31.
20. Martınez NA, Echevarrıa MM, G omez RP, Merino GS, Caba
BF, Rodrıguez RR. Multivariate study of risk factors for
arterial hypotension in pregnant patients at term
undergoing Caesarean section under subarachnoid
anesthesia. Revista Espa~nola de Anestesiologıa y
Reanimaci on 2000; 47: 189–93.
Anaesthesia 2019 doi:10.1111/anae.14957

Original Article

A randomised controlled trial of shoulder block vs.

interscalene brachial plexus block for ventilatory function
after shoulder arthroscopy
P. Rhyner,1 K. Kirkham,2 C. Hirotsu,3 A. Farron4 and E. Albrecht5

1 Fellow, 5 Program Director, Department of Anaesthesia, 3 Research Associate, Center for Investigation and Research in
Sleep, 4 Head, Department of Orthopaedic Surgery, Lausanne University Hospital, University of Lausanne, Switzerland
2 Associate Professor, Department of Anaesthesia, Toronto University Western Hospital, Toronto, ON, Canada

Summary
The shoulder block may impair ventilatory function and diaphragmatic movement less than the interscalene
brachial plexus block. We randomly allocated 30 adults who underwent shoulder arthroscopy under general
anaesthesia to ultrasound-guided shoulder block or interscalene block with 20 ml bupivacaine 0.5%. The
primary outcome, rate of ultrasound-measured hemidiaphragmatic excursion < 25% of baseline 30 min after
blockade, was reduced from 12/15 with brachial plexus block to 2/15 with shoulder block, a difference (95%CI)
of 67% (40–93%), p < 0.001. The mean (SD) numeric rating scale pain scores at rest after shoulder block were
higher than after interscalene block at two postoperative hours, 1.4 (1.2) vs. 0.3 (0.7), p = 0.02, but lower at 24
postoperative hours, 1.3 (1.3) vs. 3.4 (2.3), p = 0.008. Mean (SD) pain scores on movement in the shoulder and
interscalene blocks were similar, with respective values of 1.9 (1.9) vs. 0.7 (1.2), p = 0.13 at two postoperative
hours and 3.7 (2.3) vs. 5.3 (2.5), p = 0.41, at 24 postoperative hours.

.................................................................................................................................................................
Correspondence to: E. Albrecht
Email: eric.albrecht@chuv.ch
Accepted: 18 November 2019
Keywords: analgesia; peripheral nerve block; postoperative pain; shoulder arthroscopy
Twitter: @DrEAlbrecht, @DrKyleKirkham

Introduction ventilatory function and diaphragmatic movement, but this

The interscalene brachial plexus block provides effective
analgesia after shoulder surgery but impairs diaphragmatic
function [1]. The shoulder block is an alternative, which
blocks the suprascapular and axillary nerves. These two
nerves innervate most of the shoulder, with additional minor
contributions from the subscapular and lateral pectoral
nerves [2, 3].
A block of the suprascapular nerve can either be
performed posteriorly, at the level of the suprascapular
fossa, or anteriorly, below the belly of the omohyoid muscle
in the supraclavicular region [4]. The anterior approach is
potentially more reliable as the superficial nerve can be
easily identified on ultrasound imaging [4]. It has been
suggested that the shoulder block has little effect on
has not been tested [5].
We aimed to test whether ventilatory function and
hemidiaphragmatic movement is different after shoulder
block compared with interscalene block.
Methods
The Ethics Committee of Lausanne University Hospital
approved this prospectively registered trial, which we
report as standard [6]. We studied adults, ASA physical
status 1–3, scheduled for shoulder arthroscopy from
January 2017 to April 2019. We did not study patients with
upper limb neurological deficit; history of neck surgery or
radiotherapy; moderate or severe pulmonary disease;
chest deformity (pectus carinatum or pectus excavatum); a

© 2019 Association of Anaesthetists 1

Anaesthesia 2019 Rhyner et al. | Ventilatory function after interscalene block vs. shoulder block
contra-indication to peripheral nerve block, for instance
study drug allergy, coagulopathy or local infection. We did
not study pregnant women or patients with chronic opioid
use. Participants provided written informed consent. We
allocated participants on the day of surgery equally to
shoulder block or interscalene brachial plexus block, using
a computer-generated randomisation table in blocks of 10.
Assignments were concealed in sealed opaque envelopes.
One of the authors (EA) performed or directly
supervised blocks in a dedicated room, during which we
monitored ECG, pulse oximetry and non-invasive blood
pressure. We injected intravenous (i.v.) midazolam 1–4 mg
for anxiolysis and administered supplemental oxygen as
needed. We prepared the skin with chlorhexidine 2% in
isopropyl alcohol 70%. We used a high-frequency linear
array ultrasound transducer (13–6 MHz, BK Medicalâ Flex
Focus 400; BK Medical Holding Company Inc., MA, USA).
We performed the interscalene block with the heads of
reclined participants turned 45° away. We identified the
roots of C5–C7 in the interscalene region, as described
previously [7]. We infiltrated the skin with 1–3 ml lidocaine
1%. We advanced a 22-gauge 50-mm insulated block
needle (SonoPlex Stim cannula, Pajunkâ, Geisingen,
Germany) in-plane through the middle scalene muscle and
injected 20 ml bupivacaine 0.5% with adrenaline 1:200,000
between the C5 and C6 nerve roots.
Participants allocated to the shoulder group were
similarly positioned and the skin prepared for the
suprascapular nerve block [8]. We injected 10 ml
bupivacaine 0.5% with adrenaline 1:200,000 around the
nerve after it emerged from the superior trunk of C5, below
the omohyoid muscle towards the trapezius muscle. We
performed the axillary nerve block with participants
positioned and skin prepared as previously described [9].
We injected 10 ml bupivacaine 0.5% with adrenaline
1:200,000 cephalad to the posterior humeral circumflex
artery, the identity of which was confirmed by colour
Doppler.
We induced anaesthesia intravenously with propofol
2–4 mg, supplemented with fentanyl 1–2 lg.kg 1
facilitated tracheal intubation with rocuronium 0.6 mg.kg
We maintained anaesthesia with sevoflurane 1.6–2.4%.
Ventilation was adjusted to maintain an end-tidal PCO2 of
4.6–5.3 kPa mmHg and we injected 25–50 lg i.v. fentanyl if
blood pressure or heart rate exceeded 115%
baseline values. We injected intravenously magnesium
1 1
sulphate 50 mg.kg , dexamethasone 0.15 mg.kg
ondansetron 4 mg. We antagonised neuromuscular
blockade with neostigmine 50 lg.kg 1
and glycopyrrolate
1
5–10 lg.kg at the end of surgery. After surgery, we gave
2 © 2019 Association of Anaesthetists
participants with numeric rating scale pain scores ≥ 4 i.v.
morphine 1–2 mg every 10 min, or on request. We gave
oral paracetamol 1 g every 6 h and ibuprofen 400 mg every
8 h, supplemented with patient-controlled i.v. morphine.
We treated nausea or vomiting with i.v. ondansetron 4 mg
and metoclopramide 10 mg. All participants remained in
hospital for at least 24 h. We contacted participants
seven days after surgery and recorded complications such
as haematoma, infection, persistent paraesthesia or
weakness in the upper limb.
After regional blockade, a research assistant used a
blunt-tip needle to categorise sensation in dermatomes C5
and C6 as absent, decreased or normal and categorised
arm abduction and forearm flexion as too weak to overcome
gravity, reduced or normal. We defined successful block as
absent sensation and movement within 30 min of the block.
An author (PR) or a trained research assistant used a low-
frequency curvilinear transducer (5–2 MHz, BK Medicalâ
Flex Focus 400; BK Medical Holding Company Inc., MA,
USA) to measure hemidiaphragmatic excursion before and
30 min after regional blockade and 24 h later [1, 10, 11].
Briefly, participants were examined in the supine position
and the hemidiaphragm was identified as a hyperechoic
line with breathing-related movements using the liver or
spleen as an acoustic window. The hemidiaphragmatic
excursion was measured by real-time M-mode
ultrasonography as the difference in distance between
resting expiration and deep inspiration (see also Supporting
Information, Figure S1). Peak expiratory flow, forced vital
capacity and forced expiratory flow in 1 s were measured as
the best of three measures at the same times using a bed-
side spirometer (EasyOneTM Spirometer; ndd Medical
Technologies, Andover, UK), with the participant seated
upright.
We defined the primary outcome as a reduction in
diaphragmatic excursion to < 25% of the baseline value
30 min after blockade [1, 10]. The secondary outcomes
were: changes in the other ventilatory variables; time from
block to first dose of i.v. morphine; pain at rest and on
and movement, measured with a numeric rating scale (1–10) at 2
1
. and 24 postoperative hours; cumulative i.v. morphine
consumption at 2 and 24 postoperative hours; time to first
opioid request; rates of postoperative nausea and vomiting
at 2 and 24 postoperative hours; and satisfaction with
of overall anaesthetic management, measured with a numeric
rating scale (1–10).
and We calculated that we would need to recruit 12
participants to each group to have a 90% power to detect a
reduction in rate of hemidiaphragmatic paresis from 8/12
after interscalene block to 0/12 after shoulder block,
Rhyner et al. | Ventilatory function after interscalene block vs. shoulder block
assuming an alpha error of 0.01. We planned to recruit 30
participants in total, assuming a 20% rate of protocol
violation or drop-out. Data were analysed on an intention-
to-treat basis. Independent continuous parametric and non-
continuous data were compared using the Student’s t-test,
and Mann–Whitney U-test, when appropriate, except time
from block to first dose of i.v. morphine that was analysed
with a Kaplan–Meier survival curve. Independent
categorical data were compared with Fisher’s exact test.
Both categorical and continuous repeated measurements
were analysed using generalised estimating equations
considering the time, group and interaction between time
and group effects [12, 13]. The distribution of each model
was chosen based on the nature of the variable and the
lowest Quasi-likelihood under Independence Model
Criterion. The differences in respiratory and pain-related
Figure 1 CONSORT flow diagram.
© 2019 Association of Anaesthetists
Anaesthesia 2019
variables between two time-points were analysed with
generalised linear models [12, 13], including as covariate
the respective variable at baseline. Post-hoc Bonferroni
adjustment was used for multiple comparisons when
appropriate. We considered two-tailed p < 0.01 statistically
significant for our primary outcome, and p < 0.05 for the
secondary outcomes. Statistical analysis was performed
using SPSS Statistics for Windows, Version 25.0 (IBM Corp,
Armonk, NY).
Results
Thirty participants completed the protocol (Fig. 1; Table 1).
All blocks were successful. Shoulder block reduced the rate
of hemidiaphragmatic paresis at 30 min from 12/15 (80%)
after interscalene block to 2/15 (13%), p < 0.001, a
difference (95%CI) of 67% (40–93%). Shoulder block caused
3
Anaesthesia 2019 Rhyner et al. | Ventilatory function after interscalene block vs. shoulder block

1 1
Table 1 Characteristics and clinical data for 30 participants l, p < 0.0001; and -0.10 (1.24) l.s vs. 1.79 (1.39) l.s ,
allocated to shoulder block or interscalene block before p = 0.001, respectively (Table 2).
shoulder arthroscopy. Values are mean (SD), median (IQR Hemidiaphragmatic function had returned by 24
[range]) or number.
postoperative hours and no participant had dyspnoea
Block (Table 2 and also see Supporting Information Table S1).
Shoulder Interscalene The median (IQR [range]) time to first opioid request was
n = 15 n = 15
255 (167–750 [150–1655]) min after shoulder block and 655
Sex; male 13 10 (463–902 [165–995]) min after interscalene block, log-rank
Age; y 42.3 (17.8) 36.4 (17.9) p = 0.60. Table 3 shows other secondary outcomes. No
Height; cm 171 (10) 176 (8) participant reported a complication from blockade.
Weight; kg 71.9 (16.3) 78.9 (13.4)
BMI; kg.m 2
24.3 (3.9) 25.4 (3.9) Discussion
ASA 1/2/3 2/12/1 4/11/0 We found that shoulder block reduced the rate of
Duration of surgery; min 70.0 (34.4) 86.0 (22.2) hemidiaphragmatic paresis and ventilatory dysfunction
Surgery when compared with interscalene block.
Capsular stabilisation 11 8 We anticipated that shoulder block would not cause
Acromioclavicular resection 1 1 hemidiaphragmatic paresis. Migration of local anaesthetic
Biceps tenotomy 0 5 from the injection site around the suprascapular nerve to the
Other 3 1 phrenic may be why 14 out of 15 participants had some
reduction in hemidiaphragmatic excursion 30 min after
shoulder block, two of whom had hemidiaphragmatic
paresis [14]. Hemidiaphragmatic impairment might be
less median (IQR [range]) or mean (SD) reduction of reduced by the more difficult posterior approach to the
hemidiaphragmatic excursion, forced vital capacity, forced suprascapular nerve [4]. All participants had reduced
expiratory volume (in 1 s) and peak expiratory flow at hemidiaphragmatic excursion 30 min after interscalene
30 min than interscalene block: 0.20 ( 0.10 to 0.40 [ 1.0 to block, the rate of which might be reduced by injecting 5 ml
4.9]) cm vs. 5.4 (4.1–6.0 [0.9–6.7]) cm, p < 0.0001; 0.11 of local anaesthetic, rather than 20 ml, or by injecting above
(0.36) l vs. 1.04 (0.5) l, p < 0.0001; 0.03 (0.32) l vs. 0.87 (0.39) the clavicle or extrafascially [10, 15–17].

Table 2 Ventilatory function in 30 participants before and after shoulder block or interscalene block and after shoulder
arthroscopy. Values are median (IQR [range]).
Block
Shoulder Interscalene
Ventilatory function (n = 15) (n = 15) p value
Before block
Diaphragmatic excursion (cm) 5.0 (4.3–5.6 [3.3–7.0]) 5.8 (5.3–6.5 [4.1–6.9]) 0.48
FVC; l 4.1 (3.5–5.1 [1.6–5.6]) 4.7 (3.6–5.2 [2.2–6.2]) 1.00
FEV1; l 3.3 (2.4–3.8 [0.9-4.5]) 4.0 (3.1–4.3 [1.8–5.4]) 0.42
1
PEF; l.s 7.5 (5.4-8.5 [2.7-9.6]) 8.6 (7.7–9.4 [3.7–12.5]) 0.35
After block 30 min
Diaphragmatic excursion (cm) 4.5 (3.4–5.0 [0.0–7.2]) 0.5 (0.0–1.5 [0.0–4.5]) < 0.0001
FVC; l 4.2 (3.3–4.9 [1.7–5.6]) 3.5 (2.8–4.3 [1.3–5.1]) 1.00
FEV1; l 3.3 (2.2–3.7 [1.1–4.8]) 2.9 (2.4–3.6 [1.1–4.7]) 1.00
1
PEF; l.s 7.7 (5.3–8.6 [2.3–10.3]) 7.3 (5.6–8.5 [2.8–9.6]) 1.00
After block and arthroscopy 24 h
Diaphragmatic excursion (cm) 4.6 (3.4–5.5 [1.5–7.1]) 4.1 (3.0–5.4 [2.0–6.0]) 1.00
FVC; l 3.9 (3.4–4.3 [1.4–5.5]) 4.2 (3.5–4.7 [1.6–6.4]) 1.00
FEV1; l 3.1 (2.1–3.4 [1.1–4.5]) 3.6 (2.8–4.0 [1.1–4.3]) 1.00
1
PEF; l.s 7.5 (5.7–8.2 [2.7–9.3]) 8.0 (7.1–9.5 [3.5–10.1]) 1.00
FVC, forced vital capacity; FEV1, forced expiratory volume in 1 s; PEF, peak expiratory flow.

4 © 2019 Association of Anaesthetists

Rhyner et al. | Ventilatory function after interscalene block vs. shoulder block Anaesthesia 2019

Table 3 Pain, analgesia and nausea or vomiting after be considered hypothesis generating. We do not know
shoulder arthroscopy in 30 participants allocated to pre- whether the results of blockade by one practitioner can
operative shoulder block or interscalene block. Values are be inferred to other practitioners. We do not know
mean (SD) or number.
whether our results in participants with body mass
Block indices of around 25 can be generalised to an obese
Shoulder Interscalene population, in which blockade would be more difficult.
n = 15 n = 15 p value
We do not know whether results would be similar for
At 2 postoperative hours shoulder procedures more painful than arthroscopy, for
Pain; NRS (0–10) instance shoulder arthroplasty and rotator cuff repair, as
At rest 1.4 (1.2) 0.3 (0.7) 0.02 the shoulder block does not numb the subscapular and
On movement 1.9 (1.9) 0.7 (1.2) 0.13 lateral pectoral nerves.
Cumulative 3.5 (4.2) 1.7 (6.4) 1.00 In conclusion, the ultrasound-guided shoulder block
intravenous
caused less hemidiaphragmatic paresis and impaired
morphine; mg
ventilation less than interscalene block. The different
Nausea or vomiting 3 2 1.00
pattern of postoperative pain may reflect rebound pain after
At 24 postoperative hours
the interscalene block stopped working.
Pain; NRS (0–10)
At rest 1.3 (1.3) 3.4 (2.3) 0.008
On movement 3.7 (2.3) 5.3 (2.5) 0.41 Acknowledgements
Cumulative 10.6 (15.4) 13.2 (15.2) 1.00
EA has received grants from the Swiss Academy for
intravenous Anaesthesia Research (SACAR), Lausanne, Switzerland
morphine; mg (50,000 CHF; no grant number attributed), from B. Braun
Nausea or vomiting 2 2 1.00 Medical AG (56,100 CHF; no grant number attributed) and
Satisfaction; 9.1 (0.9) 9.3 (1.1) 0.33 from the Swiss National Science Foundation to support his
NRS (0–10)
clinical research (353,408 CHF; grant number:
NRS, numeric rating scale. 32003B_169974/1). EA has also received an honorarium
from B. Braun Medical AG. The trial was registered at
Participants had less rest pain at two postoperative Clinicaltrials.com (NCT02916342). No other external
hours after interscalene block than after shoulder block, but funding or competing interests declared.
more pain 22 h later, although morphine consumption was
similar. Increased pain after a regional block wears off – References
1. Albrecht E, Bathory I, Fournier N, Jacot-Guillarmod A, Farron A,
‘rebound’ pain – is common, for which most evidence
Brull R. Reduced hemidiaphragmatic paresis with extrafascial
comes from retrospective case series [18–21]. A similar compared with conventional intrafascial tip placement for
pattern of delayed but increased pain was described in a continuous interscalene brachial plexus block: a randomized,
controlled, double-blind trial. British Journal of Anaesthesia
randomised controlled trial of brachial plexus blockade vs. 2017; 118: 586–92.
general anaesthesia for fixation of distal radius fracture [22]. 2. Dhir S, Sondekoppam RV, Sharma R, Ganapathy S, Athwal
In a post-hoc analysis we calculated the mean (SD) GS. A Comparison of combined suprascapular and
axillary nerve blocks to interscalene nerve block for
difference in pain at 24–2 h, at rest and on movement: 3.1 analgesia in arthroscopic shoulder surgery: an equivalence
(2.4) and 4.6 (2.5) after interscalene block vs. 0.1 (1.8) and study. Regional Anesthesia and Pain Medicine 2016; 41:
564–71.
1.8 (2.4) after shoulder block, p = 0.003 and p = 0.02,
3. Tran J, Peng PWH, Agur AMR. Anatomical study of the
respectively [19]. innervation of glenohumeral and acromioclavicular joint
Our study was vulnerable to performance bias and capsules: implications for image-guided intervention. Regional
Anesthesia and Pain Medicine 2019; 44: 452–8.
assessment bias as participants and clinical personnel
4. Siegenthaler A, Moriggl B, Mlekusch S, et al. Ultrasound-
knew group allocation. Our study was small although guided suprascapular nerve block, description of a novel
consistent with our sample size calculation and we supraclavicular approach. Regional Anesthesia and Pain
Medicine 2012; 37: 325–8.
believe it would have been inappropriate to recruit more 5. Tran DQ, Elgueta MF, Aliste J, Finlayson RJ. Diaphragm-sparing
participants [23]. The fragility index for our study is 4, nerve blocks for shoulder surgery. Regional Anesthesia and
above the 75th percentile of a recent systematic review Pain Medicine 2017; 42: 32–8.
6. Schulz KF, Altman DG, Moher D. CONSORT 2010 statement:
of 43 trials [24]. We acknowledge that we did not design updated guidelines for reporting parallel group randomised
the trial to test the secondary outcomes, which should trials. PLoS Medicine 2010; 7: e1000251.

© 2019 Association of Anaesthetists 5

Anaesthesia 2019 Rhyner et al. | Ventilatory function after interscalene block vs. shoulder block
7. Albrecht E, Kirkham KR, Taffe P, et al. The maximum effective
needle-to-nerve distance for ultrasound-guided interscalene
block: an exploratory study. Regional Anesthesia and Pain
Medicine 2014; 39: 56–60.
8. Rothe C, Steen-Hansen C, Lund J, Jenstrup MT, Lange KH.
Ultrasound-guided block of the suprascapular nerve – a
volunteer study of a new proximal approach. Acta
Anaesthesiologica Scandinavica 2014; 58: 1228–32.
9. Rothe C, Asghar S, Andersen HL, Christensen JK, Lange KH.
Ultrasound-guided block of the axillary nerve: a volunteer study
of a new method. Acta Anaesthesiologica Scandinavica 2011;
55: 565–70.
10. Palhais N, Brull R, Kern C, et al. Extrafascial injection for
interscalene brachial plexus block reduces respiratory
complications compared with a conventional intrafascial
injection: a randomized, controlled, double-blind trial. British
Journal of Anaesthesia 2016; 116: 531–7.
11. Scarlata S, Mancini D, Laudisio A, Benigni A, Antonelli Incalzi R.
Reproducibility and clinical correlates of supine diaphragmatic
motion measured by M-Mode ultrasonography in healthy
volunteers. Respiration 2018; 96: 259–66.
12. Verbeke G, Fieuws S, Molenberghs G, Davidian M. The analysis
of multivariate longitudinal data: a review. Statistical Methods
in Medical Research 2017; 26: 112.
13. Zeger SL, Liang KY, Albert PS. Models for longitudinal data: a
generalized estimating equation approach. Biometrics 1988;
44: 1049–60.
14. Sehmbi H, Johnson M, Dhir S. Ultrasound-guided
subomohyoid suprascapular nerve block and phrenic nerve
involvement: a cadaveric dye study. Regional Anesthesia and
Pain Medicine 2019; 44: 561–4.
15. Riazi S, Carmichael N, Awad I, Holtby RM, McCartney CJ. Effect
of local anaesthetic volume (20 vs 5 ml) on the efficacy and
respiratory consequences of ultrasound-guided interscalene
brachial plexus block. British Journal of Anaesthesia 2008; 101:
549–56.
16. El-Boghdadly K, Chin KJ, Chan VWS. Phrenic nerve palsy and
regional anesthesia for shoulder surgery: anatomical,
physiologic, and clinical considerations. Anesthesiology 2017;
127: 173–91.
17. Kang RA, Chung YH, Ko JS, Yang MK, Choi DH. Reduced
hemidiaphragmatic paresis with a “Corner Pocket” technique
6 © 2019 Association of Anaesthetists
for supraclavicular brachial plexus block: single-center,
observer-blinded, randomized controlled trial. Regional
Anesthesia and Pain Medicine 2018; 43: 720–4.
18. Lavand’homme P. Rebound pain after regional anesthesia in
the ambulatory patient. Current Opinion in Anaesthesiology
2018; 31: 679–84.
19. Williams BA, Bottegal MT, Kentor ML, Irrgang JJ, Williams JP.
Rebound pain scores as a function of femoral nerve block
duration after anterior cruciate ligament reconstruction:
retrospective analysis of a prospective, randomized clinical
trial. Regional Anesthesia and Pain Medicine 2007; 32: 186–92.
20. Sunderland S, Yarnold CH, Head SJ, et al. Regional versus
general anesthesia and the incidence of unplanned health care
resource utilization for postoperative pain after wrist fracture
surgery: results from a retrospective quality improvement
project. Regional Anesthesia and Pain Medicine 2016; 41:
22–7.
21. Sort R, Brorson S, G€ ogenur I, Nielsen JK, Møller AM. Rebound
pain following peripheral nerve block anaesthesia in acute
ankle fracture surgery: an exploratory pilot study. Acta
Anaesthesiologica Scandinavica 2019; 63: 396–402.
22. Galos DK, Taormina DP, Crespo A, et al. Does brachial plexus
blockade result in improved pain scores after distal radius
fracture fixation? A randomized trial. Clinical Orthopaedics and
Related Research 2016; 474: 1247–54.
23. Gibbs NM, Weightman WM. Beyond effect size: consideration
of the minimum effect size of interest in anesthesia trials.
Anesthesia and Analgesia 2012; 114: 471–5.
24. Bertaggia L, Baiardo RM, Lembo R, et al. The Fragility Index in
peri-operative randomised trials that reported significant
mortality effects in adults. Anaesthesia 2019; 74: 1057–60.
Supporting Information
Additional supporting information may be found online in
the Supporting Information section at the end of the article.
Figure S1 An example of the measurement of
hemidiaphragmatic excursion.
Table S1 Individual participant data for diaphragmatic
excursion before and after block.
Anaesthesia 2020 doi:10.1111/anae.14978

Original Article

Hemidiaphragmatic paralysis following ultrasound-guided

anterior vs. posterior suprascapular nerve block: a
double-blind, randomised control trial

,1 M. Pommier,1 P. Laumonerie,2 A. Ferrier,1 R. Menut,1 L. Bosch,1 V. Balech,1
F. Ferre
N. Bonnevialle2,3 and V. Minville1,4

1 Consultant, 4 Professor, D

epartement d’Anesth
esie R

eanimation, 2 Consultant, 3 Professor, D

epartement
d’Orthop
edie Traumatologie, Centre Hospitalier Universitaire de Toulouse Purpan, H^opital Riquet, Universit

e Toulouse
3-Paul Sabatier, Toulouse, France

Summary
Interscalene brachial plexus block provides analgesia for shoulder surgery but is associated with
hemidiaphragmatic paralysis. Before considering a combined suprascapular and axillary nerve block as an
alternative to interscalene brachial plexus block, evaluation of the incidence of diaphragmatic dysfunction
according to the approach to the suprascapular nerve is necessary. We randomly allocated 84 patients
undergoing arthroscopic shoulder surgery to an anterior or a posterior approach to the suprascapular nerve
block combined with an axillary nerve block using 10 ml ropivacaine 0.375% for each nerve. The primary
outcome was the incidence of hemidiaphragmatic paralysis diagnosed by ultrasound. Secondary outcomes
included: characterisation of the hemidiaphragmatic paralysis over time; numeric rating scale pain scores; oral
morphine equivalent consumption; and patient satisfaction. The incidence of hemidiaphragmatic paralysis was
40% (n = 17) vs. 2% (n = 1) in the anterior and posterior groups, respectively (p < 0.001). In one third of patients
with hemidiaphragmatic paralysis, it persisted beyond the eighth hour. The median (interquartile range [range])
oral morphine equivalent consumption was significantly higher in the posterior approach when compared with
the anterior approach, whether in the recovery area (20 [5–31 (0–60)] mg vs. 7.5 [0–14 (0–52)] mg, respectively;
p = 0.004) or during the first 24 h (82 [61–127 (12–360) mg] vs. 58 [30–86 (0–160)] mg, respectively; p = 0.01).
Patient satisfaction was comparable between groups (p = 0.6). Compared with the anterior approach,
diaphragmatic function is best preserved with the posterior needle approach to the suprascapular nerve block.

.................................................................................................................................................................
Correspondence to: F. Ferr

e
Email: fabriceferre31@gmail.com
Accepted: 7 December 2019
Keywords: diaphragmatic paralysis; interscalene brachial plexus block; phrenic; shoulder arthroscopy; suprascapular
nerve block
This editorial accompanies an article by Mariano and El-Boghdadly, Anaesthesia 2020; https://doi.org/10.1111/anae.
14992.

Introduction 5]. However, patients frequently report shortness of

Interscalene brachial plexus block is considered to be breath caused by accidental local anaesthetic spread to
the gold standard for postoperative pain management the phrenic nerve [1, 6], which results in
after shoulder surgery [1–4]. When patients receive an hemidiaphragmatic paralysis [7–11]. This symptom is
interscalene brachial plexus block in this setting, they potentially debilitating in patients with poor respiratory
use less opioids postoperatively and recover faster [1, function [12].

© 2020 Association of Anaesthetists 1

Anaesthesia 2020
Alternative regional analgesic techniques
interscalene brachial plexus block target the peripheral
nerves that innervate the shoulder and may have a better
side-effect profile [13]. The suprascapular and axillary
nerves provide sensory innervation to most of the
glenohumeral joint [14–16]. Therefore, a combined
suprascapular and axillary nerve block is a valid alternative
to interscalene brachial plexus block [4] with less impact on
respiratory function [17–22] and hand-grip strength [23].
The anatomy of the suprascapular nerve makes two
ultrasound-guided approaches possible: an anterior
approach in the supraclavicular region; and a posterior
approach in the suprascapular notch [13, 24–26]. The
anterior approach provides better visualisation of the
suprascapular nerve and may be easier to perform [13].
However, the posterior approach allows for
anaesthetic injection further away from the interscalene
groove and could reduce the incidence
hemidiaphragmatic paralysis [6].
To our knowledge, no clinical study has evaluated the
influence of the needle approach on the incidence of
hemidiaphragmatic paralysis induced by a suprascapular
nerve block. This evaluation is necessary before considering
a combined suprascapular and axillary nerve block as an
alternative to interscalene brachial plexus block, especially
in patients at risk for pulmonary complications.
We aimed to evaluate the influence of
suprascapular nerve block approach (anterior vs. posterior)
on the incidence of hemidiaphragmatic paralysis
diagnosed by ultrasound after a combined suprascapular
and axillary nerve block for arthroscopic shoulder surgery.
We also aimed to compare the postoperative analgesic
efficacy of these two approaches.
Methods
This interventional, double-blind, randomised controlled
trial was conducted in the University Teaching Hospital of
Purpan from January 2018 to January 2019. Written
informed consent was obtained from all subjects. Approval
was given for this study by the West V Ethics Committee.
All patients > 18 years old scheduled for arthroscopic
shoulder day surgery under general anaesthesia were
eligible. The exclusion criteria were: pre-existing respiratory
failure; the inability to perform diaphragmatic ultrasound;
contra-indications to regional anaesthesia; and pregnancy.
Using computer-generated randomisation
Statistical Software, Release 14, StataCorp LP, College
Station, TX, USA), patients were successively randomly
assigned to ‘anterior’ (suprascapular nerve block with an
anterior approach in combination with an axillary nerve
2 © 2020 Association of Anaesthetists
Ferr

e et al. | Diaphragmatic paralysis and suprascapular block
to block) or ‘posterior’ (suprascapular nerve block with a
posterior approach in combination with an axillary nerve
block) groups. Allocation numbers were sealed in
envelopes and opened successively at the time of inclusion
on the day of surgery by the anaesthetist who performed the
regional anaesthesia. Patients were blinded to their group
assignment. The physician who evaluated the outcome
criteria did not know the results of the randomisation. The
overall anaesthetic management (notably, general
anaesthesia) was performed by the physician responsible
for the patient who was not involved in performing regional
anaesthesia or collecting data, and who was blinded after
assignment to interventions.
For all patients, intravenous (i.v.) access contralateral to
the surgical site was established. Standard vital sign
local monitors were placed and oxygen was delivered through a
facemask. All patients received a combined suprascapular
of and axillary nerve block before induction of general
anaesthesia with an ultrasound-guided technique. Surgery
was performed under general anaesthesia induced with
propofol (2–3 mg.kg
1) and sufentanil (0.2–0.3 lg.kg
1).
For prophylaxis of postoperative nausea and vomiting,
8 mg i.v. dexamethasone was administered to each patient
at the time of induction of anaesthesia, following loss of
consciousness. After placement of a supraglottic airway
device (Ambuâ Aura-iTM, Ambu A/S, Ballerup, Denmark),
the anaesthesia was maintained with 4–8 mg.kg
1.h
1 i.v.
propofol. Patients0 lungs were mechanically ventilated with
a tidal volume of 6 ml.kg
1 ideal bodyweight. At the end of
surgery, patients received 1 g i.v. paracetamol and 100 mg
i.v. ketoprofen in the absence of contra-indications. In the
recovery area, pain was evaluated by an 11-point numerical
rating scale (NRS) where 0 is no pain and 10 is the worst pain
imaginable. Intravenous morphine was administered when
the NRS was > 3, with an initial bolus dose of 0.1 mg.kg
1.
On discharge, all patients received 1 g paracetamol
every 6 h for 7 days; 100 mg ketoprofen every 12 h for
7 days; 30 mg sustained-release morphine sulphate every
12 h for 5 days; and immediate-release morphine sulphate
as required, maximum every 4 h in case of pain, for 5 days.
The anterior approach to the suprascapular nerve block
was performed as described by Siegenthaler et al. [13].
With the patient in the supine position, a high-frequency
linear ultrasound probe (10–15 MHz, X-Porte, Fujifilm
SonositeVR, Inc., Bothell, WA, USA) was applied to the side
(Stata of the neck in a transverse plane to visualise the brachial
plexus. The probe was then moved distally in order to
identify the emergence of the suprascapular nerve from the
superior trunk. The suprascapular nerve was followed as
laterally as possible until it was located in the supraclavicular
Ferr

e et al. | Diaphragmatic paralysis and suprascapular block
fossa under the omohyoid muscle. A 22-G, 80-mm needle
(Pajunk SonoPlex, Geisingen, Germany) was directed in-
plane from the posterior edge of the probe towards the
suprascapular nerve. A perineural injection of 10 ml
ropivacaine 0.375% combined with 2 mg dexamethasone
was administered.
The posterior approach to the suprascapular nerve
block was performed according to the methods described
by Harmon and Hearty [25]. With the patient in a seated
position, a high-frequency linear ultrasound probe (10–
15 MHz, X-Porte, Fujifilm Sonosite, Inc., Bothell, WA, USA)
was placed in the transverse plane on the medial side of
the coracoid process to view the trapezius and the
supraspinatus muscles as well as the floor of the
supraspinous fossa. The transducer was slightly tilted
anteriorly to identify the suprascapular notch. The nerve
was then located under the hyperechoic line of the
transverse ligament of the scapula, close to
suprascapular artery. A 22-G, 80-mm needle (Pajunk
SonoPlex, Geisingen, Germany) was directed in-plane from
the medial edge of the probe under the transverse
ligament. A perineural injection of 10 ml ropivacaine
0.375% combined with 2 mg dexamethasone
administered.
The axillary nerve block was administered according to
the technique described by Rothe et al. [27]. The patient
was placed in a seating position. The ultrasound probe was
placed in a sagittal plane, parallel to the long axis of the
postero-lateral humerus to enable identification of the neck
of the humerus. The axillary nerve was then identified
beside the posterior humeral circumflex artery in the
quadrilateral space of Velpeau delineated by the teres
minor muscle; the neck of the humerus; the triceps muscle;
and the deltoid muscle. The needle was introduced in-plane
through the deltoid muscle and a perineural injection of
10 ml ropivacaine 0.375% combined with
dexamethasone was administered.
The function of the diaphragm was evaluated by
ultrasound using previously described techniques [28, 29].
Patients were placed in a semi-recumbent position. A 3.5–
5 MHz curvilinear probe (X-Porte, Fujifilm Sonosite Inc.,
Bothell, WA, USA) was placed under the costal margin
between the anterior axillary and the mid-clavicular lines
with a medial, dorsal and cephalic orientation of the probe
in order to view the posterior third of the hemidiaphragm.
The liver for the right side or the spleen for the left side were
used as acoustic windows. The ultrasound was set to motion
mode, enabling the diaphragm to appear as a white
hyperechoic line undulating during the respiratory cycle.
The diaphragmatic excursion, defined as a cranio caudal
© 2020 Association of Anaesthetists
Anaesthesia 2020
movement of the diaphragm, was measured in centimetres
during a voluntary ‘sniff test’ (i.e. inspiratory manoeuvre for
which the patient is asked to forcefully inhale nasally in a
sniffing fashion) before regional anaesthesia (basal state),
and then 30 min and 4 h after administration of the block. A
fourth measurement was taken 8 h after the block was
performed if hemidiaphragmatic paralysis persisted at the
fourth hour. At every recording time, the largest of three
measurements of the excursion of the diaphragm was
entered.
The primary outcome was the incidence of
hemidiaphragmatic paralysis defined by a ≥ 25% reduction
in diaphragmatic excursion measured by ultrasound.
Complete diaphragmatic paralysis was defined as a
reduction in diaphragmatic excursion ≥ 75% in comparison
with baseline, or by paradoxical cephalic movement of the
diaphragm. Partial diaphragmatic paralysis was defined as a
the reduction in diaphragmatic excursion of 25–75% [30]. The
characterisation of hemidiaphragmatic paralysis and the
change in diaphragmatic excursion over time were also
examined.
The efficacy of the block was evaluated 30 min and 4 h
was after injection of local anaesthetic. The greatest block
efficacy between these two evaluations was considered and
used in the analysis. The top of the shoulder (axillary nerve)
and the posterior surface of the scapula (suprascapular
nerve) were tested for cold sensitivity. Sensory block was
defined as complete, partial or absent [18]. To evaluate the
intensity of the motor block, the patient was asked to
elevate, abduct and externally rotate the shoulder. Motor
block was defined as complete, partial or absent [17].
Postoperative pain was evaluated by the 11-point NRS and
the total oral morphine equivalent consumption was
assessed in the recovery area and via a telephone call at
24 h.
2 mg Patient satisfaction was evaluated with the modified
Evaluation du V

ecu de l’Anesth

esie Loco-R

egionale (EVAN-
LR) questionnaire [31]. This self-reporting questionnaire was
completed during the systematic telephone call the day
after surgery. It includes 17 items structured in five
dimensions (attention; information; discomfort;
expectation; and pain) evaluated on a 1–5 Likert scale (1,
much less than expected; 2, less than expected; 3, as
expected; 4, more than expected; 5, much more than
expected) and summarised as a global index. The score for
each dimension was linearly transformed to a scale from 0 to
100, where 100 is the best possible level of satisfaction and
0 is the worst. Scores for the items worded negatively were
inverted so that the highest scores indicated a high level of
satisfaction. The global index score was calculated as the
3
Anaesthesia 2020
Figure 1 CONSORT study flow diagram.
mean of the scores for each dimension. Concomitantly,
patients were asked whether they would agree to receive
the same anaesthetic technique for subsequent shoulder
operations, if indicated.
The feasibility of the suprascapular nerve block was
evaluated by the procedure time, that is, the time between
needle-patient contact and end of local anaesthetic
injection. Practitioners were also asked to define the
visualisation of the suprascapular nerve on ultrasound
according to a semi-quantitative scale of 0–2, where 0 is
absent (defined as an invisible nerve structure) and 2 is
good (defined as optimum visibility of the nerve) [32].
No data existed regarding the incidence of
hemidiaphragmatic paralysis after suprascapular nerve
block. According to anatomical considerations, we estimated
the incidence of hemidiaphragmatic paralysis after
suprascapular nerve block with the anterior route at
approximately 30% compared with approximately 1% with
the posterior route. The sample size was calculated
assuming a significant difference in the incidence of
hemidiaphragmatic paralysis between the groups with a
bilateral alpha risk set at 5% and a power set at 90%. Thirty-
eight patients per group were required. To compensate for
subjects lost to follow-up, generally estimated at 10%, we
aimed to include 42 subjects. The normality of the data was
verified with the Shapiro–Wilk test. Continuous variables
between the anterior and posterior groups were compared
with the Mann–Whitney U-test. Categorical variables were
compared with the Chi-squared test or Fisher’s exact test.
The change in diaphragmatic excursion over time was
4 © 2020 Association of Anaesthetists
Ferr

e et al. | Diaphragmatic paralysis and suprascapular block
examined with repeated measures ANOVA. Two factors and
their interaction were examined: the group effect (i.e.
anterior vs. posterior); and the time effect. Statistical analyses
were carried out with MedCalc version 12.6.1 (MedCalc
Software, Ostend, Belgium). A value of p < 0.05 was
considered statistically significant.
Results
Between January 2018 and January 2019, 84 patients were
included in our study (Fig. 1). Baseline characteristics were
comparable between the groups (Table 1).
The incidence of hemidiaphragmatic paralysis and its
characteristics are summarised in Table 2. The change in
diaphragmatic excursion and the incidence of hemidia-
phragmatic paralysis are summarised in Table 3 and
represented in Fig. 2. Eighty-nine percent (16/18) of
patients with hemidiaphragmatic paralysis experienced it
30 min after the block. For 8 of the 18 patients with
hemidiaphragmatic paralysis (44%), this dysfunction lasted
less than 4 h, whereas 6 of the 18 patients with
hemidiaphragmatic paralysis (33%) still experienced it after
8 h.
Examination of the global changes in diaphragmatic
excursion showed a significant difference between
measurements over time (p = 0.001), which was dependent
on the approach to the suprascapular nerve block
(p = 0.01). One patient in the anterior group experienced a
clinical impact from total hemidiaphragmatic paralysis with
respiratory distress and oxygen desaturation requiring non-
invasive ventilation delivered in the recovery area via a
Ferr

e et al. | Diaphragmatic paralysis and suprascapular block Anaesthesia 2020

Table 1 Baseline and surgical characteristics of patients landmark that allowed axillary nerve block performance in
randomly allocated to anterior or posterior approach all cases. Concerning the posterior suprascapular nerve
suprascapular nerve block. Values are mean (SD) or number block, we used the suprascapular notch as a landmark when
(proportion).
the suprascapular nerve was not seen (22%) and we injected
Anterior Posterior the local anaesthetic solution under the transverse ligament
n = 42 n = 41
of the scapula, close to the suprascapular artery.
Age; years 57.6 (11.1) 55.5 (12.1)
Table 5 illustrates the data collected in the post-
Female sex 23 (55%) 25 (61%)
anaesthesia care unit (PACU) and at 24 h.
ASA physical status
Global patient satisfaction evaluated by the modified
1 24 (57%) 21 (51%)
EVAN-LR score was comparable between the groups. All
2 16 (38%) 16 (39%) patients except for one (anterior group) responded
3 2 (5%) 4 (10%) positively to the question ‘would you agree to the same
BMI; kg.m
2 26 (4.3) 27.2 (4.6) anaesthesia technique for a subsequent shoulder operation
Side of surgery if it were indicated?’
Right 28 (67%) 29 (71%)
Left 14 (34%) 12 (29%) Discussion
Duration of surgery (min) 52.6 (19.1) 51.9 (17.9) Our work constitutes to our knowledge the first
Type of surgery published randomised study evaluating the influence of
Acromioplasty 7 (17%) 13 (32%) the suprascapular nerve block approach on the
Bankart repair 1 (2%) 2 (5%) incidence of hemidiaphragmatic paralysis diagnosed by
Rotator cuff repair 26 (62%) 14 (34%) ultrasound. We were able to demonstrate that a
Tenodesis 5 (12%) 8 (20%) combined ultrasound-guided anterior suprascapular and
Rotator cuff repair with tenodesis 1 (2%) 1 (2%) axillary nerve block resulted in a 40% incidence of partial
Calcification resection 2 (5%) 3 (7%) or complete hemidiaphragmatic paralysis. In comparison,
only one patient (2%) presented with hemidiaphragmatic
paralysis after a posterior approach to the suprascapular
Table 2 Incidence of hemidiaphragmatic paralysis in nerve.
patients randomly allocated to anterior or posterior
The relatively high incidence of hemidiaphragmatic
approach suprascapular nerve block. Values are number
(proportion). paralysis in the anterior group is consistent with data
previously published demonstrating phrenic nerve
Anterior Posterior
n = 42 n = 41 p value paralysis after regional anaesthesia in the supraclavicular
area. A supraclavicular brachial plexus block is associated
Any hemidiaphragmatic 17 (41%) 1 (2%) < 0.001
paralysis (complete with a 45% incidence of hemidiaphragmatic paralysis [29].
or partial) In this setting, Auyong et al. found a least-square mean
Hemidiaphragmatic < 0.001 (95%CI) reduction in diaphragmatic excursion of 2.1 cm
paralysis
(1.3–2.9 cm) from baseline during a vital capacity breath
Absent 25 (60%) 40 (98%) after a continuous anterior suprascapular nerve block [33].
Partial 14 (33%) 0 (0%) One of the most plausible explanations is local anaesthetic
Complete 3 (7%) 1 (2%) diffusion directly to the phrenic nerve [6]. By conducting a
cadaver study, we previously demonstrated methylene
facemask. Eight hours after the block was performed in this blue diffusion to the phrenic nerve in 20% of ultrasound-
patient, diaphragmatic function assessment revealed partial guided anterior suprascapular nerve blocks [34]. In the
hemidiaphragmatic paralysis. Non-invasive ventilation was supraclavicular region, the distance between the
discontinued and supplemental oxygen was applied at suprascapular nerve and the brachial plexus is less than
2 l.min
1 by nasal prongs. The patient was successfully 1 cm, facilitating diffusion of injectate to the phrenic nerve,
discharged from hospital the next day. especially when large volumes (e.g. > 10 ml) are used [13,
Details of the combined suprascapular and axillary 34]. By analogy with the results found after interscalene
nerves block (procedure time, visualisation, success) are brachial plexus block, a decrease in local anaesthetic
summarised in Table 4. Concerning the axillary nerve, the volume could reduce the incidence of this complication
posterior humeral circumflex artery is an easily identifiable [35, 36].

© 2020 Association of Anaesthetists 5

Anaesthesia 2020 Ferr

e et al. | Diaphragmatic paralysis and suprascapular block

Table 3 Temporal incidence of hemidiaphragmatic paralysis and diaphragmatic excursion during a sniff test in patients
randomly allocated to anterior or posterior approach suprascapular nerve block at different time-points. Values are number
(proportion) or mean (SD).
Baseline 30 min after block 4 h after block 8 h after block
Anterior Posterior p Anterior Posterior p Anterior Posterior p Anterior Posterior
n = 42 n = 41 value n = 42 n = 41 value n = 42 n = 41 value n=9 n=1 p value

Any hemidia- – – 15 (36%) 1 (2.4%) < 0.001 9 (21%) 1 (2%) 0.01 5 (56%) 1 (100%) 1.00
phragmatic
paralysis
(complete
or partial)
Hemidia– < 0.001 0.03 0.11
phragmatic
paralysis
Absent – – 27 (64%) 40 (98%) 33 (79%) 40 (98%) 4 (44%) 0
Partial – – 13 (31%) 1 (2%) 6 (14%) 1 (2%) 4 (44%) 0
Complete – – 2 (5%) 0 3 (7%) 0 1 (11%) 1 (100%)
Diaphrag- 2.9 (0.6) 2.8 (0.7) 0.42 2.4 (1.1) 2.7 (0.7) 0.03 2.6 (0.9) 2.7 (0.7) 0.11 2.7 (1.5) 0.7 NA
matic
excursion
(cm)

NA, not applicable.

By moving away from the supraclavicular region, a
significant reduction in the incidence of hemidiaphragmatic
paralysis was expected with the posterior approach to the
suprascapular nerve; however, surprisingly, one patient
presented with phrenic nerve involvement with a posterior
needle approach [37]. There are several possible
explanations for this. Visualisation of the suprascapular nerve
was absent and could have compromised the precision of
injection. It is possible that the injection was performed
anterior to the suprascapular notch. Moreover, general
anaesthesia and the consumption of opioids per se could
have affected diaphragmatic function [38]. In order to confirm
this hypothesis, a bilateral ultrasound evaluation would have
been of significant interest, but was not performed.
Two-thirds of the hemidiaphragmatic paralysis
diagnosed in our study spontaneously resolved < 8 h. This
result is consistent with the study be Hortense et al., who
found that vital capacity recovered within 6 h of an
interscalene brachial plexus block with ropivacaine [39].
Nevertheless, in our study, 33% of hemidiaphragmatic
paralysis persisted > 8 h. In some high-risk patients, this
extended duration could compromise hospital discharge
on the day of surgery.
We chose the measurement of diaphragmatic
excursion during a voluntary sniff test to evaluate respiratory
function. Other parameters of pulmonary function like SpO2,
peak expiratory flow rate or vital capacity, could have been
collected. However, the sensitivity of SpO2 and/or dyspnoea
for screening diaphragmatic dysfunction is low [6, 40].
Moreover, ultrasound evaluation of diaphragmatic
excursion is a reliable, reproducible, sensitive and specific
method that can be easily performed at the bed-side to
evaluate diaphragmatic function [6, 30, 41, 42]. Finally, with
unilateral dysfunction of the diaphragm that is observed in
the shoulder surgical setting, the reliability of ultrasound
exceeds spirometry testing, which reflects the overall
pulmonary status [6].
The clinical impact of hemidiaphragmatic dysfunction
is a matter of debate. The majority of patients remained
asymptomatic after hemidiaphragmatic paralysis induced
by interscalene brachial plexus block and did not require
specific treatment despite a decrease in their vital capacity
[6, 36, 43]. Nevertheless, one of our patients needed
hospital admission for 24 h to manage respiratory distress
induced by diaphragmatic dysfunction. By assessing the
functionality of the diaphragm, ultrasound can be used to
diagnose diaphragmatic paralysis as well as monitor
recovery [30]. Large studies that target high-risk patients
(e.g. chronic obstructive pulmonary disease, obese) should
be conducted to better evaluate the consequences of
diaphragmatic dysfunction in this population.
In our study, the success rate of sonographic
visualisation of the suprascapular nerve was higher with the
anterior when compared with the posterior approach. This
result is consistent with previously published data [13, 23].
Indeed, for the anterior approach, anatomical landmarks
like the omohyoid muscle are easily identified [13, 34].
However, the feasibility of the posterior approach is limited
by anatomical landmarks that are difficult to visualise as well
as the deep location of the suprascapular nerve [13]. The

6 © 2020 Association of Anaesthetists

Ferr

e et al. | Diaphragmatic paralysis and suprascapular block
Figure 2 Changes in diaphragmatic excursion at baseline, 30 min and 4 h after block performance, compared between
anterior (orange square) and posterior (blue circle) approaches to suprascapular nerve block. Markers represent mean values
and whiskers represent 95%CI.
suprascapular notch presents many anatomical variations
making its identification with ultrasound sometimes difficult,
which may subsequently lead to a failed block [24, 44, 45].
Overall, the posterior approach to the suprascapular nerve
seems to be reserved for experts, which increases the risk of
widening the gulf between regional anaesthesia experts
and generalists [46].
We demonstrated a higher combined suprascapular
and axillary nerve block success rate in the anterior group
for the suprascapular nerve as well as the axillary nerve.
However, an objective and reliable evaluation of the
suprascapular nerve block is made difficult by its anatomical
variations [18, 26]. A confounding factor is the presence of a
cutaneous branch in some patients for which the
distribution can intersect that of the axillary nerve [16, 47,
48]. In addition, the supraspinatus muscle is difficult to
evaluate individually [49] especially in patients with
shoulder pathology where mobilisation can be limited by
pain and pre-existing muscle weakness.
Postoperative opioid requirements were statistically
different between the two groups, with a benefit in favour
of the anterior route in the recovery area and during the
first 24 postoperative hours. To explain this observed
difference, besides failure of the posterior approach to the
© 2020 Association of Anaesthetists
Anaesthesia 2020
suprascapular nerve block, there are two possible non-
exclusive hypotheses. Firstly, there may be an articular
branch of the suprascapular nerve (medial subacromial
branch) that emerges before the suprascapular nerve
passes into the suprascapular notch, and which could only
be effectively blocked using the anterior approach [48].
Alternatively, there may be diffusion of local anaesthetic to
the brachial plexus and consequently to the other sensory
nerve branches of the shoulder (lateral pectoral nerve,
subscapular nerve and musculocutaneous nerve) when
using the anterior approach to the suprascapular nerve
[33]. In fact, Auyong et al. found comparable sensory-
motor anaesthesia in the brachial plexus roots after both
anterior suprascapular nerve block and interscalene
brachial plexus block [21, 33]. Therefore, the anterior
approach is probably not an isolated block of the
suprascapular nerve but rather a partial block of the
brachial plexus. A higher success rate for the axillary nerve
block in the anterior group supports this hypothesis.
Finally, it would have been interesting to compare grip
strength in the two groups in order to definitely confirm
brachial plexus involvement in the anterior group.
There are several limitations to our study. Firstly, it
could have benefited from a comparison with the
7
Anaesthesia 2020 Ferr

e et al. | Diaphragmatic paralysis and suprascapular block

Table 4 Block characteristics in patients randomly allocated to either anterior or posterior approach suprascapular nerve block.
Values are median (IQR [range]) or number (proportion).
Anterior Posterior
n = 42 n = 41 p value
Procedure 5 [3–6.2 (1–15)] 6 [3–10 (2–15)] 0.072
time; min
Suprascapular nerve visualisation <0.001
Absent 0 (0%) 9 (21.9%)
Poor 9 (21.5%) 23 (56.2%)
Good 33 (78.5%) 9 (21.9%)
Nerve block success: sensory assessment
Suprascapular nerve 0.006
Absent 1 (2.4%) 6 (14.6%)
Partial 12 (28.6%) 20 (48.8%)
Complete 29 (69%) 15 (36.6%)
Axillary nerve 0.031
Absent 1 (2.4%) 4 (9.8%)
Partial 4 (9.5%) 11 (26.8%)
Complete 37 (88.1%) 26 (63.4%)
Nerve block success: motor assessment 0.023
Absent 0 (0%) 6 (14.6%)
Partial 27 (64.3%) 25 (61%)
Complete 15 (35.7%) 10 (24.4%)

Table 5 Postoperative analgesia and patient satisfaction in patients randomly allocated to either anterior or posterior approach
suprascapular nerve block. Values are median (IQR [range]) or mean (SD).
Recovery area 24- h
Anterior Posterior Anterior Posterior
n = 42 n = 41 p value n = 39 n = 39 p value
Worst NRS score 3 [0–5 (0–9)] 4 [2–6 (0–8)] 0.149 3 [1–4.5 (0–9)] 3 [2–4 (0–9)] 0.613
Oral morphine equivalent 8 [0–14 (0–52)] 20 [5–31 (0–60)] 0.004 58 [30–86 (0–160)] 82 [61–127 (12–360)] 0.015
consumption; mg
Modified EVAN-LR – – 74.6 (15.7) 72.8 (13.4) 0.609
NRS, numerical rating scale; EVAN-LR, Evaluation du V

ecu de l’Anesth

esie LocoR

egionale (patient sartisfaction score).

interscalene brachial plexus block, notably for In conclusion, we showed a high incidence of
postoperative analgesia. Secondly, the choice of local hemidiaphragmatic paralysis after anterior approach to the
anaesthetic dose is open for discussion. We chose suprascapular nerve block. The posterior approach may be
ropivacaine 0.375% to minimise motor block without technically more difficult but spares diaphragmatic
limiting sensory block. However, the choice of volume of function. Further studies evaluating the optimum dose and
local anaesthetic used for the suprascapular nerve blocks is volume of local anaesthetic to reduce the incidence of
worthy of discussion. Ten millilitres is a high volume and hemidiaphragmatic paralysis are required before
there is likely to have been local anaesthetic spread to the considering the anterior suprascapular nerve block in high-
phrenic nerve with this volume. Nevertheless, it risk patients.
corresponds with the literature regarding suprascapular
nerve blocks [18, 22, 23]. Further studies using lower Acknowledgements
volume (e.g. 5 ml) for an anterior approach to the We thank Dr C. Martin, Dr C. Gris, Professor M. Kurrek and
suprascapular nerve block should be conducted. Professor O. Fourcade. The study was prospectively

8 © 2020 Association of Anaesthetists

Ferr

e et al. | Diaphragmatic paralysis and suprascapular block
registered at clinicaltrials.gov (NCT03352687). Funding was
provided by the Department of Anaesthesiology at the
Teaching Hospital of Purpan, Toulouse. No competing
interests declared.
References
1. Fredrickson MJ, Krishnan S, Chen CY. Postoperative analgesia
for shoulder surgery: a critical appraisal and review of current
techniques. Anaesthesia 2010; 65: 608–24.
2. Ilfeld BM, Vandenborne K, Duncan PW, et al. Ambulatory
continuous interscalene nerve blocks decrease the time to
discharge readiness after total shoulder arthroplasty: a
randomized, triple-masked, placebo-controlled study.
Anesthesiology 2006; 105: 999–1007.
3. Wilson AT, Nicholson E, Burton L, Wild C. Analgesia for day-
case shoulder surgery. British Journal of Anaesthesia 2004; 92:
414–15.
4. Toma O, Persoons B, Pogatzki-Zahn E, Van de Velde M, Joshi
GP. PROSPECT guideline for rotator cuff repair surgery:
systematic review and procedure-specific postoperative pain
management recommendations. Anaesthesia 2019; 74:
1320–31.
5. Abdallah FW, Halpern SH, Aoyama K, Brull R. Will the real
benefits of single-shot interscalene block please stand up? a
systematic review and meta-analysis. Anesthesia and Analgesia
2015; 120: 1114–29.
6. El-Boghdadly K, Chin KJ, Chan VWS. Phrenic nerve palsy and
regional anesthesia for shoulder surgery: anatomical,
physiologic, and clinical considerations. Anesthesiology 2017;
127: 173–91.
7. Urmey WF, McDonald M. Hemidiaphragmatic paresis during
interscalene brachial plexus block: effects on pulmonary
function and chest wall mechanics. Anesthesia and Analgesia
1992; 74: 352–7.
8. Urmey WF, Talts KH, Sharrock NE. One hundred percent
incidence of hemidiaphragmatic paresis associated with
interscalene brachial plexus anesthesia as diagnosed by
ultrasonography. Anesthesia and Analgesia 1991; 72: 498–503.
9. Verelst P, van Zundert A. Respiratory impact of analgesic
strategies for shoulder surgery. Regional Anesthesia and Pain
Medicine 2013; 38: 50–3.
10. Pakala SR, Beckman JD, Lyman S, Zayas VM. Cervical spine
disease is a risk factor for persistent phrenic nerve paresis
following interscalene nerve block. Regional Anesthesia and
Pain Medicine 2013; 38: 239–42.
11. Robaux S, Bouaziz H, Boisseau N, Raucoules-Aim
e M, Laxenaire
MC. Persistent phrenic nerve paralysis following interscalene
brachial plexus block. Anesthesiology 2001; 95: 1519–21.
12. Ferre F, Cugnin N, Martin C, et al. Regional anesthesia with
noninvasive ventilation for shoulder surgery in a patient with
severe chronic obstructive pulmonary disease: a case report.
Anesthesia and Analgesia Case Reports 2017; 8: 261–4.
13. Siegenthaler A, Moriggl B, Mlekusch S, et al. Ultrasound-
guided suprascapular nerve block, description of a novel
supraclavicular approach. Regional Anesthesia and Pain
Medicine 2012; 37: 325–8.
14. Tran DQ, Elgueta MF, Aliste J, Finlayson RJ. Diaphragm-
sparing nerve blocks for shoulder surgery. Regional Anesthesia
and Pain Medicine 2017; 42: 32–8.
15. Eckmann MS, Bickelhaupt B, Fehl J, et al. Cadaveric study of
the articular branches of the shoulder joint. Regional
Anesthesia and Pain Medicine 2017; 42: 564–70.
16. Aszmann OC, Dellon AL, Birely BT, McFarland EG. Innervation
of the human shoulder joint and its implications for surgery.
Clinical Orthopaedics and Related Research 1996; 202–7.
© 2020 Association of Anaesthetists
Anaesthesia 2020
17. Checcucci G, Allegra A, Bigazzi P, Gianesello L, Ceruso M, Gritti
G. A new technique for regional anesthesia for arthroscopic
shoulder surgery based on a suprascapular nerve block and an
axillary nerve block: an evaluation of the first results.
Arthroscopy 2008; 24: 689–96.
18. Dhir S, Sondekoppam RV, Sharma R, Ganapathy S, Athwal GS.
A comparison of combined suprascapular and axillary nerve
blocks to interscalene nerve block for analgesia in arthroscopic
shoulder surgery: an equivalence study. Regional Anesthesia
and Pain Medicine 2016; 41: 564–71.
19. Lee JJ, Kim DY, Hwang JT, et al. Effect of ultrasonographically
guided axillary nerve block combined with suprascapular nerve
block in arthroscopic rotator cuff repair: a randomized
controlled trial. Arthroscopy 2014; 30: 906–14.
20. Price DJ. The shoulder block: a new alternative to interscalene
brachial plexus blockade for the control of postoperative
shoulder pain. Anaesthesia and Intensive Care 2007; 35: 575–81.
21. Auyong DB, Hanson NA, Joseph RS, Schmidt BE, Slee AE, Yuan
SC. Comparison of anterior suprascapular, supraclavicular, and
interscalene nerve block approaches for major outpatient
arthroscopic shoulder surgery: a randomized, double-blind,
noninferiority trial. Anesthesiology 2018; 129: 47–57.
22. Neuts A, Stessel B, Wouters PF, et al. Selective suprascapular
and axillary nerve block versus interscalene plexus block for
pain control after arthroscopic shoulder surgery: a
noninferiority randomized parallel-controlled clinical trial.
Regional Anesthesia and Pain Medicine 2018; 43: 738–44.
23. Wiegel M, Moriggl B, Schwarzkopf P, Petroff D, Reske AW.
Anterior suprascapular nerve block versus interscalene
brachial plexus block for shoulder surgery in the outpatient
setting: a randomized controlled patient- and assessor-blinded
trial. Regional Anesthesia and Pain Medicine 2017; 42: 310–18.
24. Chan CW, Peng PW. Suprascapular nerve block: a narrative
review. Regional Anesthesia and Pain Medicine 2011; 36: 358–
73.
25. Harmon D, Hearty C. Ultrasound-guided suprascapular nerve
block technique. Pain Physician 2007; 10: 743–6.
26. Rothe C, Steen-Hansen C, Lund J, Jenstrup MT, Lange KH.
Ultrasound-guided block of the suprascapular nerve – a
volunteer study of a new proximal approach. Acta
Anaesthesiologica Scandinavica 2014; 58: 1228–32.
27. Rothe C, Asghar S, Andersen HL, Christensen JK, Lange KH.
Ultrasound-guided block of the axillary nerve: a volunteer study
of a new method. Acta Anaesthesiologica Scandinavica 2011;
55: 565–70.
28. Boussuges A, Gole Y, Blanc P. Diaphragmatic motion studied
by m-mode ultrasonography: methods, reproducibility, and
normal values. Chest 2009; 135: 391–400.
29. Petrar SD, Seltenrich ME, Head SJ, Schwarz SK.
Hemidiaphragmatic paralysis following ultrasound-guided
supraclavicular versus infraclavicular brachial plexus blockade:
a randomized clinical trial. Regional Anesthesia and Pain
Medicine 2015; 40: 133–8.
30. McCool FD, Tzelepis GE. Dysfunction of the diaphragm. New
England Journal of Medicine 2012; 366: 932–42.
31. Maurice-Szamburski A, Bruder N, Loundou A, Capdevila X,
Auquier P. Development and validation of a perioperative
satisfaction questionnaire in regional anesthesia. Anesthesiology
2013; 118: 78–87.
32. Stolz LA, Acuna JG, Gaskin K, et al. Echogenicity and
ultrasound visibility of peripheral nerves of the upper extremity.
Medical Ultrasonography 2018; 20: 199–204.
33. Auyong DB, Yuan SC, Choi DS, Pahang JA, Slee AE, Hanson
NA. A double-blind randomized comparison of continuous
interscalene, supraclavicular, and suprascapular blocks for total
shoulder arthroplasty. Regional Anesthesia and Pain Medicine
2017; 42: 302–9.
9
Anaesthesia 2020
34. Laumonerie P, Ferre F, Cances J, et al. Ultrasound-guided
proximal suprascapular nerve block: a cadaveric study. Clinical
Anatomy 2018; 31: 824–9.
35. Lee JH, Cho SH, Kim SH, et al. Ropivacaine for ultrasound-
guided interscalene block: 5 mL provides similar analgesia but
less phrenic nerve paralysis than 10 mL. Canadian Journal of
Anesthesia 2011; 58: 1001–6.
36. Riazi S, Carmichael N, Awad I, Holtby RM, McCartney CJ. Effect
of local anaesthetic volume (20 vs. 5 ml) on the efficacy and
respiratory consequences of ultrasound-guided interscalene
brachial plexus block. British Journal of Anaesthesia 2008; 101:
549–56.
37. Feigl GC, Anderhuber F, Dorn C, Pipam W, Rosmarin W, Likar R.
Modified lateral block of the suprascapular nerve: a safe
approach and how much to inject? A morphological study.
Regional Anesthesia and Pain Medicine 2007; 32: 488–94.
38. Sasaki N, Meyer MJ, Eikermann M. Postoperative respiratory
muscle dysfunction: pathophysiology and preventive
strategies. Anesthesiology 2013; 118: 961–78.
39. Hortense A, Perez MV, Amaral JL, Oshiro AC, Rossetti HB.
Interscalene brachial plexus block. Effects on pulmonary
function. Brazilian Journal of Anesthesiology 2010; 60:
74–8.
40. Wiesmann T, Feldmann C, Muller HH, et al. Phrenic palsy and
analgesic quality of continuous supraclavicular vs. interscalene
plexus blocks after shoulder surgery. Acta Anaesthesiologica
Scandinavica 2016; 60: 1142–51.
41. Matamis D, Soilemezi E, Tsagourias M, et al. Sonographic
evaluation of the diaphragm in critically ill patients. Technique
10
Ferr

e et al. | Diaphragmatic paralysis and suprascapular block
and clinical applications. Intensive Care Medicine 2013; 39:
801–10.
42. Boon AJ, Sekiguchi H, Harper CJ, et al. Sensitivity and
specificity of diagnostic ultrasound in the diagnosis of phrenic
neuropathy. Neurology 2014; 83: 1264–70.
43. Palhais N, Brull R, Kern C, et al. Extrafascial injection for
interscalene brachial plexus block reduces respiratory
complications compared with a conventional intrafascial
injection: a randomized, controlled, double-blind trial. British
Journal of Anaesthesia 2016; 116: 531–7.
44. Natsis K, Totlis T, Tsikaras P, Appell HJ, Skandalakis P, Koebke J.
Proposal for classification of the suprascapular notch: a study
on 423 dried scapulas. Clinical Anatomy 2007; 20: 135–9.
45. Podgorski M, Rusinek M, Cichosz M, Olewnik L, Polguj M,
Grzelak P. “Pseudo-suprascapular notch”: is it a sonographic
trap in suprascapular nerve block? Regional Anesthesia and
Pain Medicine 2019; 44: 77–80.
46. Turbitt LR, Mariano ER, El-Boghdadly K. Future directions in
regional anaesthesia: not just for the cognoscenti. Anaesthesia
2019; https://doi.org/10.1111/anae.14768.
47. Ajmani ML. The cutaneous branch of the human suprascapular
nerve. Journal of Anatomy 1994; 185: 439–42.
48. Vorster W, Lange CP, Briet RJ, et al. The sensory branch
distribution of the suprascapular nerve: an anatomic study.
Journal of Shoulder and Elbow Surgery 2008; 17: 500–2.
49. McCully SP, Suprak DN, Kosek P, Karduna AR. Suprascapular
nerve block results in a compensatory increase in deltoid
muscle activity. Journal of Biomechanics 2007; 40:
1839–46.
© 2020 Association of Anaesthetists
Anaesthesia 2020, 75, 509–528 doi:10.1111/anae.14904

Guidelines

Difficult Airway Society guidelines for awake tracheal

intubation (ATI) in adults
I. Ahmad1,2 K. El-Boghdadly,1,2 R. Bhagrath,3 I. Hodzovic,4,5 A. F. McNarry,6 F. Mir,7
E. P. O’Sullivan,8 A. Patel,9 M. Stacey10 and D. Vaughan11

1 Consultant, Department of Anaesthesia, Guy’s and St Thomas’ NHS Foundation Trust, London, UK
2 Honorary Senior Lecturer, King’s College London, London, UK
3 Consultant, Department of Anaesthesia, Barts Health NHS Trust, London, UK
4 Senior Lecturer, Department of Anaesthesia, Cardiff University School of Medicine, Cardiff, UK
5 Honorary Consultant, Department of Anaesthesia, Aneurin Bevan University Health Board, Newport, UK
6 Consultant, Department of Anaesthesia, NHS Lothian, Edinburgh, UK
7 Consultant, Department of Anaesthesia, St. George’s University Hospital NHS Foundation Trust, London, UK
8 Consultant, Department of Anaesthesia, St James’s Hospital, Dublin, Ireland
9 Consultant, Department of Anaesthesia, Royal National Throat Nose and Ear Hospital and University College London
Hospitals NHS Foundation Trust, London, UK
10 Consultant, Department of Anaesthesia, Cardiff and Vale NHS Trust (HEIW), Cardiff, UK
11 Consultant, Department of Anaesthesia, Northwick Park Hospital, London, UK

Summary
Awake tracheal intubation has a high success rate and a favourable safety profile but is underused in cases of
anticipated difficult airway management. These guidelines are a comprehensive document to support decision
making, preparation and practical performance of awake tracheal intubation. We performed a systematic
review of the literature seeking all of the available evidence for each element of awake tracheal intubation in
order to make recommendations. In the absence of high-quality evidence, expert consensus and a Delphi study
were used to formulate recommendations. We highlight key areas of awake tracheal intubation in which specific
recommendations were made, which included: indications; procedural setup; checklists; oxygenation; airway
topicalisation; sedation; verification of tracheal tube position; complications; management of unsuccessful
awake tracheal intubation; post-tracheal intubation management; consent; and training. We recognise that
there are a range of techniques and regimens that may be effective and one such example technique is
included. Breaking down the key practical elements of awake tracheal intubation into sedation, topicalisation,
oxygenation and performance might help practitioners to plan, perform and address complications. These
guidelines aim to support clinical practice and help lower the threshold for performing awake tracheal
intubation when indicated.

.................................................................................................................................................................
Correspondence to: K. El-Boghdadly
Email: elboghdadly@gmail.com
Accepted: 6 October 2019
Keywords: airway management; bronchoscopy; laryngoscopy; tracheal intubation; training; videolaryngoscopy
This article is accompanied by an editorial by Aziz and Kristensen, Anaesthesia, 2020; 75:442–6.
Twitter: @dr_imranahmad, @elboghdadly, @altgm, @ravibhagrath, @ProfEllenO, @oldandbaffled

.................................................................................................................................................................
Re-use of this article is permitted in accordance with the Creative Commons Deed, Attribution 2.5, which does not permit
commercial exploitation.

© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists 509
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and
distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
Anaesthesia 2020, 75, 509–528
Recommendations
1 Awake tracheal intubation must be considered in the
presence of predictors of difficult airway management.
2 A cognitive aid such as a checklist is recommended
before and during performance of awake tracheal
intubation.
3 Supplemental oxygen should always be administered
during awake tracheal intubation.
4 Effective topicalisation must be established and tested.
The maximum dose of lidocaine should not exceed
9 mg.kg 1
lean body weight.
5 Cautious use of minimal sedation can be beneficial. This
should ideally be administered by an independent
practitioner. Sedation should not be used as a substitute
for inadequate airway topicalisation.
6 The number of attempts should be limited to three, with
one further attempt by a more experienced operator
(3 + 1).
7 Anaesthesia should only be induced after a two-point
check (visual confirmation and capnography) has
confirmed correct tracheal tube position.
8 All departments should support anaesthetists to attain
competency and maintain skills in awake tracheal
intubation.
Why were these guidelines developed?
Awake tracheal intubation (ATI) has a high success rate
and a low-risk profile and has been cited as the gold
standard in airway management for a predicted difficult
airway. However, ATI is reported to be used in as few as
0.2% of all tracheal intubations in the UK [1]. There are
barriers preventing broad uptake and use of awake
techniques for securing the airway. We aimed to produce
generalisable guidelines to improve patient safety by
making ATI more accessible to all clinicians, trainers and
institutions. Rather than inform expert practice, these
guidelines aim to support the use of ATI by more
clinicians, with a particular focus on those that do not
regularly perform ATI. There remains heterogeneity in
clinical practice, underscoring the need for a more
consistent approach using the available evidence, which
these guidelines aim to deliver.
What other guidelines exist?
Although there are many guidelines on unanticipated
difficult airway management [2–13], there are few that
specifically focus on the anticipated difficult airway. The
ASA, Canadian Airway Focus Group, French Society for
Anesthesia and Intensive Care and German Society for
510 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Ahmad et al. | Awake tracheal intubation guidelines
Anesthesiology and Intensive Care describe clinical decision
making in the anticipated difficult airway [6, 7, 10, 12].
How do these guidelines differ from
existing guidelines?
At the time of writing, there were no nationally or
internationally agreed guidelines on the practical
performance of ATI.
Disclaimer
These guidelines are not intended to represent a minimum
standard of practice, nor are they to be regarded as a
substitute for good clinical judgement. They present key
principles and suggested strategies for preparation,
performance, consent and training to inform clinical
practice. This document is intended to guide appropriately
trained operators.
Introduction
A strategy for difficult airway management is necessary
when facemask ventilation, supraglottic airway device (SAD)
placement or ventilation, tracheal intubation or insertion of
a front-of-neck airway (FONA) is predicted to be
challenging. The incidence of difficult facemask ventilation
is 0.66–2.5% [14–17], difficult SAD placement or ventilation
0.5–4.7% [18–22], difficult tracheal intubation 1.9–10% [14,
16, 23–25] and combined difficulty in both facemask and
tracheal intubation 0.3–0.4% [16]. As a rescue technique
after failed tracheal intubation, one study reported that
SADs have a success rate as low as 65% in difficult airway
management [26]. The reported incidence of requirement
for emergency FONA and death due to airway management
are 0.002–0.07% (1:50,000–1:1400) [1, 27, 28] and 0.0006–
0.04% (1:180,000–1:2800), respectively [1, 28]. The risk and
severity of adverse outcomes during difficult airway
management is highlighted by the plethora of guidelines
and cognitive aids for airway rescue [29].
Awake tracheal intubation involves placing a tracheal
tube in an awake, spontaneously-breathing patient, most
commonly with flexible bronchoscopy (ATI:FB) or
videolaryngoscopy (ATI:VL, Table 1). This allows the airway
to be secured before induction of general anaesthesia,
avoiding the potential risks and consequences of difficult
airway management in an anaesthetised patient [30].
Awake tracheal intubation has a favourable safety
profile because both spontaneous ventilation and intrinsic
airway tone are maintained until the trachea is intubated
[31–35]. Awake tracheal intubation can be unsuccessful in
1–2% of cases, but this rarely leads to airway rescue
strategies or death [33–35]. These guidelines aim to
Ahmad et al. | Awake tracheal intubation guidelines Anaesthesia 2020, 75, 509–528

Table 1 A summary of terms used in these guidelines.

Term Definition
ATI Awake tracheal intubation
ATI:FB Awake tracheal intubation using flexible bronchoscopy
ATI:VL Awake tracheal intubation using videolaryngoscopy
FONA Front-of-neck airway
sTOP Sedation, topicalisation, oxygenation, performance
Minimal sedation Drug-induced state during which the patient responds normally to verbal commands, while the airway,
spontaneous ventilation and cardiovascular function are unaffected
Airway topicalisation Topical application of local anaesthetic to the airway
Performance The practical conduct of awake tracheal intubation
Two-point check 1. Visualisation of the tracheal lumen with ATI:FB or tracheal tube through the cords with ATI:VL to
confirm tracheal placement
2. Capnography to exclude oesophageal intubation
Unsuccessful attempt Unplanned removal of flexible bronchoscope, videolaryngoscope or tracheal tube from the airway
Unsuccessful ATI Successful tracheal intubation not achieved after 3 + 1 attemptsa
a
Three attempts by the primary operator and a fourth attempt by a more experienced operator.

increase the use of ATI by providing clear guidance for
clinicians to support decision making, preparation and
performance of ATI in the setting of a predicted difficult
airway.
Methods
The development of these guidelines followed the appraisal
of guidelines for research and evaluation (AGREE) reporting
checklist [36]. To ensure these guidelines are supported
by best evidence, a systematic review adhering to the
preferred reporting items for systematic reviews and
meta-analyses (PRISMA) recommendations [37]
performed. We sought published data of relevance to
ATI, including decision making, technical performance,
complications, training and non-technical aspects. Details
of the search, screening and study selection are shown in
the supplementary material (Supporting Information,
Appendix S1).
Data from included studies were synthesised and
consensus from all 10 members of the guideline group was
sought to formulate guideline recommendations using a
three-round Delphi method [38, 39]. The first round
entailed an initially proposed longlist
recommendations, which were each reviewed and rated
for content and clarity. The recommendations in which six
or more members of the guideline group approved were
shortlisted. A second round of rating was then undertaken,
in which the highest rated recommendations were
selected. Finally, a third round involving recommendation-
ratification in round-table discussions was undertaken.
These recommendations were based on a number of
factors, including: volume and consistency of supportive
evidence; applicability and generalisability of the evidence
to current practice; and clinical and practical implications
of recommendations.
We determined the level of evidence and graded the
strength of subsequent recommendations using a modified
version of the system developed by the Centre for Evidence-
based Medicine (Oxford, UK) (Table 2) [40]. Each
recommendation was graded A to D according to the
strength of the available evidence [41].
Over 3 years the guideline group met 21 times in
was person and 14 times remotely in order to develop, draft and
finalise these guidelines. Draft versions were presented at
the 2017 and 2018 Difficult Airway Society (DAS) annual
scientific meetings. We sent an electronic survey to DAS
members (n = 2150) to capture their opinions, preferences
and clinical experiences in ATI, of whom 632 (29%)
responded. This survey highlighted the need for guidelines
for ATI and the role of a standardised technique for training
and clinical practice. We also performed a survey of 43
international experts, seeking details on commonly used
strategies for oxygenation, topicalisation, sedation and
of performance of ATI. Patient and public involvement was
also used to explore the views and experiences of patients
who had undergone ATI. This was achieved by conducting a
fully anonymised multicentre structured survey of 100
patients, where we explored the self-reported experiences
of the overall conduct of ATI. We consulted an anaesthetic
assistant during the preparation of these guidelines and
invited a senior anaesthetic nurse and two consultant head
and neck surgeons to comment on the final draft. A draft

© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists 511
Anaesthesia 2020, 75, 509–528 Ahmad et al. | Awake tracheal intubation guidelines

Table 2 Grading of recommendations based on the level of evidence available.

Grade Level of evidence available
A • Consistent systematic reviews of RCTs, single RCTs or all-or-none studies
B • Consistent systematic reviews of low-quality RCTs or cohort studies, individual cohort study, or epidemiological
outcome studies
• Consistent systematic reviews of case–control studies, individual case–control studies
• Extrapolations from systematic reviews of RCTs, single RCTs or all-or-none studies
C • Case series, case reports
• Extrapolations from systematic reviews of low-quality RCTs, cohort studies or case–control studies,
individual cohort study, epidemiological outcome studies, individual case–control studies
• Extrapolations from systematic reviews of case–control studies
D • Expert opinion or ideas based on theory, bench studies or first principles alone
• Troublingly inconsistent or inconclusive studies of any level
RCT, randomised controlled trial.

manuscript of these guidelines was then sent to 13
international experts with clinical or academic experience
related to ATI to gather specific comments and feedback on
recommendations and to assess applicability and feasibility.
The guideline group considered the responses from expert
reviewers to inform the final recommendations. The final
draft of the guideline was then submitted to DAS executive
committee for ratification.
Indications
Prediction of difficult airway management is unreliable [14,
23, 42], but there are common features that have been
identified in patients requiring ATI. These include, but are
not limited to: patients with head and neck pathology
(including malignancy, previous surgery or radiotherapy);
reduced mouth opening; limited neck extension;
obstructive sleep apnoea; morbid obesity; and progressive
airway compromise [32, 33, 35, 43, 44]. There is limited
evidence for any individual, validated, predictive
assessment tool developed specifically for ATI. Airway
assessment including history, examination and appropriate
investigations, is indicated for all patients [1, 2, 7, 45] (Grade
D). Awake tracheal intubation must be considered in the
presence of predictors of difficult airway management
(Grade D). In an elective setting the patient should be
appropriately fasted (Grade D). In the non-fasted patient,
the potential for regurgitation or aspiration of gastric
contents still exists even with ATI. There are few relative
contra-indications to ATI (e.g. local anaesthetic allergy,
airway bleeding, unco-operative patients) but the only
absolute contra-indication is patient refusal.
Procedural setup
Awake tracheal intubation can be associated with the
greatest operator-related physical, mental and
psychological stress of all elective airway management
interventions [46]. These stressors may be associated with
suboptimal performance [47, 48], increasing the risk of
complications including failure. Teamwork, good
communication and appropriate preparation may mitigate
these challenges [48–50] and the importance of well-
trained, competent assistants should not be
underestimated. Safety should not be compromised by
time pressures presented by other staff members; therefore
planning and communication with anaesthetic assistants,
operating theatre nursing staff, surgeons and skilled
anaesthetic colleagues is essential (Grade D).
Consideration and planning of the appropriate location
is essential. Awake tracheal intubation should ideally be
performed in the operating theatre environment (Grade D).
This setting has ready access to skilled assistance, drugs,
equipment and space. For high-risk patients, including
those with significant airway obstruction, hypoxia,
respiratory failure, challenging or failed ATI, the operating
theatre may have advantages over an anaesthetic room [1,
51], such as greater space and immediate surgical
assistance. When ATI is performed outside of the theatre
environment (e.g. in the critical care unit or the emergency
department), the same standards of care should apply
(Grade D) [52].
Monitoring patients’ physiological parameters during
anaesthetic care mitigates risks and may alert operators to
impending complications [53–57]. Frequently occurring
avoidable complications in ATI that may be detected by
monitoring are airway obstruction and hypoventilation
secondary to over-sedation [33–35, 58]. Disturbances to
cardiac rhythm and blood pressure following administration
of pharmacological agents for topicalisation and sedation
are possible [35, 59–61]. In accordance with Association of
Anaesthetists’ guidelines for patients receiving sedation

512 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Ahmad et al. | Awake tracheal intubation guidelines Anaesthesia 2020, 75, 509–528

[62], it is recommended that ECG, non-invasive blood Complications or unsuccessful ATI, although
pressure, pulse oximetry and continuous end-tidal carbon uncommon, should be prepared for [33–35], and
dioxide monitoring are used throughout the process of ATI immediate access to emergency drugs, staff and equipment
(Grade C). It is acknowledged that end-tidal carbon dioxide is essential (Grade C). A plan for unsuccessful ATI, including
monitoring during ATI may be challenging in current possible postponement, FONA or high-risk general
practice. anaesthesia, should be discussed explicitly and agreed on
Workspace ergonomics have an impact on by all team members before beginning the procedure
performance and safety [51, 63], and should be considered (Grade D).
before starting the procedure (Grade D; Fig. 1; Supporting It is important to select an appropriate route for tracheal
Information, Appendix S2) [52]. This includes optimising intubation, visualisation device and tracheal tube. The route
the position of patient, operator and assistants, as well as for tracheal intubation should take into account patient
location of equipment and monitors, which should be in the anatomy, surgical access and tracheal extubation plan
direct line of sight of the operator. There is no consensus on (Grade D). For example, in patients with limited mouth
the ideal operator or patient position [64–67], but there are opening, the nasal approach may be the only option, while
physiological and anatomical advantages to having patients in patients having nasal surgery, the oral approach may be
sitting up [68–70]. the preferred route. There is no evidence or consensus

(a) (b)

Figure 1 Examples of ergonomics for awake tracheal intubation (ATI). The primary operator should have a direct line of sight of
the patient, video monitor and patient monitor, as well as immediate access to infusion pumps, anaesthetic machine, suction
and oxygen delivery device. If a second anaesthetist is present, they should be positioned with a direct line of sight of the patient
and have immediate access to infusion pumps, as well as be able to access all other equipment. The anaesthetic assistant’s
primary position should be with immediate access to the airway trolley, and in proximity to the operator. (a) Awake tracheal
intubation performed with the operator positioned facing the patient who is in a sitting up position. (b) Awake tracheal
intubation performed with the operator positioned behind the supine/semi-recumbent patient. This figure forms part of the
Difficult Airway Society guidelines for ATI in adults and should be used in conjunction with the text. ©Difficult Airway
Society 2019.

© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists 513
Anaesthesia 2020, 75, 509–528
among experts demonstrating superiority of one route if
both are feasible [33–35, 64, 71].
Awake tracheal intubation using videolaryngoscopy
has a comparable success rate and safety profile to ATI:FB
(98.3% each) [31]. Choosing between techniques is based
on patient factors, operator skills and availability of
equipment (Grade A). For example, in patients with limited
mouth opening, a large tongue or fixed flexion deformity of
the neck, ATI:FB may be more appropriate. Conversely,
patients with airway bleeding may be more suitable for an
ATI:VL technique. If the chosen ATI technique is
unsuccessful, practitioners should consider using an
alternative (e.g. ATI:FB if ATI:VL is unsuccessful or vice versa;
Grade D). A combined approach to ATI using both VL and
FB has been described [72–74] and could be considered in
complex clinical scenarios (Grade D). In a well-topicalised
patient, insertion of an SAD as a conduit for ATI:FB has also
been described [75, 76], and may provide the benefit of
maintaining airway patency. Single-use flexible
bronchoscopes are associated with a similar safety profile to
re-usable ones [77]. Operators should defer to local
availability and personal experience in determining which
flexible bronchoscope to use (Grade B). There is currently
no evidence or consensus to support the safety or efficacy of
any individual videolaryngoscope. For ATI:VL practitioners
should use videolaryngoscopes with which they are most
familiar (Grade B).
Careful selection of tracheal tube is integral to the
success of any ATI technique. This should factor in size
(internal and external diameter), shape, length, tip design
and material. For ATI:FB, reinforced, Parker Flex-TipTM
(Bridgewater, CN, USA) and intubating laryngeal mask
â TM
airway tubes (LMA Fastrach ETT, Teleflex, Beaconsfield,
UK) have been shown to be superior to standard
polyvinylchloride (PVC) tracheal tubes in terms of ease of
tracheal intubation, railroading (advancing the tracheal
tube over the flexible bronchoscope) and decreasing
laryngeal impingement [78–86]. Therefore, the use of a
standard PVC tracheal tube is not recommended (Grade A).
Using the smallest appropriate external diameter tracheal
tube is advisable, as this may reduce the incidence of
impingement [87] (Grade B). Positioning the bevel of the
tracheal tube posteriorly is recommended [80, 82, 86]
(Grade A). For ATI:VL, tracheal tube selection is similar to
that in an asleep patient and is influenced by the VL
selected.
Checklists
In the peri-operative setting the use of cognitive aids, such
as checklists, improves inter-professional communication,
514 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Ahmad et al. | Awake tracheal intubation guidelines
teamwork and patient outcomes [88–91]. In anaesthetic
practice, cognitive aids enhance performance in simulated
emergency scenarios [92, 93], and their use been
recommended in elective airway management [1]. Given
the potential benefits, we recommend a cognitive aid such
as a checklist before and during performance of ATI (Grade
D; Supporting Information, Appendix S2). The key
components of ATI are sedation, topicalisation,
oxygenation and performance (sTOP; Fig. 2). The ‘s’ is in
lower case to emphasise the optional nature of sedation.
Oxygenation
The reported incidence of desaturation (SpO2 ≤ 90%) with
1
low-flow (< 30 l.min ) oxygen techniques during ATI
ranges between 12% and 16% [58, 94, 95]. When warmed
and humidified high-flow nasal oxygen is used, the reported
incidence of desaturation is 0–1.5% [33, 96]; this was the
most common oxygenation strategy used by experts
responding to our survey. Although there are no
randomised controlled trials comparing air vs. oxygen
during ATI, data from bronchoscopy studies demonstrate
that there is a significant difference in the incidence and
severity of desaturation [97, 98]. In patients receiving
sedation in a variety of settings, administration of oxygen
has been shown to reduce the incidence of desaturation
when compared with air [97, 99–102]. United Kingdom,
European and North American recommendations for
sedation all suggest the use of supplemental oxygen [103–
105]. Whilst airway topicalisation alone may rarely be
associated with desaturation and airway obstruction [59,
106], there is no significant difference in the incidence of
desaturation between ATI:FB and ATI:VL techniques, and
therefore the recommendations apply to both approaches
[31]. The administration of supplemental oxygen during ATI
is recommended (Grade B). This should be started on
patient arrival for the procedure and continued throughout
(Grade D). If available, high-flow nasal oxygen should be the
technique of choice (Grade C).
Airway topicalisation
The success of ATI depends on effective topical application
of local anaesthetic to the airway. Vasoconstriction of the
nasal passage reduces the incidence of epistaxis [107, 108].
The use of topical nasal vasoconstrictors before
nasotracheal intubation is recommended (Grade A).
Lidocaine has theoretical safety benefits over other
local anaesthetic agents due to a favourable cardiovascular
and systemic toxicity risk profile [109]; this is the most
commonly used local anaesthetic agent for ATI. Following
airway topicalisation, clinical evidence of toxicity or levels
Ahmad et al. | Awake tracheal intubation guidelines Anaesthesia 2020, 75, 509–528

Figure 2 The Difficult Airway Society awake tracheal intubation (ATI) technique. This figure forms part of the Difficult Airway
Society guidelines for ATI in adults and should be used in conjunction with the text. HFNO, high-flow nasal oxygen; LA, local
anaesthetic; FB, flexible bronchoscopy; MAD, mucosal atomising device; TCI, target-controlled infusion; Ce, effect-site
concentration; VL, videolaryngoscopy. ©Difficult Airway Society 2019.

exceeding toxic plasma concentrations have been shown
1
with lidocaine doses of 6.0–9.3 mg.kg lean body weight
[110–114]. The dose of topical lidocaine should not exceed
1
9 mg.kg lean body weight (Grade C) [115]. The
1 1
recommendation of 9 mg.kg rather than 9.3 mg.kg is
a pragmatic decision to allow ease of calculation.
Practitioners should recognise that this is not a target but a
maximum dose, and in practice this is rarely required. The
total dose of all local anaesthetics administered, regardless
of route (e.g. regional anaesthesia or surgical infiltration),
must also be considered (Grade D). Some studies have
shown that lower concentrations of lidocaine are as effective
as higher concentrations [112, 116–118], but higher
concentrations may be associated with more rapid onset of
airway anaesthesia. As with all local techniques, a high index
of suspicion of the rare possibility of local anaesthetic
toxicity with appropriate training, procedures and
emergency drug provision (including lipid emulsion) should
be in place [119–122] (Grade D). The use of cocaine for
topicalisation and vasoconstriction can be associated with
toxic cardiovascular complications [123–126], while its
analgesic efficacy during nasotracheal tube insertion is no
better than co-phenylcaine (2.5 ml lidocaine 5%/
phenylephrine 0.5%) [127]. Cocaine in this setting is
therefore not advised, and phenylephrine in combination
with lidocaine is more appropriate (Grade A).
Depending on the delivery device used, there is
variable local anaesthetic absorption [128] but this should
not affect the maximal dose calculation. There is insufficient
evidence to recommend any individual topicalisation

© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists 515
Anaesthesia 2020, 75, 509–528
technique (e.g. mucosal atomisation, spray-as-you-go,
transtracheal injection, nebulisation) [129]. However, blocks
of the glossopharyngeal and superior laryngeal nerves have
been associated with higher plasma concentrations of local
anaesthetic [130], local anaesthetic systemic toxicity [131]
and lower patient comfort [132]. Invasive techniques should
therefore be reserved for those with expertise in their
performance (Grade B). Nebulised lidocaine can be used
but absorption is variable [133]; consequently higher doses
have been used to compensate for this [59]. Regardless of
technique used, the adequacy of topicalisation should be
tested in an atraumatic manner before airway
instrumentation [134] (Grade D), for example, with a soft
suction catheter or Yankauer sucker.
The use of an antisialogogue is not mandatory in the
performance of ATI and may be associated with undesirable
clinical consequences (Grade D; Table 3) [135]. There is
limited evidence to support their use in ATI, but in
anaesthetised patients the clarity of a visual field through a
flexible bronchoscope may be improved [136]. If used,
intramuscular antisialogogues should be injected 40–
60 min before performing ATI, for peak mucosal drying
effect, but there are few data on the intravenous (i.v.) route in
this setting.
Sedation
Awake tracheal intubation may be safely and effectively
performed without sedation [33, 59, 60]. However, its use
during ATI can reduce patient anxiety and discomfort and
increase procedural tolerance [137]. Minimal sedation is
defined as “a drug-induced state during which the patient
responds normally to verbal commands, whilst the airway,
spontaneous ventilation and cardiovascular function are
unaffected” [138]. Sedative drugs can produce a number of
effects which may be considered desirable (e.g. amnesia) or
detrimental (e.g. over-sedation). The risk of over-sedation
and its sequelae, including respiratory depression, airway
loss, hypoxia, aspiration and cardiovascular instability, make
the presence of an independent anaesthetist delivering,
monitoring and titrating sedation desirable [1] (Grade D). In
certain patient populations, the risk of over-sedation is
particularly hazardous, thus an independent practitioner
delivering sedation is strongly recommended (Grade D;
Table 4). If required, we recommend the cautious use of
minimal sedation (Grade D).
Remifentanil and dexmedetomidine are associated
with high levels of patient satisfaction and low risk of over-
sedation and airway obstruction when used for ATI [137]. A
single-agent strategy is safest for the non-expert, and if
used, remifentanil or dexmedetomidine are appropriate
516 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Ahmad et al. | Awake tracheal intubation guidelines
(Grade A). As a sole sedative agent, propofol is associated
with a greater risk of over-sedation, coughing and airway
obstruction than remifentanil [139–141] and is therefore not
advisable in this setting (Grade A) [137]. If co-administration
of sedative agents is to be performed, remifentanil and
midazolam are both reversible and therefore appropriate,
recognising the increased risk of over-sedation (Grade D).
Sedation should not be used as a substitute for inadequate
airway topicalisation (Grade D) [129]. A suggested sedation
regimen is presented in Fig. 2.
Two-point check of tracheal tube
placement
Awake tracheal intubation can result in incorrect tracheal
tube placement, including pharyngeal, oesophageal or
bronchial intubation. Oesophageal intubation occurs in
2.3% of procedures with ATI:FB and 4.9% with ATI:VL [58].
Capnography has 100% sensitivity and specificity in
identifying correct tracheal tube positioning in patients who
lungs are ventilated [142, 143]. However, in a patient who is
spontaneously breathing, a capnographic trace may also be
seen with supraglottic or bronchial placement of the
tracheal tube. A two-point check is therefore required to
confirm the position of the tracheal tube:
1 visualisation of the tracheal lumen with ATI:FB or the
tracheal tube through the vocal cords with ATI:VL to
confirm tracheal placement; and
2 capnography to exclude oesophageal intubation
(Grade C)
Anaesthesia should be induced only when the two-
point check has confirmed correct tracheal tube placement
(Grade D). Once the flexible bronchoscope is in the trachea,
the carina should be identified before advancing the
tracheal tube to minimise the risk of misplacement (Grade D).
The distance from the tracheal tube tip to the carina should
be confirmed as appropriate before removing the
bronchoscope (Grade D). On removal of the flexible
bronchoscope or videolaryngoscope, care must be taken to
maintain the correct position of the tracheal tube. The tip of
the bronchoscope should be in the neutral position and the
tracheal tube held firmly in position (Grade D). The tracheal
tube cuff can be gently inflated before, during or after
induction of anaesthesia. The decision around timing of cuff
inflation should be guided by the relative risks of aspiration,
patient movement, coughing and tracheal tube
displacement (Grade D). If there is suspicion of a cuff tear,
gentle inflation of the cuff to check integrity before
induction of anaesthesia is recommended (Grade D).
Ahmad et al. | Awake tracheal intubation guidelines Anaesthesia 2020, 75, 509–528

Table 3 Characteristics of drugs used commonly during ATI.

Duration Terminal
of elimination
Class Drug Onset action half-life Dosing Notes
Antisialogogue Glycopyrronium 20 min (i.m.) 30–60 min 40–80 min 0.2–0.4 mg Administer 30–60 min
bromide pre-procedure
3–5 min (i.v.) 30–60 min 40–80 min 0.1–0.2 mg May produce significant
tachycardia
Atropine 20 min (i.m.) 30–60 min 2h 0.3–0.6 mg Administer 30–60 min
pre-procedure – less
commonly used than
glycopyrronium bromide due
to tachycardia
2–3 min (i.v.) 30–60 min 2h 0.2–0.3 mg May produce significant
tachycardia
Hyoscine 30 min (i.m.) 4h 5h 0.2–0.6 mg Administer 30–60 min
hydrobromide 5–10 min (i.v.) pre-procedure
Longer lasting systemic effects
than glycopyrronium
bromide and atropine
May produce tachycardia,
dizziness and sedation
Topical Co-phenylcaine 2–5 min 30 min 1.5–2 h Lidocaine 1 bottle = 2.5 ml of lidocaine
anaesthesia spray 125 mg 50 mg.ml 1 and
Phenylephrine phenylephrine 5 mg.ml 1
12.5 mg
Lidocaine 1–10% 5 min 30–60 min 1.5–2 h Total dose not 1 ml of 1% = 10 mg
> 9 mg.kg 1 LBW 1 spray of 10% = 10 mg
1
Cocaine 10% 1–3 min 30–60 min 1h < 1.5 mg.kg LD50 1.2 g, but significant
toxic effects have been
reported at doses as low as
20 mg in adults
Particular care in older patients
and/or those with cardiac
disease
Sedatives Propofol 30 s 5–10 min 1.5–3 h TCI (effect-site) Caution with doses in excess of
0.5–1 lg.ml 1 1.5 lg.ml 1: risk of over-
sedation and hypoventilation,
particularly with concomitant
opioid use
Avoid bolus dosing
Midazolam 3–5 min 1–2 h 1.5–3 h Bolus 0.5–1 mg Titrate to effect
Peak effect at 5–10 min so care
with multiple doses
Dexmedeto 1–2 min 5–10 min 2h Bolus 0.5–1 lg.kg 1 Caution with bolus dosing as
midine over 5 min associated with hypertension
followed by infusion and bradycardia
(0.3–0.6 lg.kg 1.h 1)
Analgesia Remifentanil 1 min 3–5 min 1–20 min TCI (effect-site) Caution with respiratory
1–3 ng.ml 1 depression.
Avoid bolus dosing.
Fentanyl 2–5 min 30–60 min 6–10 min Bolus 0.5–1 lg.kg 1,
subsequent
doses of 0.5 lg.kg 1
as required
Alfentanil 2–3 min 15 min 90–120 min Bolus 5 lg.kg 1,
subsequent doses
of 1–3 lg.kg 1
as required

ATI, awake tracheal intubation; i.m., intramuscular; i.v., intravenous; TCI, target-controlled infusion; LD50, median lethal dose; LBW, lean
body weight.

© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists 517
Anaesthesia 2020, 75, 509–528 Ahmad et al. | Awake tracheal intubation guidelines

Table 4 Special circumstances that may affect standard performance of ATI with suggested management options.
Special
circumstance Considerations Modification Potential management options

Critically ill Limited physiological reserve and greater adverse Sedation Avoid or minimise sedation
consequences associated with sedation
Higher risk of local anaesthetic systemic toxicity Topicalisation Cautious use of local anaesthetic
Increased secretions Suction airway before instrumentation
Increased oxygen demand and reduced oxygen Oxygenation Supplemental oxygen essential
reserves
Unstable for transfer to operating theatre Performance Do not transfer patient out of critical care settings
Maintain same standards of equipment and
monitoring
Time-critical performance of ATI
Early consideration for high-risk general anaesthesia
Obstetrics Fetal sedation with benzodiazepines, long-acting Sedation Sedation with dexmedetomidine or remifentanil
opioids or propofol Warn neonatologists
Higher risk of local anaesthetic systemic toxicity; Topicalisation Cautious dosing of local anaesthetic; consider using
concomitant use of local anaesthetics via epidural pre-pregnancy body weight for dosing
analgesia
Increased oxygen demand and reduced oxygen Oxygenation Supplemental oxygen essential
reserves
Increased upper airway oedema and perfusion thus Performance Oral approach to ATI
increasing risk of nasal haemorrhage Identify and mark cricothyroid membrane early
FONA more difficult Airway ultrasound to identify cricothyroid
membrane
Obesity Critical adverse consequences of over-sedation Sedation Avoid or minimise sedation
Risk of local anaesthetic overdose Topicalisation Local anaesthetic dosing on lean body weight
Increased oxygen demand and reduced oxygen Oxygenation Supplemental oxygen essential
reserves
Diaphragmatic splinting and reduced functional Performance Sitting position or reverse Trendelenburg
residual capacity Operator facing patient
FONA more difficult Identify and mark cricothyroid membrane early
Airway ultrasound to identify cricothyroid
membrane
Trauma Critical adverse consequences of over-sedation Sedation Avoid or minimise sedation
Difficult administration due to airway soiling Topicalisation Clear soiled airway before topicalisation
Increased oxygen demand and reduced oxygen Oxygenation Supplemental oxygen essential
reserves
Unstable for transfer to operating theatre Performance Do not transfer patient out of critical care settings
Maintain same standards of equipment and
monitoring
Airway soiling from haemorrhage, secretions, vomitus ATI: VL
and tissue oedema Tracheal intubation via SAD
Suspected base of skull or facial fracture Avoid HFNO
Oral approach to ATI
Trismus Critical adverse consequences of over-sedation Sedation Avoid or minimise sedation
Limited pharyngeal access Topicalisation Nebulised lidocaine
Spray-as-you-go
Transtracheal lidocaine injection
Insertion of mucosal atomiser and patient gargling
Potentially increased oxygen demand Oxygenation Supplemental oxygen essential
Limited mouth opening Performance Nasal approach to ATI: FB

(continued)

518 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Ahmad et al. | Awake tracheal intubation guidelines
Table 4 (continued)
Special
circumstance Considerations
Stridor Critical adverse consequences of over-sedation
Risk of laryngospasm
Airway obstruction
Narrowed airway
ATI, awake tracheal intubation; VL, videolaryngoscopy; SAD, supraglottic airway device; HFNO, high-flow nasal oxygen; FB, flexible
bronchoscopy; FONA, front-of-neck airway.
Difficult Airway Society ATI technique
An example of a practical approach to the sTOP ATI
technique is shown in Fig. 2. This technique has been
specifically considered for simplicity and generalisability.
We recognise that there are a range of different techniques
and regimens which also address the key sTOP components
that will be equally effective.
Special circumstances
Specific patient pathophysiology may dictate modifications
to the performance of ATI that must be considered and
planned for. As with all other aspects of ATI, these
modifications can be categorised based on sTOP. Examples
of suggested changes to technique are presented in
Table 4.
Managing complications
The reported overall complication rate in patients
undergoing ATI either flexible bronchoscopic
videolaryngoscopic, is up to 18% [33–35, 58, 144–146].
Complications during ATI occur due to inadequate sTOP. In
the event of a complication, its aetiology should be
determined and appropriately managed (Grade D; Fig. 3).
We define an unsuccessful attempt at ATI as the
unplanned removal of flexible bronchoscope,
videolaryngoscope or tracheal tube from the airway.
Patients in whom ATI is indicated are at greater risk of the
adverse consequences of multiple attempts, such as airway
trauma, airway obstruction, bleeding and unsuccessful ATI
[1]. It is therefore advisable to minimise the number of
attempts at ATI (Grade D). Operators should consider if they
require more experienced support before commencing ATI
(Grade D). Operators should ensure sTOP is optimised
before the first attempt (Grade D). If unsuccessful with the
first attempt, operators should re-assess, correct any
© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Anaesthesia 2020, 75, 509–528
Modification Potential management options
Sedation Avoid or minimise sedation
Topicalisation Consider nebulised and/or lower concentrations of
lidocaine
Oxygenation HFNO highly recommended
Performance Recognise that airway narrowing may preclude oral
or nasal tracheal intubation
Prime for emergency FONA
Use smaller tracheal tube
Most experienced practitioner to perform
May require combined technique
inadequate sTOP components, and call for help before
proceeding with a second attempt (Grade D). If
unsuccessful with the second attempt, a third may be
considered only if conditions can be further optimised
(Grade D). A fourth and final attempt (3 + 1) should only be
undertaken by a more experienced operator, which may
include a surgeon (Grade D). Each attempt subsequent to
the first should involve a change in the elements of
performance to improve the likelihood of success (Grade D).
The use of an alternative device (e.g. FB to VL or vice versa)
should be counted in the total number of attempts. Each
failed attempt may adversely affect patient and operator
confidence. Seeking expert help at the earliest opportunity
is recommended (Grade D). If unsuccessful after 3 + 1
attempts, the unsuccessful ATI algorithm should be
followed (Grade D; Fig. 4).
Management of unsuccessful ATI
or The unsuccessful ATI algorithm is a guide for the rare
occasions where successful tracheal intubation has not
been achieved in 3 + 1 attempts. Immediate actions should
include a call for help, ensuring 100% oxygen is applied and
stopping (if necessary, reversing) any sedative drugs (Grade
D). Operators should ‘stop and think’ to determine
subsequent airway management, while also ‘priming’ for
emergency FONA [2, 52] (Grade D). The default action in
the event of unsuccessful ATI should be to postpone the
procedure (Grade D). Operators should only proceed with
immediate airway management if essential (e.g. if airway
patency, ventilation or neurology is compromised; urgent
or immediate surgery is required; or clinical deterioration is
expected) (Grade D).
If airway management is deemed essential, the
preferred option for securing the airway after unsuccessful
ATI:VL or ATI:FB should be ATI using FONA (ATI:FONA),
519
Anaesthesia 2020, 75, 509–528 Ahmad et al. | Awake tracheal intubation guidelines

Figure 3 Managing procedural complications during awake tracheal intubation (ATI). This provides a framework for managing
complications, but is not meant to be a comprehensive guide. This figure forms part of the Difficult Airway Society guidelines for
ATI in adults and should be used in conjunction with the text. FIO2, fractional inspired concentration of oxygen; O2, oxygen
©Difficult Airway Society 2019.

which includes cricothyroidotomy or tracheostomy (Grade D).
The most appropriately skilled clinician available should
perform this (Grade C). The considerations for the
appropriateness of ATI:FONA include: patient factors; skill;
and equipment availability. If inappropriate or unsuccessful,
a high-risk general anaesthetic is the only remaining option.
In this scenario, the operator should formulate an
achievable A to D airway management strategy informed by
the unsuccessful attempts at ATI and based on the 2015
DAS guidelines [2], recognising that they are primarily for
the unanticipated difficult tracheal intubation (Grade D).
This strategy should include an i.v. induction of anaesthesia
with full neuromuscular blockade [2, 52] (Grade D).
Videolaryngoscopes may improve tracheal intubation
success rates in cases of difficult tracheal intubation [147];
therefore, the first attempt at tracheal intubation in this
scenario should be with a videolaryngoscope (Grade A). All
attempts with any device should be performed by the most
appropriately skilled clinician present (Grade C).
Post-tracheal intubation management
Patients who have had ATI due to predicted difficult airway
management are at high risk of complications at tracheal
extubation [1, 148], and require an appropriate tracheal
extubation strategy. Planning, preparation, performing and
post-tracheal extubation care should follow DAS guidelines
[148] (Grade D).
Before tracheal extubation, laryngoscopy, either with a
direct laryngoscope or videolaryngoscope, may provide
useful information for risk stratification of tracheal
extubation and any subsequent airway management. The
view at laryngoscopy may be altered by the presence of a

520 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Ahmad et al. | Awake tracheal intubation guidelines Anaesthesia 2020, 75, 509–528

Figure 4 The Difficult Airway Society management of unsuccessful awake tracheal intubation (ATI) in adults. This algorithm
forms part of the Difficult Airway Society guidelines for ATI in adults and should be used in conjunction with the text. HFNO,
high-flow nasal oxygen; SAD, supraglottic airway device; FONA, front-of-neck airway; GA, general anaesthesia. ©Difficult Airway
Society 2019.

tracheal tube [148]. Therefore, verification of laryngoscopy lidocaine is up to 2 h, patients should remain nil by mouth
grade may rule out, but not rule in, easy subsequent asleep for at least 2 h following airway topicalisation for ATI (Grade
tracheal intubation. D).
Topical lidocaine has a dose-dependent duration of Documentation of ATI in clinical records is necessary to
analgesic action of up to 40 min although this may vary with inform and guide future patient management [1, 152]. This
concentration and method of administration [149, 150]. should include: documentation of oxygenation;
However, the time to return of laryngeal reflexes can be topicalisation; sedation strategy; device and tracheal tube
longer [151]. Given that the terminal elimination half-life of used; approach (e.g. right nasal, left nasal, oral); number of

© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists 521
Anaesthesia 2020, 75, 509–528 Ahmad et al. | Awake tracheal intubation guidelines

attempts; and any complications or notes (Grade D;
Supporting Information, Appendix S3).
Consent
Clinicians should adhere to the Association
Anaesthetists’ guidelines on consent for anaesthesia [153]
(Grade D). Informed consent must be taken, with patients
being given information (ideally including a patient
information leaflet [154]) in a timely manner (Grade D).
The risks of ATI and its alternative (induction of general
anaesthesia before securing the airway) should be
discussed (Table 5) (Grade D). Appropriate explanation is
vital and a good rapport can increase the confidence and
co-operation of the patient in the procedure and is
strongly encouraged (Grade D). The consent process
should be documented (Grade D).
Training
Successful performance of ATI has been shown to be
independent of seniority, but related to experience [33].
There are many strategies used for training in the technical
aspects of ATI, including the use of manikins, simulators,
cadavers and patients [59, 155–161]. All anaesthetists
should seek every opportunity to attain and maintain skills in
ATI and all departments should support this [1] (Grade C).
Awake tracheal intubation is a skill in the compulsory higher
training curriculum of the Royal College of Anaesthetists
[162], but opportunities for training are known to be limited
[163–167]. These guidelines provide a common stem for
sedation, topicalisation, oxygenation and performance to
encourage training in ATI. Experience using a range of tools
should be sought, complementing active clinical practice to
develop non-technical aspects of ATI (Grade B). We
recognise this may be difficult to achieve [163–167], but
local hospital airway leads are ideally placed to facilitate
training and provision of ATI skills and equipment. Team
training in ATI should include anaesthetic assistants,
operating department practitioners and theatre staff.
of Awake tracheal intubation may be performed solely for the
purposes of training provided appropriate consent is taken
(Grade D).
Future directions
These guidelines highlight the paucity of high-quality
evidence in ATI, as demonstrated by the need for expert
opinion for the majority of recommendations. This presents
an opportunity for further research to be undertaken to
improve both clinical and patient-centred outcomes [168]. In
particular, the ideal topicalisation and sedation strategies are
yet to be elucidated, with a limited evidence base for
individual drugs, administration methods (e.g. infusion vs.
bolus, combinations of sedatives, mucosal atomiser vs.
nebulisation) and their related outcomes [129]. There
remains uncertainty regarding many aspects of procedural
performance such as ideal patient and operator positioning,
the role of checklists and cognitive aids and immediate
management of complications. Moreover, training in ATI has
thus far focussed on technical aspects, primarily with FB, but
training with alternative devices and non-technical skills have
had little attention in the published literature and warrant
further investigation. Novel technology for ATI must also be
developed, such as improved capnography and monitoring,
safer sedation delivery devices and better image visualisation
and guidance technology. Finally, the impact these
guidelines have on clinical practice should be examined to
allow further iterations to be improved upon. This will require
updates of these guidelines using similar methodology when
a more robust evidence base becomes available.

Table 5 Incidence of complications when asleep or awake tracheal intubation is performed. The rates reported for asleep
tracheal intubation include data for all patients, and patients who are predicted to have difficult airway management. The rates
reported for awake tracheal intubation are only for patients who are predicted to be at risk of difficult airway management.
Asleep tracheal intubation
All patients Predicted difficult tracheal intubation Awake tracheal intubation
Difficult facemask ventilation 2.2–2.5% 18.6–22% Not applicable
Impossible facemask ventilation 0.15% Not currently available Not applicable
Difficult tracheal intubation 1.9–10% 25% Not applicable
Failed tracheal intubation 0.15% 0.36%a 1–2%
a
CICO 0.04% 0.75% 0–0.06%
Front-of-neck airway 0.002–0.07% 0.12%a 0–0.38%
Death 0.0006–0.04% Not currently available Not currently available
CICO, cannot intubate, cannot oxygenate.
a
Unpublished data from the Danish Anaesthesia Database.

522 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Ahmad et al. | Awake tracheal intubation guidelines
Discussion
The primary aims of these guidelines are to provide
practitioners with a comprehensive document on ATI. These
guidelines should support clinical practice and lower the
performance threshold thereby increasing the use of ATI
when indicated. The quality of evidence supporting many
recommendations is limited, with interventions and
outcomes being highly heterogeneous. This is likely
influenced by the fact that ATI can be successfully
performed in a wide range of settings and patients with
varying techniques [129]. For example, the use of SADs as a
conduit to ATI or optical stylets in awake patients have not
been well-described but warrant future investigation as
their role becomes more defined. Similarly, the lack of
previously published specific guidelines means that
research in ATI is disparate and inconsistent [169]. However,
we have sought and appraised the available evidence in
ATI, and in its absence we have incorporated the practical
and theoretical experience of international experts. We
have involved patients, DAS members and international
experts in order to further understand current practice and
the need for these guidelines. Formal resource implication
analysis has not been conducted; however, the tools to
practically perform ATI are available widely, and thus we
expect the resource impact to be modest.
These guidelines prioritise patient safety and provide
recommendations for best clinical practice. It is hoped that
they will lead to a paradigm shift in clinical practice and
improve the care of patients with predicted difficult airway
management in the UK and beyond.
Acknowledgements
The systematic review was registered at PROSPERO
(registration ID CRD42017072707). We thank Ms M. Hillier
(librarian, UK) for her assistance with the literature search. We
thank Mr A. Diver (anaesthetic assistant, UK), for his
contribution to the development of the recommendations.
We thank Mr A. Fry (consultant surgeon, UK), Mr R. Oakley
(consultant surgeon, UK), Mrs I. Anastasescu (anaesthetic
nurse, UK) and Mrs E. Jacovou (anaesthetic nurse, UK) for
their independent review of the manuscript. For their
contribution to the expert surveys and/or manuscript review,
we thank Prof. M. Aziz (USA), Dr P. Baker (New Zealand), Dr E.
Burdett (UK), Dr S. Charters (UK), Dr N. Chrimes (Australia),
Dr S. Clarke (UK), Professor T. Cook (UK), Professor R. Cooper
(Canada), Professor P. Diemunsch (France), Dr J. Doyle
(UAE), Professor C. Frerk (UK), Professor K. Greenland
(Australia), Professor R. Greif (Switzerland), Dr P. Groom (UK),
Professor T. Heidegger (Austria), Dr A. Higgs (UK), Dr E.
Hodgson (South Africa), Dr R. Hoffmeyer (South Africa), Dr J.
© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Anaesthesia 2020, 75, 509–528
Huitink (The Netherlands), Dr F. Kelly (UK), Dr M. Kristensen
(Denmark), Professor J. A. Law (Canada), Dr B. McGuire (UK),
Professor C. Mendonca (UK), Professor M. Mushambi (UK),
Professor S. Myatra (India), Dr R. Coloma Navarro (Chile),
Professor V. Nekhendzy (USA), Dr H. Osses (Chile), Professor
J. Pandit (UK), Dr B. Patel (UK), Professor W. Rosenblatt (USA),
Dr N. Shallik (Qatar), Professor A. Smith (UK), Dr M. Sorbello
(Italy) and Dr N. Woodall (UK). We thank Drs A. Nørskov,
C. Rosenstock and L. Lundstrøm for data provided from the
Danish Anaesthesia Database in Table 5. Costs related to
Guideline Group meetings and graphic design were met by
DAS. IA has previously received honoraria for consulting for
Ambu, honoraria and funding for travel and accommodation
from Fisher & Paykel Healthcare, Ambu and Verathon
Medical to give lectures at international meetings. KE is an
Editor for Anaesthesia and this manuscript underwent
external review. KE has previously received honoraria for
consulting for Ambu. RB has received products for
departmental and workshop use by Ambu, Armstrong, Cook
Medical, Fisher & Paykel Healthcare, Karl Storz, PROACT and
Teleflex. AM has received equipment for evaluation and
teaching (including the running of workshops) from Ambu,
Cook Medical, Fisher & Paykel Healthcare, Medtronic, Karl
Storz, Teleflex, VBM/Freelance. AM has acted as an advisor
to the Medicines and Healthcare products Regulatory
Agency (MHRA) and participated on an advisory board and
speaker panel for Medtronic. He has received travel
expenses for teaching sessions from Fisher & Paykel
Healthcare. IH has been given trial products for clinical use
and evaluation from Ambu, Cook Medical, Storz, Verathon,
Venner Medical and Fannin. IH has also received funding for
travel and accommodation to give lectures from Covidien
and has received equipment to conduct airway workshops
from Storz, Ambu, Verathon and Fannin. AP has helped to
develop a videolaryngoscope. He has received travel,
consulting and research support from Fisher & Paykel
Healthcare. FM has received funding for travel and
accommodation from Fisher & Paykel Healthcare. There has
been no involvement of any industry in any aspect of this
project. No external funding or competing interest declared.
‘[Correction added on 20 February 2020, after first
online publication: In the Acknowledgements section, Mr A.
Diver’s name was previously wrong and has been corrected
in this version.]
References
1. Cook TM, Woodall NM, Frerk CM. 4th National Audit Project of
the Royal College of Anaesthetists and the Difficult Airway
Society. British Journal of Anaesthesia 2011; 106: 617–31.
2. Frerk C, Mitchell VSS, McNarry AFF, et al. Difficult Airway
Society 2015 guidelines for management of unanticipated
523
Anaesthesia 2020, 75, 509–528
difficult intubation in adults. British Journal of Anaesthesia
2015; 115: 827–48.
3. Mushambi MC, Kinsella SM, Popat M, et al. Obstetric
Anaesthetists’ Association and Difficult Airway Society
guidelines for the management of difficult and failed tracheal
intubation in obstetrics. Anaesthesia 2015; 70: 1286–306.
4. Petrini F, Accorsi A, Adrario E, et al. Recommendations for
airway control and difficult airway management. Minerva
Anestesiologica 2005; 71: 617–57.
5. Australian and New Zealand College of Anaesthetists.
Guidelines for the management of evolving airway
obstruction: transition to the can’t intubate can’t oxygenate
airway emergency. Australian and New Zealand College of
Anaesthetists (ANZCA) 2016.; 2016.
6. Law JA, Broemling N, Cooper RM, et al. The difficult airway
with recommendations for management – Part 1 – Intubation
encountered in an unconscious/induced patient. Canadian
Journal of Anesthesia 2013; 60: 1089–118.
7. Apfelbaum JL, Hagberg CA, Caplan RA, et al. Practice
guidelines for management of the difficult airway: an updated
report by the American Society of Anesthesiologists Task
Force on Management of the Difficult Airway. Anesthesiology
2013; 118: 251–70.
8. Myatra SN, Shah A, Kundra P, et al. All India Difficult Airway
Association 2016 guidelines for the management of
unanticipated difficult tracheal intubation in adults. Indian
Journal of Anaesthesia 2016; 60: 885–98.
9. Ramkumar V, Dinesh E, Shetty SR, et al. All India difficult airway
association 2016 guidelines for the management of
unanticipated difficult tracheal intubation in obstetrics. Indian
Journal of Anaesthesia 2016; 60: 899–905.
10. Langeron O, Bourgain J-LL, Francon D, et al. Difficult
intubation and extubation in adult anaesthesia. Anaesthesia
Critical Care and Pain Medicine 2018; 37: 639–51.
11. Rehn M, Hyldmo PK, Magnusson V, et al. Scandinavian SSAI
clinical practice guideline on pre-hospital airway
management. Acta Anaesthesiologica Scandinavica 2016; 60:
852–64.
12. Piepho T, Cavus E, Noppens R, et al. S1-Leitlinie
Atemwegsmanagement: Leitlinie der Deutschen Gesellschaft
f€
ur An€asthesiologie und Intensivmedizin (DGAI). Anaesthesist
2015; 64: 27–40.
13. Anon JSA. airway management guideline 2014: to improve
the safety of induction of anesthesia. Journal of Anaesthesia
2014; 28: 482–93.
14. Nørskov AK, Rosenstock CV, Wetterslev J, Astrup G, Afshari
A, Lundstrøm LH. Diagnostic accuracy of anaesthesiologists’
prediction of difficult airway management in daily clinical
practice: a cohort study of 188 064 patients registered in
the Danish Anaesthesia Database. Anaesthesia 2015; 70:
272–81.
15. Nørskov AK, Wetterslev J, Rosenstock CV, et al. Prediction of
difficult mask ventilation using a systematic assessment of risk
factors vs. existing practice – a cluster randomised clinical trial
in 94,006 patients. Anaesthesia 2017; 72: 296–308.
16. Kheterpal S, Healy D, Aziz MF, et al. Incidence, predictors, and
outcome of difficult mask ventilation combined with difficult
laryngoscopy: a report from the multicenter perioperative
outcomes group. Anesthesiology 2013; 119: 1360–9.
17. Kheterpal S, Martin L, Shanks AM, Tremper KK. Prediction and
outcomes of impossible mask ventilation: a review of 50,000
anesthetics. Anesthesiology 2009; 110: 891–7.
18. Verghese C, Brimacombe JR. Survey of laryngeal mask airway
usage in 11,910 patients: safety and efficacy for conventional
and non-conventional usage. Anesthesia and Analgesia 1996;
82: 129–33.
19. Saito T, Liu W, Chew STH, Ti LK. Incidence of and risk factors
for difficult ventilation via a supraglottic airway device in a
524 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Ahmad et al. | Awake tracheal intubation guidelines
population of 14,480 patients from South-East Asia.
Anaesthesia 2015; 70: 1079–83.
20. Francksen H, Renner J, Hanss R, Scholz J, Doerges V, Bein B. A
comparison of the i-gel with the LMA-Unique in non-paralysed
anaesthetised adult patients. Anaesthesia 2009; 64: 1118–24.
21. Rose DK, Cohen MM. The airway: problems and predictions in
18,500 patients. Canadian Journal of Anesthesia 1994; 41:
372–83.
22. Cook TM, Tr€ umpelmann P, Beringer R, Stedeford J. A
randomised comparison of the Portex SoftsealTM laryngeal
mask airway with the LMA-UniqueTM during anaesthesia.
Anaesthesia 2005; 60: 1218–25.
23. Shiga T, Wajima Z, Inoue T, Sakamoto A. Predicting difficult
intubation in apparently normal patients: a meta-analysis of
bedside screening test performance. Anesthesiology 2005;
103: 429–37.
24. Lundstrøm LH, Møller AM, Rosenstock C, Astrup G, G€ atke MR,
Wetterslev J. A documented previous difficult tracheal
intubation as a prognostic test for a subsequent difficult
tracheal intubation in adults. Anaesthesia 2009; 64: 1081–8.
25. Detsky ME, Jivraj N, Adhikari NK, et al. Will this patient be
difficult to intubate? the rational clinical examination
systematic review. Journal of the American Medical
Association 2019; 321: 493–503.
26. Thomsen JLD, Nørskov AK, Rosenstock CV. Supraglottic
airway devices in difficult airway management: a retrospective
cohort study of 658,104 general anaesthetics registered in the
Danish Anaesthesia Database. Anaesthesia 2019; 74: 151–7.
27. Rosenstock CV, Nørskov AK, Wetterslev J, Lundstrøm LH.
Danish Anaesthesia Database. Emergency surgical airway
management in Denmark: a cohort study of 452,461 patients
registered in the Danish Anaesthesia Database. British Journal
of Anaesthesia 2016; 117(Suppl): i75–82.
28. Huitink JM, Lie PP, Heideman I, et al. A prospective, cohort
evaluation of major and minor airway management
complications during routine anaesthetic care at an academic
medical centre. Anaesthesia 2017; 72: 42–8.
29. Edelman DA, Perkins EJ, Brewster DJ. Difficult airway
management algorithms: a directed review. Anaesthesia
2019; 74: 1175–85.
30. Cook TM. Strategies for the prevention of airway complications
– a narrative review. Anaesthesia 2018; 73: 93–111.
31. Alhomary M, Ramadan E, Curran E, Walsh SR.
Videolaryngoscopy vs. fibreoptic bronchoscopy for awake
tracheal intubation: a systematic review and meta-analysis.
Anaesthesia 2018; 73: 1151–61.
32. Ajay S, Singhania A, Akkara AG, Shah A, Adalja M. A study of
flexible fiberoptic bronchoscopy aided tracheal intubation for
patients undergoing elective surgery under general
anesthesia. Indian Journal of Otolaryngology and Head and
Neck Surgery 2013; 65: 116–19.
33. El-Boghdadly K, Onwochei DN, Cuddihy J, et al. A
prospective cohort study of awake fibreoptic intubation
practice at a tertiary centre. Anaesthesia 2017; 72: 694–703.
34. Law JA, Morris IR, Brousseau PA, de la Ronde S, Milne AD. The
incidence, success rate, and complications of awake tracheal
intubation in 1,554 patients over 12 years: an historical cohort
study. Canadian Journal of Anesthesia 2015; 62: 736–44.
35. Joseph TT, Gal JS, DeMaria SJ, Lin H-M, Levine AI, Hyman JB. A
retrospective study of success, failure, and time needed to
perform awake intubation. Anesthesiology 2016; 125: 105–14.
36. Brouwers MC, Kerkvliet K, Spithoff K, Consortium ANS. The
AGREE Reporting Checklist: a tool to improve reporting of
clinical practice guidelines. British Medical Journal 2016; 352:
i1152.
37. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, John
PA. The PRISMA statement for reporting systematic reviews
and meta-analyses of studies that evaluate healthcare
Ahmad et al. | Awake tracheal intubation guidelines
interventions: explanation and elaboration. British Medical
Journal 2009; 339: b2700.
38. Boulkedid R, Abdoul H, Loustau M, Sibony O, Alberti C. Using
and reporting the Delphi method for selecting healthcare
quality indicators: a systematic review. Wright JM, ed. PLoS
ONE 2011; 6: e20476.
39. Diamond IR, Grant RC, Feldman BM, et al. Defining
consensus: a systematic review recommends methodologic
criteria for reporting of Delphi studies. Journal of Clinical
Epidemiology 2014; 67: 401–9.
40. Pandit JJ, Popat MT, Cook TM, et al. The Difficult Airway
Society “ADEPT” Guidance on selecting airway devices: the
basis of a strategy for equipment evaluation. Anaesthesia
2011; 66: 726–37.
41. Du Rand IA, Blaikley J, Booton R, et al. Summary of the British
Thoracic Society guideline for diagnostic flexible
bronchoscopy in adults. Thorax 2013; 68: 786–7.
42. Roth D, Pace NL, Lee A, et al. Airway physical examination
tests for detection of difficult airway management in
apparently normal adult patients. Cochrane Database of
Systematic Reviews 2018; 2018: CD008874.
43. Benumof JL. Obstructive sleep apnea in the adult obese
patient: implications for airway management. Anesthesiology
Clinics of North America 2002; 20: 789–811.
44. Petrini F, Di Giacinto I, Cataldo R, et al. Perioperative and
periprocedural airway management and respiratory safety for
the obese patient: 2016 SIAARTI Consensus. Minerva
Anestesiologica 2016; 82: 1314–35.
45. Law JA, Broemling N, Cooper RM, et al. The difficult airway
with recommendations for management – Part 2 – Difficult
airway. Canadian Journal of Anesthesia 2013; 60: 1119–38.
46. Weinger MB, Vredenburgh AG, Schumann CM, et al.
Quantitative description of the workload associated with
airway management procedures. Journal of Clinical
Anaesthesia 2000; 12: 273–82.
47. Reason J. Human error: models and management. British
Medical Journal 2000; 320: 768–70.
48. Rall M, Dieckmann P. Safety culture and crisis resource
management in airway management: general principles to
enhance patient safety in critical airway situations. Best
Practice and Research: Clinical Anesthesiology 2005; 19:
539–57.
49. Smith AF, Mishra K. Interaction between anaesthetists, their
patients, and the anaesthesia team. British Journal of
Anaesthesia 2010; 105: 60–8.
50. Flin R, Fioratou E, Frerk C, Trotter C, Cook TM. Human factors
in the development of complications of airway management:
preliminary evaluation of an interview tool. Anaesthesia 2013;
68: 817–25.
51. Davis M, Hignett S, Hillier S, Hames N, Hodder S. Safer
anaesthetic rooms: human factors/ergonomics analysis of work
practices. Journal of Perioperative Practice 2016; 26: 274–80.
52. Higgs A, McGrath BA, Goddard C, et al. Guidelines for the
management of tracheal intubation in critically ill adults.
British Journal of Anaesthesia 2018; 120: 323–52.
53. Eichhorn JH, Cooper JB, Cullen DJ, Maier WR, Philip JH,
Seeman RG. Standards for patient monitoring during
anesthesia at Harvard Medical School. Journal of American
Medical Association 1986; 256: 1017–20.
54. Webb RK, Currie M, Morgan CA, et al. The Australian incident
monitoring study: an analysis of 2000 incident reports.
Anaesthesia and Intensive Care 1993; 21: 520–8.
55. Williamson JA, Webb RK, Sellen A, Runciman WB, Van der
Walt JH. The Australian incident monitoring study. Human
failure: an analysis of 2000 incident reports. Anaesthesia and
Intensive Care 1993; 21: 678–83.
56. Keenan RL, Boyan CP. Decreasing frequency of anesthetic
cardiac arrests. Journal of Clinical Anaesthesia 1991; 3: 354–7.
© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Anaesthesia 2020, 75, 509–528
57. McKay WPS, Noble WH. Critical incidents detected by pulse
oximetry during anaesthesia. Canadian Journal of Anesthesia
1988; 35: 265–9.
58. Rosenstock CV, Thogersen B, Afshari A, Christensen AL,
Eriksen C, Gatke MR. Awake fiberoptic or awake video
laryngoscopic tracheal intubation in patients with anticipated
difficult airway management: a randomized clinical trial.
Anesthesiology 2012; 116: 1210–16.
59. Woodall NM, Harwood RJ, Barker GL, et al. Complications of
awake fibreoptic intubation without sedation in 200 healthy
anaesthetists attending a training course. British Journal of
Anaesthesia 2008; 100: 850–5.
60. Patil V, Barker GL, Harwood RJ, Woodall NM. Training course
in local anaesthesia of the airway and fibreoptic intubation
using course delegates as subjects. British Journal of
Anaesthesia 2002; 89: 586–93.
61. Sun Y, Liu J-X, Jiang H, Zhu Y-S, Xu H, Huang Y. Cardiovascular
responses and airway complications following awake nasal
intubation with blind intubation device and fibreoptic
bronchoscope: a randomized controlled study. European
Journal of Anesthesiology 2010; 27: 461–7.
62. Checketts MR, Alladi R, Ferguson K, et al. Recommendations
for standards of monitoring during anaesthesia and recovery
2015: association of Anaesthetists. Anaesthesia 2016; 71: 85–
93.
63. Carayon P, Xie A, Kianfar S. Human factors and ergonomics as
a patient safety practice. British Medical Journal Quality and
Safety 2014; 23: 196–205.
64. Ovassapian A, Tuncbilek M, Weitzel EK, Joshi CW. Airway
management in adult patients with deep neck infections: a
case series and review of the literature. Anesthesia and
Analgesia 2005; 100: 585–9.
65. Meghjee SPL, Marshall M, Redfern EJ, McGivern DV. Influence
of patient posture on oxygen saturation during fibre-optic
bronchoscopy. Respiratory Medicine 2001; 95: 5–8.
66. van Zwam JP, Kapteijns EFG, Lahey S, Smit HJM. Flexible
bronchoscopy in supine or sitting position. Journal of
Bronchology and Interventional Pulmonology 2010; 17: 29–
32.
67. Ling IT, Piccolo F, Mulrennan SA, Phillips MJ. Posture
influences patient cough rate, sedative requirement and
comfort during bronchoscopy: an observational cohort study.
Cough 2011; 7.
68. Altermatt FR, Mu~ noz HR, Delfino AE, Cortınez LI. Pre-
oxygenation in the obese patient: effects of position on
tolerance to apnoea. British Journal of Anaesthesia 2005; 95:
706–9.
69. Isono S, Ishikawa T, Nishino T, Tanaka A, Tagaito Y. Sitting
posture decreases collapsibility of the passive pharynx in
anesthetized paralyzed patients with obstructive sleep apnea.
Anesthesiology 2010; 113: 812–18.
70. Ikeda H, Ayuse T, Oi K. The effects of head and body
positioning on upper airway collapsibility in normal subjects
who received midazolam sedation. Journal of Clinical
Anesthesia 2006; 18: 185–93.
71. Heidegger T, Gerig HJ, Ulrich B, Schnider TW. Structure and
process quality illustrated by fibreoptic intubation: analysis of
1612 cases. Anaesthesia 2003; 58: 734–9.
72. Gomez-Rıos MA,  Nieto SL. Combined use of an Airtraqâ
optical laryngoscope, Airtraq video camera, Airtraq wireless
monitor, and a fibreoptic bronchoscope after failed tracheal
intubation. Canadian Journal of Anesthesia 2011; 58: 411–
12.
73. Chung MY, Park B, Seo J, Kim CJ. Successful airway
management with combined use of McGrath MAC video
laryngoscope and fiberoptic bronchoscope in a severe obese
patient with huge goiter -a case report. Korean Journal of
Anesthesiology 2018; 71: 232–6.
525
Anaesthesia 2020, 75, 509–528
74. Liew GHC, Wong TGL, Lu A, Kothandan H. Combined use of
the glidescope and flexible fibrescope as a rescue technique
in a difficult airway. Proceedings of Singapore Healthcare
2015; 24: 117–20.
75. Shiraishi T. Awake insertion of the air-QTM intubating laryngeal
airway device that facilitates safer tracheal intubation in
morbidly obese patients. British Journal of Anaesthesia 2013;
111: 1024–5.
76. Lim WY, Wong P. Awake supraglottic airway guided flexible
bronchoscopic intubation in patients with anticipated difficult
airways: a case series and narrative review. Korean Journal of
Anesthesiology 2019.
77. Kristensen MS, Fredensborg BB. The disposable Ambu
aScope vs. a conventional flexible videoscope for awake
intubation – a randomised study. Acta Anaesthesiologica
Scandinavica 2013; 57: 888–95.
78. Brull SJ, Wiklund R, Ferris C, Connelly NR, Ehrenwerth J,
Silverman DG. Facilitation of fiberoptic orotracheal intubation
with a flexible tracheal tube. Anesthesia and Analgesia 1994;
78: 746–8.
79. Hakala P, Randell T, Valli H. Comparison between tracheal
tubes for orotracheal fibreoptic intubation. British Journal of
Anaesthesia 1999; 82: 135–6.
80. Jafari A, Gharaei B, Reza Kamranmanesh M, et al. Wire
reinforced endotracheal tube compared with Parker Flex-Tip
tube for oral fiberoptic intubation: a randomized clinical trial.
Minerva Anestesiologica 2014; 80: 324–9.
81. Lucas DN, Yentis SM. A comparison of the intubating laryngeal
mask tracheal tube with a standard tracheal tube for fibreoptic
intubation. Anaesthesia 2000; 55: 358–61.
82. Sharma D, Bithal PK, Rath GP, Pandia MP. Effect of orientation
of a standard polyvinyl chloride tracheal tube on success rates
during awake flexible fibreoptic intubation. Anaesthesia 2006;
61: 845–8.
83. Kristensen MS. The Parker Flex-Tip tube versus a standard
tube for fiberoptic orotracheal intubation: a randomized
double-blind study. Anesthesiology 2003; 98: 354–8.
84. Rai MR, Scott SH, Marfin AG, Popat MT, Pandit JJ. A
comparison of a flexometallic tracheal tube with the
intubating laryngeal mask tracheal tube for nasotracheal
fibreoptic intubation using the two-scope technique.
Anaesthesia 2009; 64: 1303–6.
85. Barker KF, Bolton P, Cole S, Coe PA. Ease of laryngeal passage
during fibreoptic intubation: a comparison of three
endotracheal tubes. Acta Anaesthesiologica Scandinavica
2001; 45: 624–6.
86. Lomax SL, Johnston KD, Marfin AG, Yentis SM, Kathawaroo S,
Popat MT. Nasotracheal fibreoptic intubation: a randomised
controlled trial comparing the GlideRiteâ (Parker-Flexâ Tip)
nasal tracheal tube with a standard pre-rotated nasal RAETM
tracheal tube. Anaesthesia 2011; 66: 180–4.
87. Koga K, Asai T, Latto IP, Vaughan RS. Effect of the size of a
tracheal tube and the efficacy of the use of the laryngeal mask
for fibrescope-aided tracheal intubation. Anaesthesia 1997;
52: 512.
88. Arriaga AF, Bader AM, Wong JM, et al. Simulation-based trial
of surgical-crisis checklists. New England Journal of Medicine
2013; 368: 246–53.
89. Ziewacz JE, Arriaga AF, Bader AM, et al. Crisis checklists for
the operating room: development and pilot testing. Journal of
the American College of Surgeons 2011; 213: 212.e10–
217.e10.
90. Russ S, Rout S, Sevdalis N, Moorthy K, Darzi A, Vincent C. Do
safety checklists improve teamwork and communication in the
operating room? A systematic review Annals of Surgery 2013;
258: 856–71.
91. Haynes AB, Weiser TG, Berry WR, et al. A surgical safety
checklist to reduce morbidity and mortality in a global
526 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Ahmad et al. | Awake tracheal intubation guidelines
population. New England Journal of Medicine 2009; 360:
491–9.
92. Harvey R, Foulds L, Housden T, et al. The impact of didactic
read-aloud action cards on the performance of cannula
cricothyroidotomy in a simulated ‘can’t intubate can’t
oxygenate’ scenario. Anaesthesia 2017; 72: 343–9.
93. Marshall SD, Mehra R. The effects of a displayed cognitive aid
on non-technical skills in a simulated “can’t intubate, can’t
oxygenate” crisis. Anaesthesia 2014; 69: 669–77.
94. Sidhu VS, Whitehead EM, Ainsworth QP, Smith M, CalderI. A
technique of awake fibreoptic intubation. Experience in patients
with cervical spine disease. Anaesthesia 1993; 48: 910–13.
95. Fuchs G, Schwarz G, Baumgartner A, Kaltenb€ ock F, Voit-
Augustin H, Planinz W. Fiberoptic intubation in 327
neurosurgical patients with lesions of the cervical spine.
Journal of Neurosurgical Anesthesiology 1999; 11: 11–16.
96. Badiger S, John M, Fearnley RA, Ahmad I, Asai T. Optimizing
oxygenation and intubation conditions during awake fibre-
optic intubation using a high-flow nasal oxygen-delivery
system. British Journal of Anaesthesia 2015; 115: 629–32.
97. Rozario L, Sloper D, Sheridan MJ. Supplemental oxygen
during moderate sedation and the occurrence of clinically
significant desaturation during endoscopic procedures.
Gastroenterology Nursing 2008; 31: 281–5.
98. Milman N, Faurschou P, Grode G, Jørgensen A. Pulse oximetry
during fibreoptic bronchoscopy in local anaesthesia:
frequency of hypoxaemia and effect of oxygen
supplementation. Respiration 1994; 61: 342–7.
99. Reed MWR, O’leary DP, Duncan JL, Majeed AW, Wright B,
Reilly CS. Effects of sedation and supplemental oxygen during
upper alimentary tract endoscopy. Scandinavian Journal of
Gastroenterology 1993; 28: 319–22.
100. Reshef R, Shiller M, Kinberg R, et al. A prospective study
evaluating the usefulness of continuous supplemental oxygen
in various endoscopic procedures. Israel Journal of Medical
Sciences 1996; 32: 736–40.
101. Deitch K, Chudnofsky CR, Dominici P, Latta D, Salamanca Y.
The utility of high-flow oxygen during emergency department
procedural sedation and analgesia with propofol: a
randomized, controlled trial. Annals of Emergency Medicine
2011; 58: 360–4.
102. Deitch K, Chudnofsky CR, Dominici P. The utility of
supplemental oxygen during emergency department
procedural sedation and analgesia with midazolam and
fentanyl: a randomized. Controlled Trial. Annals of Emergency
Medicine 2007; 49: 1–8.
103. Hinkelbein J, Lamperti M, Akeson J, et al. European Society of
Anaesthesiology and European Board of Anaesthesiology
guidelines for procedural sedation and analgesia in adults.
European Journal of Anesthesiology 2018; 35: 6–24.
104. Surgeons M. Practice guidelines for moderate procedural
sedation and analgesia 2018: a Report by the American
Society of Anesthesiologists Task Force on Moderate
Procedural Sedation and Analgesia, the American Association
of Oral and Maxillofacial Surgeons, American. Anesthesiology
2018; 128: 437–79.
105. Newton M, Blightman K. Guidance on the provision of
sedation services 2016. In: Royal College of Anaesthetists.
Guidelines for the Provision of Anaesthesia Services. London:
Royal College of Anaesthetists, 2016: Chapter 19.
106. Ho AMH, Chung DC, To EWH, Karmakar MK. Total airway
obstruction during local anesthesia in a non-sedated patient
with a compromised airway. Canadian Journal of Anesthesia
2004; 51: 838–41.
107. Song J. A comparison of the effects of epinephrine and
xylometazoline in decreasing nasal bleeding during
nasotracheal intubation. Journal of Dental Anesthesia and
Pain Medicine 2017; 17: 281.
Ahmad et al. | Awake tracheal intubation guidelines
108. O’Hanlon J, Harper KW. Epistaxis and nasotracheal
intubation–prevention with vasoconstrictor spray. Irish
Journal of Medical Science 1994; 163: 58–60.
109. El-Boghdadly K, Pawa A, Chin KJ. Local anesthetic systemic
toxicity: current perspectives. Local and Regional Anesthesia
2018; 11: 35–44.
110. Milman N, Laub M, Munch EP, Angelo HR. Serum
concentrations of lignocaine and its metabolite
monoethylglycinexylidide during fibre-optic bronchoscopy in
local anaesthesia. Respiratory Medicine 1998; 92: 40–3.
111. Hasmoni MH, Rani MFA, Harun R, Manap RA, Tajudin NAA,
Anshar FM. Randomized-controlled trial to study the
equivalence of 1% versus 2% lignocaine in cough suppression
and satisfaction during bronchoscopy. Journal of
Bronchology 2008; 15: 78–82.
112. Mainland PA, Kong AS, Chung DC, Chan CH, Lai CK.
Absorption of lidocaine during aspiration anesthesia of the
airway. Journal of Clinical Anaesthesia 2001; 13: 440–6.
113. Efthimiou J, Higenbottam T, Holt D, Cochrane GM. Plasma
concentrations of lignocaine during fibreoptic bronchoscopy.
Thorax 1982; 37: 68–71.
114. Williams KA, Barker GL, Harwood RJ, Woodall NM. Combined
nebulization and spray-as-you-go topical local anaesthesia of
the airway. British Journal of Anaesthesia 2005; 95: 549–53.
115. Ingrande J, Lemmens HJM. Dose adjustment of anaesthetics
in the morbidly obese. British Journal of Anaesthesia 2010;
105: i16–23.
116. Woodruff C, Wieczorek PM, Schricker T, Vinet B, Backman SB.
Atomised lidocaine for airway topical anaesthesia in the
morbidly obese: 1% compared with 2%. Anaesthesia 2010;
65: 12–17.
117. Xue FS, Liu HP, He N, et al. Spray-as-you-go airway topical
anesthesia in patients with a difficult airway: a randomized,
double-blind comparison of 2% and 4% lidocaine. Anesthesia
and Analgesia 2009; 108: 536–43.
118. Wieczorek PM, Schricker T, Vinet B, et al. Airway topicalisation
in morbidly obese patients using atomised lidocaine: 2%
compared with 4%. Anaesthesia 2007; 62: 984–8.
119. Haldar R, Gyanesh P, Samanta S. Lignocaine toxicity from
topical anaesthesia of airway during awake fibreoptic
intubation. Journal of Neuro Anesthesiology and Critical Care
2014; 1: 146–7.
120. Giordano D, Panini A, Pernice C, Raso MG, Barbieri V.
Neurologic toxicity of lidocaine during awake intubation in a
patient with tongue base abscess. Case report. American
Journal of Otolaryngology – Head and Neck Medicine and
Surgery 2014; 35: 62–5.
121. El-Boghdadly K, Chin KJ. Local anesthetic systemic toxicity:
continuing professional development. Canadian Journal of
Anesthesia 2016; 63: 330–49.
122. Association of Anaesthetists. Safety Guideline: management
of severe local anaesthetic toxicity, 2010. https://anaesthe
tists.org/Home/Resources-publications/Guidelines/Manage
ment-of-severe-local-anaesthetic-toxicity (accessed 17/10/
2019).
123. Osula S, Stockton P, Abdelaziz MM, Walshaw MJ. Intratracheal
cocaine induced myocardial infarction: an unusual
complication of fibreoptic bronchoscopy. Thorax 2003; 58:
733–4.
124. Chiu YC, Brecht K, Dasgupta DS, Mhoon E. Myocardial
infarction with topical cocaine anesthesia for nasal surgery.
archives of otolaryngology-head and neck. Surgery 1986.
125. Higgins TS, Hwang PH, Kingdom TT, Orlandi RR,
Stammberger H, Han JK. Systematic review of topical
vasoconstrictors in endoscopic sinus surgery. Laryngoscope
2011; 121: 422–32.
126. Lenders GD, Jorens PG, De Meyer T, Vandendriessche T,
Verbrugghe W, Vrints CJ. Coronary spasm after the topical
© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Anaesthesia 2020, 75, 509–528
use of cocaine in nasal surgery. American Journal of Case
Reports 2013; 14: 76–9.
127. Cara DM, Norris AM, Neale LJ. Pain during awake nasal
intubation after topical cocaine or phenylephrine/lidocaine
spray. Anaesthesia 2003; 58: 777–80.
128. Takaenoki Y, Masui K, Oda Y, Kazama T. The
pharmacokinetics of atomized lidocaine administered via the
trachea: a randomized trial. Anesthesia and Analgesia 2016;
123: 74–81.
129. Cabrini L, Baiardo Redaelli M, Ball L, et al. Awake fiberoptic
intubation protocols in the operating room for anticipated
difficult airway: a systematic review and meta-analysis of
randomized controlled trials. Anesthesia and Analgesia 2019;
128: 971–80.
130. Reasoner DK, Warner DS, Todd MM, Hunt SW, Kirchner J. A
comparison of anesthetic techniques for awake intubation in
neurosurgical patients. Journal of Neurosurgical
Anesthesiology 1995; 7: 94–9.
131. Wiles JR, Kelly J, Mostafa SM. Hypotension and bradycardia
following superior laryngeal nerve block. British Journal of
Anaesthesia 1989; 63: 125–7.
132. Sitzman BT, Rich GF, Rockwell JJ, Leisure GS, Durieux ME,
DiFazio CA. Local anesthetic administration for awake direct
laryngoscopy. Are glossopharyngeal nerve blocks superior?
Anesthesiology 1997; 86: 34–40.
133. O’Callaghan C, Barry PW. The science of nebulised drug
delivery. Thorax 1997; 52: S31–44.
134. Kundra P, Kutralam S, Ravishankar M, Scand AA, Care C,
Education PM. Local anaesthesia for awake fibreoptic
nasotracheal intubation. Acta Anaesthesiologica
Scandinavica 2000; 44: 511–16.
135. Ali-Melkkila T, Kaila T, Kanto J. Glycopyrrolate;
pharmacokinetics and some pharmacodynamics findings.
Acta Anaesthesiologica Scandinavica 1989; 33: 513–17.
136. Brookman CA, Teh HP, Morrison LM. Anticholinergics improve
fibreoptic intubating conditions during general anaesthesia.
Canadian Journal of Anaesthesia 1997; 44: 165–7.
137. Johnston KD, Rai MR. Conscious sedation for awake fibreoptic
intubation: a review of the literature. Canadian Journal of
Anesthesia 2013; 60: 584–99.
138. Academy of Medical Royal Colleges. Safe sedation practice
for healthcare procedures, 2013. https://www.rcoa.ac.uk/
system/files/PUB-SafeSedPrac2013.pdf (accessed 17/10/
2019).
139. Rai MR, Parry TM, Dombrovskis A, Warner OJ. Remifentanil
target-controlled infusion vs propofol target-controlled
infusion for conscious sedation for awake fibreoptic
intubation: a double-blinded randomized controlled trial.
British Journal of Anaesthesia 2008; 100: 125–30.
140. Lallo A, Billard V, Bourgain JL. A comparison of propofol and
remifentanil target-controlled infusions to facilitate fiberoptic
nasotracheal intubation. Anesthesia and Analgesia 2009;
108: 852–7.
141. Zhang X, He W, Wu X, Zhou X, Huang W, Feng X. TCI
remifentanil vs. TCI propofol for awake fiber-optic intubation
with limited topical anesthesia. International Journal of
Clinical Pharmacology and Therapeutics 2012; 50: 10–16.
142. Silvestri S, Ladde JG, Brown JF, et al. Endotracheal tube
placement confirmation: 100% sensitivity and specificity with
sustained four-phase capnographic waveforms in a cadaveric
experimental model. Resuscitation 2017; 115: 192–8.
143. Grmec S.  Comparison of three different methods to confirm
tracheal tube placement in emergency intubation. Intensive
Care Medicine 2002; 28: 701–4.
144. Merry AF, Mitchell SJ. Complications of anaesthesia.
Anaesthesia 2018; 73: 7–11.
145. Wahba SS, Tammam TF, Saeed AM. Comparative study of
awake endotracheal intubation with Glidescope video
527
Anaesthesia 2020, 75, 509–528
laryngoscope versus flexible fiber optic bronchoscope in
patients with traumatic cervical spine injury. Egyptian Journal
of Anaesthesia 2012; 28: 257–60.
146. Kramer A, M€ uller D, Pf€
ortner R, Mohr C, Groeben H. Fibreoptic
vs. videolaryngoscopic (C-MACâ D-BLADE) nasal awake
intubation under local anaesthesia. Anaesthesia 2015; 70:
400–6.
147. Lewis SR, Butler AR, Parker J, Cook TM, Smith AF.
Videolaryngoscopy versus direct laryngoscopy for adult
patients requiring tracheal intubation. Cochrane Database of
Systematic Reviews 2016; 2016: CD011136.
148. Popat M, Mitchell V, Dravid R, Patel A, Swampillai C, Higgs A.
Difficult Airway Society Guidelines for the management of
tracheal extubation. Anaesthesia 2012; 67: 318–40.
149. Kirkpatrick MB, Sanders RV, Bass JB. Physiologic effects and
serum lidocaine concentrations after inhalation of lidocaine
from a compressed gas-powered jet nebulizer. American
Review of Respiratory Disease 1987; 136: 447–9.
150. Schønemann NK, van der Burght M, Arendt-Nielsen L,
Bjerring P. Onset and duration of hypoalgesia of lidocaine
spray applied to oral mucosa – a dose response study. Acta
Anaesthesiologica Scandinavica 1992; 36: 733–5.
151. Roberts MH, Gildersleve CD. Lignocaine topicalization of the
pediatric airway. Pediatric Anesthesia 2016; 26: 337–44.
152. General Medical Council. Good medical practice Domain 1:
Knowledge skills and performance Manchester, UK. 2019.
153. Yentis SM, Hartle AJ, Barker IR, et al. Association of
Anaesthetists: consent for anaesthesia 2017. Anaesthesia
2017; 72: 93–105.
154. Difficult Airway Society. DAS ATI Patient Information Leaflet.
https://das.uk.com/files/awake_leaflet.pdf (accessed 21/02/
2019).
155. Martin KM, Larsen PD, Segal R, Marsland CP. Effective
nonanatomical endoscopy training produces clinical airway
endoscopy proficiency. Anesthesia and Analgesia 2004; 99:
938–44.
156. Baker PA, Weller JM, Baker MJ, et al. Evaluating the ORSIM â
simulator for assessment of anaesthetists’ skills in flexible
bronchoscopy: aspects of validity and reliability. British
Journal of Anaesthesia 2016; 117: i87–91.
157. Casso G, Schoettker P, Savoldelli GL, Azzola A, Cassina T.
Development and initial evaluation of a novel, ultraportable,
virtual reality bronchoscopy simulator. Anesthesia and
Analgesia 2018; 1.
158. Jiang B, Ju H, Zhao Y, Yao L, Feng Y. Comparison of the efficacy
and efficiency of the use of virtual reality simulation with high-
fidelity mannequins for simulation-based training of fiberoptic
bronchoscope manipulation. Simulation in Healthcare: Journal
of the Society for Simulation in Healthcare 2018; 13: 83–7.
159. Marsland C, Larsen P, Segal R, et al. Proficient manipulation of
fibreoptic bronchoscope to carina by novices on first clinical
attempt after specialized bench practice. British Journal of
Anaesthesia 2010; 104: 375–81.
160. Cole AFD, Mallon JS, Rolbin SH, Ananthanarayan C.
Fiberoptic intubation using anesthetized, paralyzed, apneic
528 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Ahmad et al. | Awake tracheal intubation guidelines
patients results of a resident training program.
Anesthesiology 1996; 84: 1101–6.
161. De Oliveira GS, Glassenberg R, Chang R, Fitzgerald P,
McCarthy RJ. Virtual airway simulation to improve dexterity
among novices performing fibreoptic intubation. Anaesthesia
2013; 68: 1053–8.
162. The Royal College of Anaesthetists. CCT in Anaesthetics –
Advanced Level Training (Annex E) Edition 2, 2017.
https://www.rcoa.ac.uk/system/files/TRG-CCT-ANNEXE.pdf
(accessed 29/03/2019).
163. McNarry AF, Dovell T, Dancey FML, et al. Perception of
training needs and opportunities in advanced airway skills: a
survey of British and Irish trainees. European Journal of
Anesthesiology 2007; 24: 498–504.
164. Ezri T, Szmuk P, Warters RD, Katz J, Hagberg CA. Difficult
airway management practice patterns among
anesthesiologists practicing in the United States: have we
made any progress? Journal of Clinical Anaesthesia 2003; 15:
418–22.
165. Wiles MD, McCahon RA, Armstrong JAM. An audit of
fibreoptic intubation training opportunities in a UK teaching
hospital. Journal of Anesthesiology 2014; 2014: 1–4.
166. Dawson AJ, Marsland C, Baker P, Anderson BJ. Fibreoptic
intubation skills among anaesthetists in New Zealand.
Anaesthesia and Intensive Care 2005; 33: 777–83.
167. Goldmann K, Braun U. Airway management practices at
German university and university-affiliated teaching hospitals
– Equipment, techniques and training: results of a nationwide
survey. Acta Anaesthesiologica Scandinavica 2006; 50: 298–
305.
168. Ahmad I, Onwochei DN, Muldoon S, Keane O, El-Boghdadly
K. Airway management research: a systematic review.
Anaesthesia 2019; 74: 225–36.
169. Ahmad I, El-Boghdadly K. From evidence based on practice to
evidence-based practice: time for a difficult airway
management research strategy. Anaesthesia 2019; 74: 135–
9.
Supporting Information
Additional supporting information may be found online via
the journal website.
Appendix S1. Summary of the systematic review
methodology.
Appendix S2. The Difficult Airway Society checklist for
awake tracheal intubation (ATI) in adults. SAD, supraglottic
airway device.
Appendix S3. The Difficult Airway Society documentation
sticker for awake tracheal intubation (ATI) in adults. FB,
flexible bronchoscopy; VL, videolaryngoscopy; L, left; R,
right.
Anaesthesia 2020, 75, 529–540 doi:10.1111/anae.14891

Review Article

A systematic review and cost effectiveness analysis of

reusable vs. single-use flexible bronchoscopes
J. M. Mouritsen,1 L. Ehlers,2 J. Kovaleva,3 I. Ahmad4,5 and K. El-Boghdadly4,5

1 Student, Department of Health Science and Technology, Faculty of Medicine, Aalborg University, Aalborg, Denmark
2 Professor, Danish Center of Healthcare Improvements, Institute of Business and Management, Aalborg University,
Aalborg, Denmark
3 Clinical Microbiologist/Clinical Pathologist, Sint-Jozefkliniek Bornem and Willebroek, Bornem, Belgium
4 Consultant, Guy’s and St Thomas’ NHS Foundation Trust, London, UK
5 Honorary Senior Lecturer, King’s College London, UK

Summary
The cost effectiveness of reusable vs. single-use flexible bronchoscopy in the peri-operative setting has yet to be
determined. We therefore aimed to determine this and hypothesised that single-use flexible bronchoscopes
are cost effective compared with reusable flexible bronchoscopes. We conducted a systematic review of the
literature, seeking all reports of cross-contamination or infection following reusable bronchoscope use in any
clinical setting. We calculated the incidence of these outcomes and then determined the cost per patient of
treating clinical consequences of bronchoscope-induced infection. We also performed a micro-costing analysis
to quantify the economics of reusable flexible bronchoscopes in the peri-operative setting from a high-
throughput tertiary centre. This produced an accurate estimate of the cost per use of reusable flexible
bronchoscopes. We then performed a cost effectiveness analysis, combining the data obtained from the
systematic review and micro-costing analysis. We included 16 studies, with a reported incidence of cross-
contamination or infection of 2.8%. In the micro-costing analysis, the total cost per use of a reusable flexible
bronchoscope was calculated to be £249 sterling. The cost per use of a single-use flexible bronchoscope was
£220 sterling. The cost effectiveness analysis demonstrated that reusable flexible bronchoscopes have a cost
per patient use of £511 sterling due to the costs of treatment of infection. The findings from this study suggest
benefits from the use of single-use flexible bronchoscopes in terms of cost effectiveness, cross-contamination
and resource utilisation.

.................................................................................................................................................................
Correspondence to: K. El-Boghdadly
Email: elboghdadly@gmail.com
Accepted: 19 September 2019
Keywords: airway management; bronchoscope; economics; infection; intubation
Twitter: @elboghdadly; @dr_imranahmad

Introduction considered, even when they are reprocessed according to

Flexible bronchoscopes allow visualisation of the infection control guidelines and recommendations [2–5].
nasopharynx, oropharynx, larynx, trachea and its However, the transmission of pathogenic organisms via
subsequent divisions for diagnostic and therapeutic contaminated reusable flexible bronchoscopes remains an
purposes. It is estimated that 500,000 bronchoscopic evident risk [6–8], even if appropriate decontamination
procedures are performed annually in the USA alone [1]. procedures are adhered to [9]. There are unquantifiable
The risk of transmission of infection following bronchoscopy risks of cross-contamination and infection from reusable
with reusable flexible bronchoscopes is often under- flexible bronchoscopes, along with uncertainty regarding

© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists 529
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and
distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
Anaesthesia 2020, 75, 529–540
their cost effectiveness. Reusable flexible bronchoscopes
are often used by anaesthetists to place tracheal tubes,
either awake or asleep, and to check adequate positioning
of double-lumen tubes. Therefore, cross-contamination risk
will also apply to these patients.
Single-use flexible bronchoscopes are delivered sterile
and thus should minimise the risk of infection transmission
and cross-contamination compared with reusable flexible
bronchoscopes. A previously reported cost effectiveness
study of single-use flexible bronchoscopes in a typical
intensive care unit (ICU) setting in the USA demonstrated
that subsequent implementation is cost effective when
looking at cross-contamination and potential subsequent
infection, and it is associated with increased patient safety
[10]. There are several reports of cross-contamination of
reusable flexible bronchoscopes due to inappropriate
cleaning, disinfection or lack of leak testing and drying
[11,12]. These reports do not provide a quantifiable risk for
cross-contamination and subsequent infection, but it is
accepted that there is a risk and cases are under-reported
[8, 13–15]; consequently, the literature lacks a quantified
risk of cross-contamination and subsequent infection due to
flexible bronchoscopy [8, 16, 17]. Moreover, several micro-
costing studies of reusable flexible bronchoscopes do not
include costs of infections, which is why there is some
uncertainty regarding these estimates [18–20].
We therefore aimed to determine the cost per use and
cross-contamination risk of reusable flexible bronchoscopes
and to ascertain the cost effectiveness of single-use flexible
bronchoscopes compared with reusable flexible
bronchoscopes in various clinical settings. To achieve this
we conducted a micro-costing analysis of flexible
bronchoscope utilisation from a high-throughput tertiary
centre [21] and a systematic review of the literature.
Our primary hypothesis was that single-use flexible
bronchoscopes are equally or more cost effective than
reusable flexible bronchoscopes.
Methods
The preferred reporting items for systematic reviews and
meta-analyses (PRISMA) guidance was adhered to in the
conduct of the systematic review [22]. Given an evident risk of
patient cross-contamination and infection with reusable
flexible bronchoscopes [6,8] and no risk using single-use
flexible bronchoscope due to its single-use modality, the
effect measure in this cost effectiveness analysis was defined
as the avoided risk of infection using single-use compared
with reusable flexible bronchoscopes. A comprehensive
search strategy was conducted of the PubMed, MEDLINE and
Embase databases to identify relevant literature for the risk of
530 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Mouritsen et al. | Cost effectiveness of flexible bronchoscopes
cross-contamination and infection associated with reusable
flexible bronchoscopes. The following search terms or
combinations were employed: ‘bronchoscopy’;
‘bronchoscope’ determine ‘infection’ determine ‘cross-
infection’ determine ‘pseudo-outbreak’ determine ‘outbreak’
determine ‘device contamination’ determine and ‘hospital
infection’, including word variations and assorted
permutations. English-language studies on humans were
included from 1982 to 2018. Retrospective and prospective
observational studies were included when studying the risk of
cross-contamination and infection post-bronchoscopy.
Reference lists of review articles were hand-searched to
locate studies that may have been missed in our initial search.
Eligibility criteria of studies were met if cross-contamination of
patients who underwent bronchoscopic procedures was
detected by traditional typing systems based on phenotypes
or more recent methods that examine the relatedness of
isolates at a molecular level, such as polymerase chain
reaction or pulse-field gel electrophoresis [23]. To quantify
the risk of infection, we needed the number of:
bronchoscopic procedures; patients who underwent
bronchoscopy; contaminated patients; and infected patients.
Studies were included in the quantitative analysis if at least
three of these four variables were reported. Studies were
excluded if they had a non-quantifiable risk. A study with a
non-quantifiable risk was defined as one in which less than
three of the above-mentioned parameters were reported.
Other exclusion criteria were other endoscopic procedures
(e.g. gastroscopy), if bronchoscope contamination was not
detected or reported by recognised typing methods, or if the
setting of the studies was not clinically relevant for answering
our hypothesis. The risk of infection was used as the
measurement of effectiveness. In the event of incomplete
data on the number of bronchoscopic procedures and
number of patients included, a simple regression method
was applied to predict missing data in this large
heterogeneous group of patients eligible for bronchoscopy.
The relationship was used to predict missing data points
within these two variables. Once regression methods were
applied, all studies were included for quantitative synthesis.
The risk of cross-contamination and infection was determined
by a weighted average using a fixed-effects model to reflect a
more precise estimate. There are currently no reported
cases of cross-contamination using single-use flexible
bronchoscopes, which is why the risk is expected to be 0%.
The cost effectiveness of single-use vs. reusable flexible
bronchoscopes was estimated by using a literature review
to obtain the best available evidence of effects. The effect
measure was the risk of infection. The time horizon of
the cost-effectiveness analysis was within 1 year. The
Mouritsen et al. | Cost effectiveness of flexible bronchoscopes Anaesthesia 2020, 75, 529–540

micro-costing analysis was conducted at Guy’s and St.
Thomas’ NHS Foundation Trust Department of Anaesthesia.
The total cost per use of reusable flexible bronchoscopes
for tracheal intubation and double-lumen tube position
verification was estimated, and the cost per use of single-
use flexible bronchoscopes (including the monitor) was
determined. All costs identified were adjusted to pounds
sterling (£) in 2017. To determine the present value of
capital expenditures, a discount rate of 3.5% was used.
Capital acquisition costs were annualised across a 5-year
period for bronchoscopes and related equipment, and a
30-year period for buildings.
A decision tree was constructed using Tree Age (2016
version, TreeAge Software, MA, USA), which enabled the
comparison of the cost effectiveness of single-use flexible
bronchoscopes to reusable flexible bronchoscopes (Fig. 1).
The modelling approach was based on principles of good
practice for decision analytic modelling in healthcare
analyses [24].
The cost perspective used in this analysis was a UK
government third-party payer, and the clinical setting was
an anaesthetics department where tracheal intubations and
double-lumen tube position verification were frequently
carried out. Multiple bronchoscopies for either tracheal
intubation or double-lumen tube position verification were
examined to determine the resources and costs associated
with the procedures. The procedures were monitored in
detail from start to finish.
Data obtained retrospectively from various fiscal years
(2000–2017) were used to quantify costs for capital
acquisitions, repairs, consumables, disposables and service
agreements. Additionally, labour time associated with the
reprocessing cycles of the reusable flexible bronchoscopes
was meticulously measured. The department has 19
reusable flexible bronchoscopes. Of these, 12 are used for
tracheal intubation (either awake or asleep), whereas seven
are reserved for double-lumen tube position verification.
The activities of performing bronchoscopy are dispersed
across various operating theatres on one floor, whereas the
reprocessing is divided on two different floors. The reusable
flexible bronchoscopes undergo a precleaning cycle on the
same floor as the operating theatre using equipment
including detergents and brushes, followed by a second-
stage manual clean performed in a central cleaning facility
on another floor. Costs were estimated using the mean
annual number of bronchoscopic procedures and
reprocessing volumes done by the automated endoscope
reprocessors in this tertiary hospital.
In the Supporting Information (Data S1), all collected cost
data from the micro-costing analysis are presented in detail.
This includes: (1) capital and repair costs of reusable flexible
bronchoscopes and rack systems; (2) capital and repair costs
of reprocessing capital and additional equipment; (3) time
measurements of the specific reprocessing steps; (4) average
cost of labour-related reprocessing; (5) reprocessing
equipment and costs incurred; and (6) the allocation keys and
reason for usage [25].
The costs of the clinical outcome were determined by
identifying treatment costs related to the clinical
manifestations of a post-bronchoscopic contamination or
infection. The incidence of postoperative pulmonary
complications (including pneumonia and sepsis) is
estimated to be up to 23% in an unselected group of
patients having general anaesthesia [26]. This figure is
higher in patients undergoing thoracic or head and neck
surgery [27], which are the cohort of patients in whom

Figure 1 Decision tree model used in this cost effectiveness analysis.

© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists 531
Anaesthesia 2020, 75, 529–540 Mouritsen et al. | Cost effectiveness of flexible bronchoscopes

flexible bronchoscopes are most commonly used. However,
there are no data directly demonstrating bronchoscope-
induced cross-contamination or infection in these patients,
as infection is often assumed to be multifactorial. We were
therefore only able to use published data of infection and
cross-contamination from the ICU or elective endoscopy or
bronchoscopy population. The cost of the clinical outcome
was estimated from the studies that were included
according to the clinical manifestations that were reported
in included studies, such as respiratory tract infection
prophylaxis and therapy [28–36] and sepsis [33, 35]. In the
cohort of patients of interest for this study, we considered
ventilator-associated pneumonia, sepsis and community-
acquired pneumonia as suitable outcomes. The average
cost of ventilator-associated pneumonia was identified from
a review of 28 US community hospitals to be £25,426
sterling per patient [37]. From a systematic review of
hospital-related cost of sepsis, the treatment-related costs
were identified as £27,123 sterling per patient [38]. The
costs of inpatient and outpatient community-acquired
pneumonia were estimated from > 28,000 community-
acquired pneumonia episodes from a large US database
study at £13,151 and £1948, respectively, per patient [39].
The weighted average was defined as the treatment-related
costs per patient infected. This value was imputed in the
cost-effectiveness analysis. A summary of all costs and effect
inputs for the cost-effectiveness analysis is presented in
Table 1.
Using the results from the literature review and the
micro-costing analysis to compute the cost-effectiveness
analysis, a base-case result was generated. Sensitivity
analyses were undertaken to capture uncertainty within
parameters and to provide sufficient insight to decision-
makers. Uncertainty is captured through deterministic and
probabilistic sensitivity analyses. One-way (univariate)
sensitivity analyses were applied to all parameters in the
model to test its robustness by examining the impact on the
incremental cost effectiveness ratio. All cost parameters
were varied by  50%. Considering the average period of
an 8-month investigation across the 16 studies, the effect
parameters were varied from a low value of 0% risk to a high
value of 20% risk. A scenario analysis was conducted from a
previous Delphi approach to the general risk of patient
cross-contamination and infection [10]. Furthermore, the
impact of the incremental cost-effectiveness ratio was
observed by varying the amortisation period of capital
investments related to bronchoscopy materials to 10 years
from the previous 5 years.
The probabilistic sensitivity analysis quantifies the
overall uncertainty within parameters using pre-specified
distributions (Table 1). The modality of the probabilistic
sensitivity analysis was a second-order Monte Carlo
simulation with 10,000 iterations of the mean incremental
cost-effectiveness ratios. These 10,000 iterations were
drawn up in a cost effectiveness scatterplot to represent the
expected avoidance of infection risk using a single-use
compared with reusable flexible bonchoscopes.
According to International Guidelines in Health
Economics, the mean is used as it is the only relevant
measure for economic decision making [40]. In economic
calculations, we aim to capture the uncertainty of the
sample mean, that is, parameter uncertainty, rather than
variability or heterogeneity, that is, stochastic variability.
The uncertainty in the expected mean is the standard
error (SE) [40]. Consequently, we report all cost data as
mean (SE).

Table 1 Inputs for the cost effectiveness model. Base-case value, the standard error (SE) and the distribution are provided.
Parameters Base-case value (SE) Distribution
Effects
Risk of patient contamination using a reusable FB 0.153 (0.009) Beta
Risk of subsequent infection using a reusable FB 0.181 (0.018) Beta
Risk of patient contamination using a single-use FB 0.0 (0.001) Beta
Risk of subsequent infection using a single-use FB 0.0 (0.001) Beta
Costs
Capital cost per use of a reusable FB (reusable FB, stack systems, reprocessing capital) £116.4a (29.10) Gamma
Repair cost per use of a reusable FB (reusable FB, stack systems, reprocessing capital) £92.9* (23.20) Gamma
Reprocessing cost per use a reusable FB (labour time and equipment) £39.9* (10.00) Gamma
Cost of per clinical outcome [37–39] (per patient infection) £9,454 (£1,158) Gamma
Cost per use of a single-use FB, including monitor £220 (21.80) Gamma
a
SE was not estimated. However, a conservative approach was taken by varying the parameter with a SE of 25%. FB, flexible
bronchoscope.

532 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Mouritsen et al. | Cost effectiveness of flexible bronchoscopes
Results
We identified 890 citations, of which 12 were duplicates.
Seven additional studies were considered through hand-
searching of two review articles [7, 8]. Across numerous
studies a non-quantifiable risk was identified [41–62], thus
not fulfilling the inclusion criteria for the quantitative
analysis. After screening based on title and abstract and
full-text review, we identified 16 studies for quantitative
analysis of the cross-contamination and infection risk
(Fig. 2) [12, 28–36, 63–68]. The setting of these studies
were patients who underwent bronchoscopy in a hospital
intensive care or respiratory unit setting, or during
bronchoscopy or endoscopy in the UK, USA, France, Spain,
Figure 2 Study flow chart.
© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Anaesthesia 2020, 75, 529–540
Australia or Taiwan (Table 2). The study designs were
prospective observational and retrospective studies, and
the period of contamination and infection investigation of
patients undertaken was from one to 23 months [30, 33].
A total of 2351 patients underwent 3120 various
bronchoscopic procedures. Of these procedures, 476 cases
of patient contamination were detected. Eighty-six of these
patients were reported to have a bronchoscope-linked
manifestation of infection, including pneumonia or other
respiratory tract infection. Seven of the included studies
contained a missing data-point relating to the number of
bronchoscopic procedures (five studies) and the number
of patients who underwent a bronchoscopy (two studies)
533
 Table 2 Sixteen studies identified for quantitative analysis of the cross-contamination and infection risk. Values are number or mean (SD).

534
Period of
investi- Patient Cases of Detection Typing
gation age Proce- contami- Cases of of contami- system
Source Study design (months) Country Setting (years) dures Patients nation infection nation used Infection(s)
Blanc et al. [68] Observational 6 USA NA NA 410 299a 35 0 Yes Ribotyping No infections
Botana-Rial et al. [30] Observational 1 Spain BU NA 154 118a 39 21 Yes REP-PCR Treatment to
prevent the
development
of pneumonia
Anaesthesia 2020, 75, 529–540

Chroneou et al. [63] Observational 2 USA NA 71 (49–86) 76a 57 9 0 Yes REP-PCR No infections
DiazGranados et al. [35] Observational 2 USA BU NA 27a 20 12 2 Yes PFGE Pneumonia
and sepsis
Waite et al. [64] Retrospective 12 UK ICU NA 63a 47 18 0 Yes PFGE No infections
Wang et al. [31] Retrospective 4 Taiwan NA 60 (45–79) 163a 123 18 8 Yes AFB Patients were
treated as
mycobacterial
infected
Silva et al. [12] Retrospective 22 USA EU NA 429a 324 41 0 Yes Ribotyping No infections
Nye et al. [65] Retrospective 6 UK EU 61 (40–80) 58 7 7 0 Yes Culturing and No infections
lipid analysis
Guy et al. [32] Retrospective 7 France ICU 62 (49–73) 216 157 10 8 Yes PFGE Treatment to
prevent the
development
of pneumonia
Sammartino et al. [36] Retrospective 3 USA NA 56 (36–76) 19 19 11 1 Yes Serotyping Pneumonia
Campagnaro et al. [66] Retrospective 5 Australia NA NA 65 65 12 0 Yes DNA probing No infections
Corne et al. [28] Retrospective 8 France ICU NA 61 36 16 4 Yes PFGE Pneumonia
Kirschke et al. [34] Retrospective 4 USA NA 59 (24–88) 66 60 20 1 Yes PFGE Pneumonia
Srinivasan et al. [33] Retrospective 23 USA EU NA 665 414 39 39 Yes PFGE Pneumonias,
sepsis,
respiratory
tract infection
Pappas et al. [29] Retrospective 11 USA NA NA 195 187 72 2 Yes Culturing No data
C^
etre et al. [67] Observational/ 8 France EU NA 453 418 117 0 Yes PFGE No infections
retrospective
Total 3120 2351 476 86

AFB, acid-fast bacillus testing; BU, bronchoscopic unit; DNA, deoxyribonucleic acid; EU, endoscopic unit; ICU, intensive care unit; NA, not available; PFGE, pulsed-field gel
electrophoresis; REP-PCR, repetitive extragenic palindromic-polymerase chain reaction.
a
Predicted data-point.
Mouritsen et al. | Cost effectiveness of flexible bronchoscopes

© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Mouritsen et al. | Cost effectiveness of flexible bronchoscopes
[12, 30, 31, 35, 63, 64, 68]. These missing data-points were
predicted using a linear regression method. There was a
strong correlation between the number of patients
who underwent bronchoscopy and the number of
bronchoscopic procedures (r = 0.92) (see Supporting
Information, Data S2). From this, 118 and 299 patients were
predicted, respectively, for two of the studies with a missing
data point [30, 68]. A strong correlation was also
demonstrated between the number of bronchoscopic
procedures carried out and the number of patients who
underwent bronchoscopy (r = 0.93). Five studies contained
a missing data-point in terms of number of bronchoscopic
procedures, which were therefore calculated as 163, 429,
27, 76 and 63, respectively [12, 31, 35, 63, 64].
From the included studies, the differentiation between
a pre-existing infection and a flexible bronchoscope-related
infection was determined using traditional bacterial
recognition or more recent methods that examine the
association of isolates at a molecular level (Table 2).
One thousand annual bronchoscopic procedures
are performed by the Department of Anaesthesia at
Guy’s Hospital. The reprocessing volumes by
automated endoscope reprocessor were 10,075 cycles
per year. The results from the micro-costing analysis
revealed a mean (SE) capital cost per use of a reusable
flexible bronchoscope at £116.40 (£29.10), whereas the
repair and reprocessing cost per use of a reusable
flexible bronchoscope was estimated at £92.90 (£23.20)
and £39.90 (£10.00), respectively. This equates to a
total cost per use of a reusable flexible bronchoscope
at £249.20 sterling. The mean (SE) cost per use of a
single-use flexible bronchoscope were provided by
â TM
Ambu (Ambu aScope 4, Copenhagen, Denmark) at
£220.00 sterling (£21.80), including the monitor.
In the Supporting Information (Data S1), a detailed
overview and description are provided of all costs incurred,
and the allocation keys that were employed to more
accurately reflect reality.
In the cost-effectiveness analysis, we found reusable
flexible bronchoscopes to have a mean (SE) cost per patient
of £511.00 sterling (£59.60), with an associated risk of
infection of 2.8%. The mean (SE) cost per patient with single-
use flexible bronchoscopes are estimated at £220.00
(£21.80) and a 0% risk of infection. Base-case results indicate
a net saving of £291.00 to hospitals and an avoided risk
of infection of patients undergoing bronchoscopy at
2.8% with single-use flexible bronchoscopes compared
with reusable flexible bronchoscopes. The base-case
incremental cost-effectiveness ratio is £10,505, which is
interpreted as the cost of avoided patient infection.
© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Anaesthesia 2020, 75, 529–540
Varying cost inputs by  50% did not have a
significant impact on the expected value of the incremental
cost-effectiveness ratio, which varied from £15,232 to
£5,778. From Fig. 3, the cost parameter with the greatest
impact on cost effectiveness is the cost of clinical outcome,
whereas the parameter with the lowest impact is the
reprocessing cost per use of a reusable flexible
bronchoscope.
Considering the average of 8 months of investigation
across the 16 studies of patient contamination and infection,
one-way sensitivity results from the CEA and a scenario
analysis is presented in Table 3.
The probabilistic sensitivity analysis was indicative
of a potential net savings to hospitals ranging from
£34 to 577 sterling per use and eliminating the risk of
infection of approximately 1.71–4.07% using single-use
flexible bronchoscopes compared with reusable flexible
bronchoscopes (Fig. 4).
Discussion
This systematic review and cost-effectiveness analysis found
the that single-use flexible bronchoscopes are cost effective
and associated with a lower risk of infection compared
with reusable flexible bronchoscopes. Sensitivity analyses
support these findings. Our systematic review
demonstrated that the risk of patient infection post-
bronchoscopy was 2.8%, with a cost per use of a reusable
flexible bronchoscope of £249 sterling and of a single-use
flexible bronchoscope of £220 sterling. Our cost-
effectiveness analysis demonstrated that reusable flexible
bronchoscopes have a cost per patient use of £511 sterling
due to the potential costs of treatment of infection.
The risk for patient contamination (15%) and infection
(18%) resulted in a 2.8% risk of patient infection post-
bronchoscopy. In a previous study, a Delphi approach was
used to estimate the general risk of patient contamination
(3.4%) and infection (21%) in critical care settings [10]. This
generated a risk of post-bronchoscopy infection of 0.7%.
The estimate from this present study (2.8%) is of high
accuracy because patient contamination and infection were
linked to bronchoscopes and the data were sourced from
international, multicentre settings with more than 2300
patients undergoing approximately 3100 various
procedures.
The cost per use of reusable flexible bronchoscopes for
tracheal intubation in a UK hospital carrying out 141 flexible
bronchoscope-assisted tracheal intubations annually was
previously estimated as £340 sterling [69]. In this present
study, the total cost per use of reusable flexible
bronchoscopes in an institution performing 1000 annual
535
Anaesthesia 2020, 75, 529–540 Mouritsen et al. | Cost effectiveness of flexible bronchoscopes

Figure 3 Tornado chart showing multiple one-way (univariate) sensitivity analyses of cost input parameters varied by  50%.
The incremental cost-effectiveness ratio (ICER) midpoint is £10,505 sterling and is equal to the base-case result from the cost-
effectiveness analysis. Low values (blue) for cost of clinical outcome, capital, repair and reprocessing cost per use of a reusable
flexible bronchoscope increase the ICER, whereas high values (red) reduce the ICER. Low values (blue) for cost of a single-use
bronchoscope reduce the ICER, whereas high values (red) increase the ICER. FB, flexible bronchoscope.

Table 3 Base-case result and one-way sensitivity analyses of effect parameters.

Difference Difference ICER (cost per avoided
Description in cost in effects patient infection)
Base-case £291 2.8% Dominant
One-way sensitivity of effects
Cross-contamination risk of 0% £29 No difference Dominant
Cross-contamination risk of 20% £371 3.6% Dominant
Infection risk of 0% £29 No difference Dominant
Infection risk of 20% £318 3.1% Dominant
One-way sensitivity of amortisation of capital investments
Capital investments amortised across 10 years £239 2.8% Dominant
Scenario analysis using estimates of cross-contamination and infection risk obtained from a Delphi approach[10]
Cross-contamination risk of 3.38% and infection risk of 21.3% £97 0.7% Dominant

ICER, incremental cost effectiveness ratio.

procedures was estimated to be £249 sterling. The cost per
use for capital, reprocessing and repairs are highly
dependent on local and clinical setting. The reprocessing
costs are dependent on the length of non-usage before
reusable flexible bronchoscopes need reprocessing again.
This is a variable time, but can be as low as 12 h [5]. Staffing
costs and use of productive working hours are also factors
that may have an economic impact. Furthermore, repair
costs per use of reusable flexible bronchoscopes are also
highly dependent on the local setting. McCahon and
Whynes conducted an analysis at a teaching hospital with
repair cost per use of £146 sterling [69], which contrasts to
the repair cost per use at Guy’s Hospital of £93 sterling. This
could be due to service agreements to cover repairs of all
capital equipment in some institutions but not others.
Finally, capital costs per use are dependent on economies
of scale advantages, that is, the volume of annual
bronchoscopies. Comparing McCahon and Whynes data to
that in Guy’s Hospital, the capital cost per use was £141 and
£114, respectively. Other costing analyses have been
conducted with total cost per use ranging from £111 to £540
sterling [18–20]. From these studies, the clinical setting in
terms of capital and repair expenditures play an important
role when determining the capital and repair cost per use,
whereas the reprocessing costs were similar between
settings.
There are two main strengths of this study, the first
being the micro-costing analysis as the cost comparison
between reusable and single-use equipment is complex.
Numerous overhead cost elements must be considered,
and this study captures more than previous studies [18–20].
The other main strength is the fact that this is the first study
to identify risk of patient contamination and infection from
the published literature.

536 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Mouritsen et al. | Cost effectiveness of flexible bronchoscopes
Figure 4 Probabilistic sensitivity analysis with 10,000 iterations (blue) and the base-case value (red).
In general, cost analyses lack precision in terms of
including all relevant overhead costs [18–20]. This is difficult
to manage because numerous elements contribute to the
overall cost per use estimate, such as water consumption,
electricity, training of new personnel in cleaning techniques,
maintaining updated and compliant guidelines on
reprocessing, handling of automated endoscope
reprocessor cycle failures, among others. Moreover, the
cost of the tracking systems is often left out [18–20]. If all
relevant indirect and overhead costs were identified and
included in those analyses, evidently it would add to the
cost per use. To advance the precision of the cost per use
estimate in this present micro-costing analysis, some of the
overhead costs mentioned above could have been
included. If done so, it would have added to the cost
effectiveness of single-use flexible bronchoscopes.
Findings from sensitivity analyses of cost per use of single-
use flexible bronchoscopes and reusable flexible
bronchoscopes support the cost effectiveness of single-use
flexible bronchoscopes, even when varying cost parameters
 50%.
In the institution examined in this study, the availability
of reusable flexible bronchoscopes remains limited due to
the unplanned requirement for flexible bronchoscopes, the
additional time required for reprocessing and the ongoing
requirement for bronchoscopes to be repaired. Availability
was still an issue, thus investment in a subset of single-use
flexible bronchoscope for emergencies was undertaken.
The availability of single-use flexible bronchoscopes is
constant as long as stocks are replenished, but the
© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Anaesthesia 2020, 75, 529–540
availability of reusable flexible bronchoscopes are
hampered by the need for reprocessing.
The environmental impact of clinical care is also an
element to consider. The disposable equipment, chemical
detergents, water, electricity, and other resources used
during the reprocessing cycles have an environmental
impact, whereas single-use flexible bronchoscopes are
disposed of after each use. From research comparing
carbon dioxide (CO2) emissions and resource consumption
from a single-use flexible bronchoscope (Ambu aScopeTM 4)
to a reusable flexible bronchoscope, results show that the
materials used for reprocessing are substantial when
comparing the two types of bronchoscopes [70]. The use of
cleaning materials and personal protective equipment
contributes to a comparable or potentially higher material
and energy consumption as well as emissions of CO2
equivalents and value of resource consumption for reusable
compared with single-use flexible bronchoscopes [70].
This study has several limitations. Our cost-
effectiveness analysis show that, ceteris paribus, the single-
use technology is superior in terms of costs and patient
safety, and a mix of single-use and reusable equipment may
be the only realistic alternative. But not all overhead costs
related to reusable equipment will be possible to eliminate,
and the predicted savings may be smaller under such
circumstances. Further research should be conducted to
investigate the cost effectiveness of a mixed usage strategy
for single-use and reusable bronchoscopes. Another
potential limitation is that we were unable to perform a
formal risk of bias assessment of the included studies
537
Anaesthesia 2020, 75, 529–540
as no appropriate, validated tool was suitable. An
additional limitation was that we included data from
patients presenting for outpatient bronchoscopy, who
may sometimes have systemic pathology, could be
immunocompromised and are, therefore, prone to infection
[6–8]. However, the cohort of patients sought in the micro-
costing analysis also included patients undergoing thoracic
surgical procedures and patients undergoing major head
and neck surgery, both of which may be associated with
high risk of infection transmission and immunocompromise.
Moreover, the risk of transmission of infection remains,
regardless of patient baseline characteristics. Furthermore,
our cost calculations are primarily relevant to the peri-
operative setting, but the published data used come from
outpatient bronchoscopic and critical care settings, and
therefore there may be some discrepancy that is
unaccounted for due to this assumption. There was
heterogeneity in the lengths of investigations (1–
23 months), patient contamination rates (4.6–58%) and
infection risk (0–100%) among included studies. The
variation was partly accounted for by utilising a fixed-effects
model. When varying risks of patient cross-contamination
and infection in the sensitivity analysis, single-use flexible
bronchoscopes remained financially superior to reusable
flexible bronchoscopes. In addition, one of the assumptions
that were made was that the risk of infection of single-use
flexible bronchoscopes was 0%. However, there always
remain unaccounted-for risks of infection transmission, even
with single-use flexible bronchoscopes, due to unsterile
operators and surrounding equipment or re-use in the same
patient [71]. Finally, there remains a previously unreported
risk of nosocomial infection from tracheal intubation in
elective patients who are not critically unwell and who are
intubated for a limited duration of time, which might
account for some of the incidence of infection after
bronchoscopy. However, given that there are virtually no
data reporting nosocomial infection in these settings, this
may have a limited impact on our data. Moreover, the data
we used on the published incidence of infection could be
secondary to tracheal intubation rather than the use of a
flexible bronchoscope. However, given that only 44 of the
included patients with evidence of infection were in critical
care settings, it is unlikely that the incidence of infection
could be primarily attributed to non-bronchoscopic
sources.
In conclusion, our systematic review has demonstrated
that the risk of patient infection following bronchoscopy
with reusable flexible bronchoscopes is significant,
warranting a need for guidelines on reprocessing to be
stricter to ensure greater patient safety [8, 9, 72]. The total
538 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Mouritsen et al. | Cost effectiveness of flexible bronchoscopes
cost per use of a reusable flexible bronchoscope was
calculated to be £249. The cost per use of a single-use
flexible bronchoscope was £220 sterling. When considering
the risk of infection in the cost analysis, reusable flexible
bronchoscopes have a mean cost per patient of £511
sterling and an associated risk of infection at 2.8%. The
findings from this study suggest benefits of single-use
flexible bronchoscopes in terms of cost effectiveness, cross-
contamination and resource utilisation.
Acknowledgements
This review was registered on PROSPERO (CRD420191-
34573). KE is an Editor of Anaesthesia. IA and KE have
previously received honoraria for consulting for Ambu, but
there was no involvement of any industry in any aspect of
this study, nor has anyone bar the authors had sight of the
results of this study before submission for publication. No
external funding or competing interests declared for JM, LE
and JK.
References
1. Ernst A, Silvestri GA, Johnstone D. Interventional pulmonary
procedures. Chest 2003; 123: 1693–4.
2. Van Wicklin SA, Conner R, Spry C. Guideline for processing
flexible endoscopes. In: Connor R, ed. 2016 Guidelines for
Perioperative Practice. Denver, CO: Association of Peri-
operative Registered Nurses, 2016. 675–758.
3. Society of Gastroenterology Nurses and Associates. Standard
of infection prevention in the gastroenterology setting.
2015. https://www.sgna.org/Portals/0/Standard of Infection
Prevention_FINAL_2.22.pdf?ver = 2016-02-22-153052-487 (acc-
essed 06/05/2019).
4. Society of Gastroenterology Nurses and Associates. Standards
of infection prevention in reprocessing flexible gastrointestinal
endoscopes. 2015. https://www.sgna.org/Portals/0/Education/
PDF/Standards-Guidelines/sgna_stand_of_infection_control_
0812_FINAL.pdf (accessed 06/05/2019).
5. Advancing Safety in Healthcare Technology. American National
Standard - Flexible and semi-rigid endoscope processing in
health care facilities. 2015. https://my.aami.org/aamiresources/
previewfiles/ST91_1504_preview.pdf (accessed 06/05/2019).
6. Kovaleva J, Meessen N, Peters F, et al. Is bacteriologic
surveillance in endoscope reprocessing stringent enough?
Endoscopy 2009; 41: 913–16.
7. Culver DA, Gordon SM, Mehta AC. Infection control in the
bronchoscopy suite: a review of outbreaks and guidelines for
prevention. American Journal of Respiratory and Critical Care
Medicine 2003; 167: 1050–6.
8. Kovaleva J, Peters FTM, van der Mei HC, Degener JE.
Transmission of infection by flexible gastrointestinal endoscopy
and bronchoscopy. Clinical Microbiology Reviews 2013; 26:
231–54.
9. Ofstead CL, Quick MR, Wetzler HP, et al. Effectiveness of
reprocessing for flexible bronchoscopes and endobronchial
ultrasound bronchoscopes. Chest 2018; 154: 1024–34.
10. Terjesen CL, Kovaleva J, Ehlers L. Early assessment of the likely
cost effectiveness of single-use flexible video bronchoscopes.
PharmacoEconomics – Open 2017; 1: 133–41.
11. Bou R, Aguilar A, Perpi~ n J, et al. Nosocomial outbreak of
na
Pseudomonas aeruginosa infections related to a flexible
bronchoscope. Journal of Hospital Infection 2006; 64: 129–35.
Mouritsen et al. | Cost effectiveness of flexible bronchoscopes
12. Silva CV, Magalh~ aes VD, Pereira CR, Kawagoe JY, Ikura C, Ganc
AJ. Pseudo-outbreak of Pseudomonas aeruginosa and Serratia
marcescens related to bronchoscopes. Infection Control and
Hospital Epidemiology 2003; 24: 195–7.
13. Mehta AC, Prakash UBS, Garland R, et al. American College of
Chest Physicians and American Association for Bronchoscopy
Consensus Statement: prevention of flexible bronchoscopy-
associated infection. Chest 2005; 128: 1742–55.
14. Srinivasan A. Epidemiology and prevention of infections
related to endoscopy. Current Infectious Disease Reports 2003;
5: 467–72.
15. Mughal MM, Minai OA, Culver DA, Mehta AC. Reprocessing
the bronchoscope: the challenges. Seminars in Respiratory and
Critical Care Medicine 2004; 25: 443–9.
16. Ramsey AH, Oemig TV, Davis JP, Massey JP, T€ or€ok TJ.
An outbreak of bronchoscopy-related Mycobacterium
tuberculosis infections due to lack of bronchoscope leak
testing. Chest 2002; 121: 976–81.
17. Larson JL, Lambert L, Stricof RL, Driscoll J, McGarry MA, Ridzon
R. Potential nosocomial exposure to Mycobacterium
tuberculosis from a bronchoscope. Infection Control and
Hospital Epidemiology 2006; 24: 825–30.
18. Tvede MF, Kristensen MS, Nyhus-Andreasen M. A cost analysis
of reusable and disposable flexible optical scopes for
intubation. Acta Anaesthesiologica Scandinavica 2012; 56:
577–84.
19. Perbet S, Blanquet M, Mourgues C, et al. Cost analysis of
single-use (Ambuâ aScopeTM) and reusable bronchoscopes in
the ICU. Annals of Intensive Care 2017; 7: 3.
20. Gupta D, Wang H. Cost-effectiveness analysis of flexible optical
scopes for tracheal intubation: a descriptive comparative study
of reusable and single-use scopes. Journal of Clinical
Anesthesia 2011; 23: 632–5.
21. El-Boghdadly K, Onwochei DN, Cuddihy J, et al. A prospective
cohort study of awake fibreoptic intubation practice at a tertiary
centre. Anaesthesia 2017; 72: 694–703.
22. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, John
PA. The PRISMA statement for reporting systematic reviews and
meta-analyses of studies that evaluate healthcare interventions:
explanation and elaboration. British Medical Journal 2009;
339: b2700.
23. Sabat AJ, Budimir A, Nashev D, et al. Overview of molecular
typing methods for outbreak detection and epidemiological
surveillance. European Communicable Disease Bulletin 2013;
18: 20380.
24. Weinstein MC, O’Brien B, Hornberger J, et al. Principles of
good practice for decision analytic modeling in health-care
evaluation: report of the ISPOR Task Force on Good Research
Practices-Modeling Studies. Value in Health 2003; 6: 9–17.
25. Ofstead CL, Quick MR, Eiland JE, Adams SJ. A glimpse at the
true cost of reprocessing endoscopes. 2017. https://www.
bostonscientific.com/content/dam/bostonscientific/uro-wh/
portfolio-group/LithoVue/pdfs/Sterilization-Resource-Handout.
pdf (accessed 19/09/2019).
26. Miskovic A, Lumb AB. Postoperative pulmonary complications.
British Journal of Anaesthesia 2017; 118: 317–34.
27. Brueckmann B, Villa-Uribe JL, Bateman BT, et al. Development
and validation of a score for prediction of postoperative
respiratory complications. Anesthesiology 2013; 118:
1276–85.
28. Corne P, Godreuil S, Jean-Pierre H, et al. Unusual implication of
biopsy forceps in outbreaks of Pseudomonas aeruginosa
infections and pseudo-infections related to bronchoscopy.
Journal of Hospital Infection 2005; 61: 20–6.
29. Pappas SA, Schaaff DM, DiCostanzo MB, King FW, Sharp JT.
Contamination of flexible fiberoptic bronchoscopes. American
Review of Respiratory Disease 1983; 127: 391–2.
© 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Anaesthesia 2020, 75, 529–540
30. Botana-Rial M, Leiro-Fern andez V, N ~ ez-Delgado M, et al. A
un
pseudo-outbreak of Pseudomonas putida and Stenotrophomonas
maltophilia in a bronchoscopy unit. Respiration 2016; 92: 274–8.
31. Wang HC, Liaw YS, Yang PC, Kuo SH, Luh KT. A
pseudoepidemic of Mycobacterium chelonae infection caused
by contamination of a fibreoptic bronchoscope suction
channel. European Respiratory Journal 1995; 8: 1259–62.
32. Guy M, Vanhems P, Dananch e C, et al. Outbreak of pulmonary
Pseudomonas aeruginosa and Stenotrophomonas maltophilia
infections related to contaminated bronchoscope suction
valves, Lyon, France, 2014. Eurosurveillance 2016; 21: 30286.
33. Srinivasan A, Wolfenden LL, Song X, et al. An outbreak of
Pseudomonas aeruginosa Infections associated with flexible
bronchoscopes. New England Journal of Medicine 2003; 348:
221–7.
34. Kirschke DL, Jones TF, Craig AS, et al. Pseudomonas
aeruginosa and Serratia marcescens contamination associated
with a manufacturing defect in bronchoscopes. New England
Journal of Medicine 2003; 348: 214–20.
35. DiazGranados Ca, Jones MY, Kongphet-Tran T, et al. Outbreak
of Pseudomonas aeruginosa infection associated with
contamination of a flexible bronchoscope. Infection Control
and Hospital Epidemiology 2009; 30: 550–5.
36. Sammartino MT, Israel RH, Magnussen CR. Pseudomonas
aeruginosa contamination of fiberoptic bronchoscopes.
Journal of Hospital Infection 1982; 3: 65–71.
37. Anderson DJ, Kirkland KB, Kaye KS, et al. Underresourced
hospital infection control and prevention programs: penny
wise, pound foolish? Infection Control and Hospital
Epidemiology 2007; 28: 767–73.
38. Arefian H, Heublein S, Scherag A, et al. Hospital-related cost of
sepsis: a systematic review. Journal of Infection 2017; 74: 107–
17.
39. Sato R, Gomez Rey G, Nelson S, Pinsky B. Community-acquired
pneumonia episode costs by age and risk in commercially
insured US adults aged ≥ 50 years. Applied Health Economics
and Health Policy 2013; 11: 251–8.
40. Briggs AH, Weinstein MC, Fenwick EAL, Karnon J, Sculpher MJ,
Paltiel AD. Model parameter estimation and uncertainty
analysis: a report of the ISPOR-SMDM modeling good research
practices task force working group-6. Medical Decision Making
2012; 32: 722–32.
41. Anon. Bronchoscopy-related infections and pseudoinfections –
New York, 1996 and 1998. Morbidity and Mortality Weekly
Report 1999; 48: 557–60.
42. Schelenz S, French G. An outbreak of multidrug-resistant
Pseudomonas aeruginosa infection associated with
contamination of bronchoscopes and an endoscope washer-
disinfector. Journal of Hospital Infection 2000; 46: 23–30.
43. Kressel A, Kidd F. Pseudo-outbreak of Mycobacterium
chelonae and Methylobacterium mesophilicum caused by
contamination of an automated endoscopy washer. Infection
Control and Hospital Epidemiology 2000; 21: 394–7.
44. Bennett SN, Peterson DE, Johnson DR, Hall WN, Robinson-
Dunn B, Dietrich S. Bronchoscopy-associated Mycobacterium
xenopi pseudoinfections. American Journal of Respiratory and
Critical Care Medicine 1994; 150: 245–50.
45. Webb SF, Vall Spinosa A. Outbreak of Serratia marcescens
associated with the flexible fiberbronchoscope. Chest 1975;
68: 703–8.
46. Siegman-igra Y, Inbar G, Campus A. An, “outbreak” of
pulmonary pseudoinfection by Serratia marcescens. Journal of
Hospital Infection 1985; 6: 218–20.
47. Vandenbroucke-Grauls CMJE, Baars ACM, Visser MR, Hulstaert
PF, Verhoef J. An outbreak of Serratia marcescens traced to a
contaminated bronchoscope. Journal of Hospital Infection
1993; 23: 263–70.
539
Anaesthesia 2020, 75, 529–540
48. Goldstein B, Abrutyn E. Pseudo-outbreak of Bacillus species:
related to fibreoptic bronchoscopy. Journal of Hospital
Infection 1985; 6: 194–200.
49. Hagan ME, Klotz SA, Bartholomew W, Potter L, Nelson M. A
pseudoepidemic of Rhodotorula rubra: a marker for microbial
contamination of the bronchoscope. Infection Control and
Hospital Epidemiology 1995; 16: 727–8.
50. Machida H, Seki M, Yoshioka N, et al. Correlation between
outbreaks of multidrug-resistant Pseudomonas aeruginosa
infection and use of bronchoscopes suggested by
epidemiological analysis. Biological and Pharmaceutical
Bulletin 2014; 37: 26–30.
51. Peaper DR, Havill NL, Aniskiewicz M, et al. Pseudo-outbreak of
Actinomyces graevenitzii associated with bronchoscopy.
Journal of Clinical Microbiology 2015; 53: 113–17.
52. Guimar~ aes T, Chimara E, do Prado GVB, et al. Pseudooutbreak
of rapidly growing mycobacteria due to Mycobacterium
abscessus subsp bolletii in a digestive and respiratory
endoscopy unit caused by the same clone as that of a
countrywide outbreak. American Journal of Infection Control
2016; 44: e221–6.
53. Sorin M, Segal-Maurer S, Mariano N, Urban C, Combest A,
Rahal JJ. Nosocomial transmission of imipenem-resistant
pseudomonas aeruginosa following bronchoscopy associated
with improper connection to the steris system I processor.
Infection Control and Hospital Epidemiology 2001; 22: 409–13.
54. Peltroche-Llacsahuanga H, L€ utticken R, Haase G. Temporally
overlapping nosocomial outbreaks of Serratia marcescens
infections: an unexpected result revealed by pulsed-field gel
electrophoresis. Infection Control and Hospital Epidemiology
1999; 20: 387–8.
55. Barton E, Borman A, Johnson E, Sherlock J, Giles A. Pseudo-
outbreak of Fusarium oxysporum associated with
bronchoscopy. Journal of Hospital Infection 2016; 94: 197–8.
56. Raz DJ, Nelson RA, Grannis FW, Kim JY. Natural history of
typical pulmonary carcinoid tumors: a comparison of
nonsurgical and surgical treatment. Chest 2015; 147: 1111–17.
57. Nelson KE, Larson PA, Schraufnagel DE, Jackson J.
Transmission of tuberculosis by flexible fiberbronchoscopes.
American Review of Respiratory Disease 1983; 127: 97–100.
58. Wheeler PW, Lancaster D, Kaiser AB. Bronchopulmonary cross-
colonization and infection related to mycobacterial
contamination of suction valves of bronchoscopes. Journal of
Infectious Diseases 1989; 159: 954–8.
59. Agerton T, Valway S, Gore B, et al. Transmission of a highly
drug-resistant strain (strain W1) of Mycobacterium
tuberculosis. Community outbreak and nosocomial
transmission via a contaminated bronchoscope. Journal of the
American Medical Association 1997; 278: 1073–7.
60. Prigogine T, Glupczynski Y, Van Molle P, Schmerber J.
Mycobacterial cross-contamination of bronchoscopy
specimens. Journal of Hospital Infection 1988; 11: 93–5.
540 © 2019 The Authors. Anaesthesia published by John Wiley & Sons Ltd on behalf of Association of Anaesthetists
Mouritsen et al. | Cost effectiveness of flexible bronchoscopes
61. Uttley AH, Honeywell KM, Fitch LE, Yates MD, Collins CH,
Simpson RA. Cross contamination of bronchial washings.
British Medical Journal 1990; 301: 1274.
62. Takigawa K, Fujita J, Negayama K, et al. Eradication of
contaminating Mycobacterium chelonae from
bronchofibrescopes and an automated bronchoscope
disinfection machine. Respiratory Medicine 1995; 89: 423–7.
63. Chroneou A, Zimmerman SK, Cook S, et al. Molecular typing of
mycobacterium chelonae isolates from a pseudo-outbreak
involving an automated bronchoscope washer. Infection
Control and Hospital Epidemiology 2008; 29: 1088–90.
64. Waite TDD, Georgiou A, Abrishami M, Beck CRR. Pseudo-
outbreaks of Stenotrophomonas maltophilia on an intensive
care unit in England. Journal of Hospital Infection 2016; 92:
392–6.
65. Nye K, Chadha DK, Hodgkin P, Bradley C, Hancox J, Wise R.
Mycobacterium chelonei isolation from broncho-alveolar
lavage fluid and its practical implications. Journal of Hospital
Infection 1990; 16: 257–61.
66. Campagnaro RL, Teichtahl H, Dwyer B. A pseudoepidemic of
Mycobacterium chelonae: contamination of a bronchoscope
and autocleaner. Australian and New Zealand Journal of
Medicine 1994; 24: 693–5.
67. C^etre JC, Nicolle M-CC, Salord H, et al. Outbreaks of
contaminated broncho-alveolar lavage related to intrinsically
defective bronchoscopes. Journal of Hospital Infection 2005;
61: 39–45.
68. Blanc DS, Parret T, Janin B, Raselli P, Francioli P. Nosocomial
infections and pseudoinfections from contaminated
bronchoscopes: two-year follow up using molecular markers.
Infection Control and Hospital Epidemiology 1997; 18: 134–6.
69. McCahon RA, Whynes DK. Cost comparison of re-usable and
single-use fibrescopes in a large English teaching hospital.
Anaesthesia 2015; 70: 699–706.
70. Lilholt Sørensen B, Gr€ uttner H. Comparative study on
environmental impacts of reusable and single-use
bronchoscopes. American Journal of Environmental Protection
2019; 7: 55.
71. McGrath BA, Ruane S, McKenna J, Thomas S. Contamination of
single-use bronchoscopes in critically ill patients. Anaesthesia
2017; 72: 36–41.
72. Rutala WA, Weber DJ. Gastrointestinal endoscopes a need to
shift from disinfection to sterilization? Journal of the American
Medical Association 2014; 312: 1405–6.
Supporting Information
Additional supporting information may be found online via
the journal website.
Data S1. Micro-costing analysis.
Data S2. Data prediction.
Anaesthesia 2019 doi:10.1111/anae.14916

Review Article

Learning from the law. A review of 21 years of litigation

for nerve injury following central neuraxial blockade in
obstetrics
K. McCombe1,2 and D. G. Bogod3

1 Consultant, Department of Anaesthesia, Mediclinic City Hospital, Dubai Healthcare City, Dubai, UAE
2 Adjunct Associate Professor, Mohammed Bin Rashid University, Dubai, UAE
3 Consultant, Department of Anaesthesia, Nottingham University Hospitals NHS Trust, Nottingham, UK

Summary
Medicolegal claims for neurological injury following the use of central neuraxial blockade in childbirth
represent the second most common claim against obstetric anaesthetists. We present an analysis of 55 cases
from a database of 368 obstetric anaesthetic claims. Common themes that emerge from the analysis include:
consent; nature of nerve injury (non-anaesthetic; direct; chemical; compressive); recognition; and
management. Specific advice arising from these cases includes: the importance of informing patients of the
risks of nerve damage; keeping below the conus of the cord for intrathecal procedures; responding
appropriately if a patient complains of paraesthesia; and having a high index of suspicion if recovery of normal
neurological function is delayed. As ever, principles of good practice, including respect for patient autonomy,
early provision of information, good communication and a high standard of record-keeping, will minimise the
frustration of patients that can then lead them to seek a legal route to redress if they suffer an injury following
central neuraxial blockade.

.................................................................................................................................................................
Correspondence to: D. G. Bogod
Email: david.bogod@me.com
Accepted: 15 October 2019
Keywords: ethical principles; autonomy; pregnancy
Twitter: @BogodDavid

Introduction The Third National Audit Project (NAP3) of the Royal

In this article, the second in our ‘Learning from the Law’ College of Anaesthetists confirmed that central neuraxial
series [1], we draw again from the database of over 360 blockade in the obstetric population are ‘very safe’ and that,
negligence claims relating to obstetric anaesthesia for “the risk balance of regional techniques in the obstetric
which DB has acted as an expert witness to the courts population is so far tipped towards the benefit side of the
between 1994 and 2015. This time we focus on nerve injury equation, that no sensible commentator would argue
following central neuraxial blocks. against its continued use” [2]. However, although
Over 700,000 central neuraxial blocks are carried out in complication rates are very low, when complications do
UK hospitals each year, 45% of which are performed for occur they can be devastating. Even more so perhaps
obstetric indications [2]. Approximately 25% of women who because the women affected are usually healthy before
labour in the UK will choose epidural analgesia (around anaesthetic intervention and, not unreasonably, expect a
140,000 per year) and 92% of caesarean sections are carried positive birth experience free from long-term
out under central neuraxial blockade. consequences.

© 2019 Association of Anaesthetists 1

Anaesthesia 2019 McCombe and Bogod | Litigation after obstetric nerve injury

As defined more fully in our previous article [1], another process (i.e. childbirth)’ [7]. However, although
medical negligence is a civil tort that occurs when a the mother’s signature is not required, she must be
patient suffers harm due to the (in)actions of her apprised of all material risks and the details of this
doctor. To succeed, the claimant must prove that, on conversation must be documented in the patient record.
the balance of probability: Regardless of the legal position, many hospitals insist on
written consent, and anaesthetists should follow local
1 the doctor owed her a duty of care
protocols.
2 the doctor’s practice fell below an acceptable standard
In our database, the information provided to
[3, 4]
patients relating to neuraxial anaesthesia was deemed
3 she suffered harm as a direct consequence of the
to be inadequate and consequently negligent in 8 out
substandard care
of 55 cases (15%). While this figure represents the
Neurological injury is the second most common minority it is important to appreciate that this would be
claim against obstetric anaesthetists after inadequate a far greater issue if today’s legal standards were
regional anaesthesia resulting in pain during caesarean applied to historical claims. In more paternalistic times
delivery. However, despite relatively fewer claims for when ‘doctor knew best’, it was common practice not to
nerve injury, the associated cost is considerably warn of the risk of nerve damage, given its rarity.
greater, reflecting the spectrum of potential injury However, in today’s post-Montgomery era, this is
from mild, temporary paraesthesia, to devastating considered wholly unacceptable [8].
paraplegia [5]. The UK Supreme Court ruled in Montgomery that
These national figures were reflected in our database patients must be warned of ‘all material risks’ and defined a
where 55 out of the 368 (15%) negligence claims involved a material risk as any to which, ‘that particular patient might
case of neurological injury following regional anaesthesia, attach significance’, no matter how unlikely it is to
compared with 76 (21%) claims for pain during caesarean materialise. Therefore, with regard to nerve injury, we
section. Anaesthetic practice was assessed by DB to have should counsel the patient of the following: temporary
been negligent in 141 out of 368 (38%) cases in the obstetric nerve damage 1:1000 (rare); effects lasting > 6 months
database overall, and in 25 out of the 55 cases involving 1:13,000 (rare); and severe injury, including paralysis
nerve damage (46%). 1:250,000 (very rare) [9].
As stated earlier, certain common themes emerged These data, derived from NAP3, are the best
from this series of claims. These were consent; nerve injury available given the difficulties we face when estimating
(non-anaesthetic causes of nerve injury; direct trauma to the risk involving very low probabilities. It can be equally
nerve; chemical injury; compressive injury); and recognition difficult to set risk in a meaningful context for a specific
and management of complications. patient [10], and this challenge may be even more
apparent in the obstetric setting where, “drugs, fatigue,
pain or anxiety may compromise the capacity of the
Theme 1: consent
parturient” [7]. In the eyes of the law [11] and of the
The consent process involves the two-way exchange of
Association of Anaesthetists, which echoes the legal
information. To fulfil the requirements of genuinely
position in its guidance [7], the influence of pain and
informed consent (i.e. not merely imparting information,
drugs does not cause the mother to lose capacity
but allowing time for reflection and discussion), this
except in the most exceptional of circumstances.
process should start in the antenatal setting and should
Consequently, information should be shared and
not be confined to labour. There are obvious systemic
consent sought in the normal way.
and cultural barriers to this occurring routinely, but
Clinicians may be dubious on learning that
resources such as the Obstetric Anaesthetists’
women who labour are regarded legally as having
Association’s information leaflets [6] are readily available
capacity in all but the most extreme situations. To
to help in this process.
consider otherwise would be to enter a world of loss
The quality of consent is not enhanced by gaining
of autonomy for pregnant women in labour, a
the patient’s signature on a form and for this reason,
position that society would almost certainly not be
written consent is not a legal requirement for obstetric
prepared to tolerate.
central neuraxial blocks because it is given ‘to facilitate

2 © 2019 Association of Anaesthetists

McCombe and Bogod | Litigation after obstetric nerve injury
Consent and birth plans
A birth plan is a statement of a woman’s wishes and
values. However, birth is an epistemically transformative
process and so a woman may specify that she does not
want epidural anaesthesia in her birth plan and go on
to change her mind once labour starts. A woman with
capacity (i.e. the vast majority, as described above) is
not bound by her previous statement and has the
absolute right to alter her views in the face of her new
knowledge and experience, or even for no reason at all.
In this situation, the epidural should be sited in the
usual way.
If a capacitous woman refuses to give consent to
any procedure during labour, this refusal must be
respected. Should a woman genuinely lose capacity, the
Association of Anaesthetists advises that the birth plan
be viewed as an advance decision [12]. Therefore, if
there is real conviction that the mother has lost her
capacity, the anaesthetist should abide by her expressed
wishes.
During the insertion of combined spinal-epidural
anaesthesia for caesarean section, Mrs A was screaming
in pain which she felt down her lower back and left leg
with each of several attempts. She cried out for a general
anaesthetic; the anaesthetist told her ‘she would be
grateful in the end.’
As well as changing her mind to request neuraxial
anaesthesia, a woman is free to withdraw consent at
any time during the process. Should the woman ask
the anaesthetist to stop, they must obey, discuss her
wishes with her and respect her subsequent decision,
whether or not they agree with it. “A mentally
competent patient has the absolute right to refuse to
consent to medical treatment for any reason, rational or
irrational, or no reason at all” [13]. Under UK law, the
fetus has no rights until the moment of birth and so
any perceived fetal interests do not trump the wishes
of the mother [11].
Theme 2: nerve injury
This theme can be further subdivided according to the
mechanism of nerve damage; non-anaesthetic causes of
nerve injury; direct nerve trauma; chemical injury
(arachnoiditis); and compressive injury (epidural abscess or
vertebral canal haematoma).
© 2019 Association of Anaesthetists
Anaesthesia 2019
Non-anaesthetic causes of nerve injury
Mrs B was admitted for induction in her second
pregnancy at 10 days past term. An epidural was
inserted easily and uneventfully. The epidural, which was
initially effective, needed topping up on three separate
occasions due to pain in the right groin, before the
delivery of a 4.5-kg baby in the occipito-posterior (OP)
position.
Following delivery, Mrs B complained of numbness
in her right leg and was found to have mild sensory and
motor deficit in the distribution of the femoral nerve (L2,
3 and 4). The MRI was normal and nerve conduction
studies confirmed a peripheral nerve lesion either of the
femoral nerve or of the roots supplying it. The
neurophysiologist concluded:
“This is probably a complication of the epidural with
the anaesthetic being introduced into the subarachnoid
space and pooling to produce both sacral and lumbar
root damage.”
Anaesthetists at the defendant hospital disagreed,
citing birth trauma as the most likely cause; these views
were dismissed until expert opinion was sought. DB
agreed that the epidural was not responsible for the
nerve damage. The Tuohy needle would have had to
contact multiple nerve roots for it to have caused the
demonstrated neuropathy. This is not anatomically
feasible in the context of a single puncture and
uncomplicated epidural insertion, during which no pain
or paraesthesia was elicited.
The argument for drugs ‘pooling’ in the
subarachnoid space was spurious because there was no
evidence of dural puncture; no evidence of spinal block;
and a mixture of bupivacaine and fentanyl was used.
These drugs are injected routinely into the subarachnoid
space and do not cause nerve root damage. DB
postulated that the damage was caused by compression
of the nerve roots in the pelvis by the fetal head. This
argument was supported by the following facts; the
baby was large and in the OP position, thus enlarging
further the head diameter entering the pelvis; the patient
felt pain in the right groin during the second stage and
this fits with compression of the upper right lumbar
nerve roots; and the lesion fits very well the symptoms
and signs of nerve root compression [14].
The claim against the anaesthetist was subsequently
rejected.
3
Anaesthesia 2019
Neurological lesions can result solely from the process
of childbirth. Bromage estimated that pelvic neural
compression occurs in 1 in 3000 deliveries, compression of
nutrient arteries in 1 in 15,000 women and problems arising
from arteriovenous malformations 1 in 20,000 times. These
figures combined to make an incidence of postpartum
neurological complications relating to obstetric causes of 1
in 2000 deliveries [15]. A prospective audit of all postpartum
women found an incidence of 1 in 2530 for a neurological
deficit lasting longer than 6 weeks, with epidural
considered contributory in only 1 in 13,000 [16]. A recent
prospective French study found an incidence of postpartum
neuropathy of 0.3%, with 84% of lesions being in the femoral
nerve territory and 69% resolving within six weeks [17]. This
strongly suggests that childbirth by itself is a more common
mechanism of nerve injury than neuraxial anaesthesia. In
addition, positioning; instrumental delivery; ischaemic
injury to the nerves as a result of hypotension or obstruction
of the internal iliac arteries by the fetal head; or femoral
compression resulting from oedema in late pregnancy may
all result in postpartum neuropathies. Despite these
recognised causes, if a woman has received neuraxial
anaesthesia during childbirth, it seems that blame will often
first be directed at the anaesthetist. By way of emphasis, we
have a case in our series in which a woman suffering from
postpartum paraesthesia attempted a claim against the
anaesthetist although she had received neither epidural nor
spinal anaesthesia! It is often only when an anaesthetic
medicolegal opinion is sought that the obstetric nature of
the injury is appreciated.
Direct nerve trauma
With respect to direct nerve trauma, negligence is assessed
by considering two criteria. First, the level of insertion. When
inserting a spinal needle, it is incumbent upon anaesthetists
to take all reasonable precautions to ensure that the needle
tip enters the subarachnoid space at a point below the
termination of the spinal cord. Much has been written about
this in the literature and this body of work will inform the
opinion of any medical expert or judge assessing a case of
neurological damage. Second, the actions following
contact with a nerve. Should a correctly sited needle appear
to contact the cord or filum terminale, a medical expert will
examine the subsequent actions of the doctor to assess for
negligence.
The spinal cord terminates at or above the L1/2
interspace in the majority, but extends down to the
upper border of L2 in 43% of individuals and to its
lower border in up to 20%. Thus, a needle inserted in a
4 © 2019 Association of Anaesthetists
McCombe and Bogod | Litigation after obstetric nerve injury
cranial direction at the L2/3 interspace would,
theoretically, run the risk of hitting the cord in up to
20% of patients [18]. Tuffier’s line (also known as the
intercristal line) is commonly accepted as the landmark
for the lower border of the body of L4 or the L4/5
interspace. However, studies show that Tuffier’s line
intersects the midline at or above L2/3 interspace in 33–
51% of patients [19, 20]. This means that reliance on
Tuffier’s line can result in unintentionally high spinal
placement in a significant number of patients.
To compound this problem, a paper published in 2000
demonstrated that anaesthetists were inaccurate in their
identification of the lumbar spinal interspace at which a
marker was positioned. Only 29% identified the space
correctly. Of the remaining 71%, 68% thought the space was
lower than it actually was; 51% were one space out in their
estimate; 15.5% two spaces; 1% three spaces; and 0.5%
were four spaces out [21]. This inaccuracy cannot be
accounted for solely by the variability in Tuffier’s line and it
probably reflects a degree of overoptimistic assessment by
anaesthetists, who tend to find higher spaces technically
easier for insertion of spinal needles. The paper was
accompanied by an editorial in which Professor F. Reynolds
stated that “the L2/3 interspace should not be an option”
[22]. This opinion has become accepted as conventional
wisdom, and its implications for negligence claims are
commented upon by DB in his subsequent editorial;
“The anaesthetist who is unfortunate enough to hit
and damage a normally-terminating cord with a spinal
needle is likely to find himself in a difficult position
when it comes to a claim for medical negligence. With
the professional literature replete with papers
showing us the errors we tend to make when
identifying spinal levels and warnings about the risk of
placing a spinal needle too high, the Defence will be
on the back foot from the outset” [23].
An anaesthetist may believe the spinal needle entered
at a particular level but if cord damage should result, an MRI
scan is likely to reveal the true level of needle insertion; the
courts will accept radiological evidence over the level
documented. The use of ultrasound is not, at least at the
time of writing, a standard of care for the insertion of
neuraxial anaesthesia.
Of course, the dictat of needle insertion below L3
applies only to intrathecal (spinal) anaesthesia. It is
acceptable to insert an obstetric epidural at any level in
the lumbar, or even lower thoracic region, as we do not
intend to breach the dura with the Tuohy needle.
McCombe and Bogod | Litigation after obstetric nerve injury
Mrs C underwent elective caesarean for the delivery of her
fifth child. Spinal insertion proved difficult and three
attempts at L2/3 were documented, the last one causing
pain and paraesthesia, described as “immediate hot
pain. . .like having a red-hot poker pushed down both
legs”. Success was finally achieved at a level documented
as L3/4. The anaesthetist commented that the patient may
have an unusually low termination of the cord. Mrs C
subsequently developed paraesthesia, weakness and
urinary complications and an MRI performed several
months later showed a syrinx centred around T12/L1.
The anaesthetist’s estimation of level was judged to
be incorrect and the damage was considered, on the
balance of probability, to be caused by the spinal needle.
Causation was attributed to this error, which represented
substandard practice. The anaesthetist was found to be
negligent and the case was settled by the Trust.
Close attention must be paid to the patient’s response
while inserting the spinal/Tuohy needle and when injecting
the anaesthetic solution. Any complaint of tingling,
lancinating shocks or pain distant to the site of insertion
(especially in the legs) is highly suggestive of direct nerve
contact. If any of these signs is elicited, the anaesthetist must
stop. The needle should be partially or fully withdrawn, and
inserted at a different angle or different interspace. It is not
uncommon for the patient to experience fleeting, mild
paraesthesia during the threading of an epidural catheter;
this occurs frequently and is generally regarded as benign.
During spinal anaesthesia, injectate should be
administered only after confirming that cerebrospinal fluid
(CSF) flows freely from the needle hub as this suggests that
the tip is lying free in the subarachnoid space and is not
partially or fully embedded within nerve tissue. It is not
sufficient simply to see fluid in the hub as this could have
entered the needle during its passage through the CSF into
the nerve; the fluid must flow freely. Many practitioners
aspirate CSF from the spinal needle twice: once before
beginning injection and for a second time half-way through
to ensure that the tip remains free in the CSF.
The spinal procedure should be documented
meticulously, including the number of attempts, any symptoms
elicited, actions taken in light of these and the presence of
free-flowing CSF. Significant negatives should also be
recorded, for example, ‘No paraesthesia, no complications’.
Should nerve injury follow delivery, the absence of
paraesthesia or lancinating pain during insertion of the
neuraxial block should reassure the anaesthetist, as it is
highly unlikely for damage to be attributable to the needle/
© 2019 Association of Anaesthetists
Anaesthesia 2019
injectate in the absence of these signs. If appropriate
consent has been sought and given, a safe level of insertion
chosen and a recognised technique used, then contact with
the nerve is not, in itself, negligent providing the correct
remedial steps are taken following this complication.
Chemical injury – adhesive arachnoiditis
The link between neuraxial anaesthesia and arachnoiditis
has long been recognised. In 1949, Mr Woolley and Mr Roe
were both left paraplegic after undergoing spinal
anaesthesia for routine surgery on the same day, in the same
hospital, with the same anaesthetist. The Court found that
phenol, used to sterilise the local anaesthetic ampoules, had
penetrated the glass ampoules through microscopic cracks,
but it is now believed that the needles and syringes were
contaminated with descaling fluid, used to clean the
sterilising baths over the preceding weekend [24]. The
publicity surrounding the lawsuit led to a drastic decline in
the use of spinal anaesthesia in the UK until its resurgence as
a technique in the 1970s.
Rare though spinal/epidural related arachnoiditis may
be, our database contains at least two such cases, and
possibly as many as seven. One of these is the case of Mrs
Angelique Sutcliffe [25], which received damning coverage
in the media [26].
In 2001, Angelique Sutcliffe developed a progressive
and debilitating adhesive arachnoiditis after an
apparently uneventful spinal anaesthesia for elective
caesarean section, for which only hyperbaric
bupivacaine 0.5% was administered. The path of her
deterioration was steep and inexorable. Within a few
days she had severe back pain, with urinary retention
following shortly afterwards. Two weeks after delivery,
she had signs of raised intracranial pressure,
necessitating the insertion of a ventriculoperitoneal
shunt to treat obstructive hydrocephalus. She
developed worsening and ascending sensory and
motor neuropathy in her legs over the following weeks
and having undergone further surgery to treat recurrent
raised intracranial pressure, became progressively
paraplegic with limited use of her arms. Her magnetic
resonance imaging scans show a spinal cord severely
damaged as a result of multiple dense adhesions [27].
At the time of acting as expert witness to this case, DB
was unconvinced by the judge’s findings that the equipment
must, somehow, have become contaminated with
chlorhexidine solution; there did not appear to be a
plausible explanation as to how this happened. However, he
5
Anaesthesia 2019
retracted this view following the devastating paralysis
suffered in Australia by Grace Wang following the
accidental injection of 8 ml of chlorhexidine 2% in alcohol
via a Tuohy needle into her epidural space [27]. This mistake
occurred when, following a bloody tap, the saline in the loss-
of-resistance syringe was returned to the gallipot, staining
the remaining saline pink. This blood–saline mixture was
positioned next to the gallipot containing chlorhexidine,
thereby aligning the holes in the ‘Swiss Cheese’ to
catastrophic effect [28]. The error that befell Grace Wang
demonstrated unequivocally the devastating effects of
injecting chlorhexidine into the neuraxium and her
deterioration mirrored exactly that of Angelique Sutcliffe.
The Association of Anaesthetists has produced a safety
guideline for the use of chlorhexidine to achieve skin
antisepsis, the salient points of which are in Table 1 [29].
NHS England also issued a Patient Safety Alert in 2015,
warning against the practice of providing skin antisepsis
agents and solutions intended for injection in ‘open
systems’ (e.g. gallipots) in proximity to each other [30].
Summary of safety guideline: skin
antisepsis for central neuraxial
blockade [29]
Chlorhexidine in alcohol should be used for skin
antisepsis.
Meticulous care in taking measures to prevent
chlorhexidine from reaching the CSF:
1 Chlorhexidine should be kept well away from the
drugs and equipment and should not be poured into
containers on or near the same surface as the
equipment for central neuraxial blocks. Equipment
should be covered or protected while the antiseptic
is applied by swab, applicator or spray.
2 The solution must be allowed to dry before the skin is
palpated or punctured.
3 The operator should check his/her gloves for
contamination with chlorhexidine. If there is any
doubt, they should be changed before continuing
the procedure.
Given the lack of convincing evidence of the
antimicrobial superiority of a 2% solution of
chlorhexidine in alcohol over a 0.5% solution, but the
presence of clear evidence of the neurotoxicity of
chlorhexidine, the use of a 0.5% solution should be
preferred over a 2% solution for skin antisepsis before
central neuraxial blocks.
6 © 2019 Association of Anaesthetists
McCombe and Bogod | Litigation after obstetric nerve injury
Needless to say, practitioners and Trusts who choose to
ignore this pragmatic advice will reap the legal
consequences should harm befall their patients as a result.
Compressive injury
The risks of epidural abscess in the obstetric population are
quoted as 1 in 50,000 [9]. We have no cases of suspected or
confirmed epidural abscess in our database. Vertebral canal
haematoma is a rare complication of neuraxial blockade.
NAP3 estimated the risk in the obstetric population to be 1
in 170,000 [2]. We have one case of vertebral canal
haematoma in our database. It left the patient paraplegic
and serves as a stark reminder to remain ever vigilant for
post-anaesthetic complications.
Ms D was pregnant with her first child. She underwent
repair of a congenital cardiac abnormality as a child and,
in later life, insertion of spinal rods to correct scoliosis.
She was seen in the high-risk obstetric anaesthetic clinic
where it was agreed that epidural should be attempted
early in labour to maintain cardiovascular stability. She
was fully informed of all material risks and the consent
process was deemed satisfactory.
Insertion of the epidural was performed by a senior
consultant anaesthetist who repeated the consent
process before insertion. The epidural space was
located easily and, apart from transient paraesthesia on
threading of the catheter, which settled in the usual way,
the procedure was easy and unremarkable. The epidural
required several top ups during her protracted labour,
and Ms D’s baby was eventually delivered, by forceps
extraction, in the early hours of the following morning,
some 18 h after epidural insertion. The epidural
required a further top up to facilitate this procedure. The
epidural catheter was removed 2 h after delivery.
Ms D was reviewed on the morning anaesthetic
ward round, nearly 5 h after the last dose of anaesthetic
and 3 h after removal of the catheter. No assessment of
motor function was made during this visit as the
clinicians were focussed on her cardiovascular system.
Later that morning, it was noted by a midwife that Ms D
could not move her legs to allow examination but no
further action was taken.
Ms D mentioned to a passing anaesthetist that her
legs still felt heavy in the middle of the afternoon. She was
given reassurance, but not formally assessed. In the late
afternoon, 14 h after delivery, her vaginal pack was
removed by an obstetrician who noted her immobile legs
and requested a formal anaesthetic review. This took
place 90 min later, over 24 h after catheter removal.
McCombe and Bogod | Litigation after obstetric nerve injury
On assessment she was found to have full motor
block and was referred to the neurosurgeons for urgent
MRI. There was no out-of-hours MRI service at the
hospital where she was an inpatient and so she was
transferred to another facility. The MRI, performed 4 h
after anaesthetic assessment, revealed a vertebral canal
haematoma. She underwent emergency spinal
decompression that night but her motor function did not
recover and she remains paraplegic at the time of
writing.
All aspects of the anaesthetic management before and
during labour were deemed acceptable in the case of Ms D.
However, the anaesthetists and midwives failed in their duty
of care after delivery by failing to recognise and appreciate
the implications of the lack of block regression. This leads us
to our final theme.
Theme 3: recognition and management
of complications
Follow-up of patients must be timely and include inquiry
into the return of their motor and sensory function.
Regardless of the total dose of epidural anaesthesia
administered during labour and delivery, a block should
be regressing by 4 h after the last dose. This ‘4-h rule’
is referred to often in NAP3 [2] and will inform any
expert assessing a claim. Failure of block regression by
this time should alert the anaesthetist to the possibility
of vertebral canal pathology. Urgent MRI (the gold
standard for imaging suspected space-occupying
lesions) and referral to the neurosurgeons must be
considered.
Any patient who complains of pain or weakness
following neuraxial anaesthesia should be assessed by
an anaesthetist in a timely fashion. A full neurological
examination should be conducted to identify the
affected area and, if possible, to determine the cause of
injury. If injury appears to result from direct nerve
trauma (and does not represent an expanding lesion
requiring emergency management), then in-patient
referral to a neurologist for assessment is advisable.
Electrophysiological studies can distinguish between
central and peripheral nerve injury and can help in
prognostication. They can be useful in assessing the
longevity of nerve injury since new injuries take time to
evolve. Therefore, electromyography performed within
72 h of suspected injury will reveal old neuropathies,
but will not help in diagnosing new ones.
© 2019 Association of Anaesthetists
Anaesthesia 2019
Conclusion
Although rare, nerve damage sustained in the course of
neuraxial anaesthesia can be devastating. Financial awards
against anaesthetists and their employers can therefore be
concomitantly high. Although rare, the life-changing
neurological complications associated with neuraxial blocks
certainly constitute a ‘material risk’ and so each patient must
be fully apprised of these and give their consent freely.
Neurological complications can occur by chance even
in the most experienced and fastidious of hands, and nerve
injury does not necessarily imply negligence on the part of
the anaesthetist. However, given the propensity to blame
the anaesthetic for any abnormal postpartum neurology it is
prudent to be aware of this and to ensure that all entries into
the medical records are sufficiently detailed and
meticulous.
Acknowledgements
DB derives income from acting as a medicolegal expert,
accepting instructions from both Claimant and Defendant
solicitors. No external funding or competing interests
declared.
References
1. McCombe K, Bogod DG. Learning from the Law: a review of
21 years of litigation for pain during caesarean section.
Anaesthesia 2018; 73: 223–30.
2. Cook TM, Counsell D, Wildsmith JAW. Major complications
of central neuraxial block: report on the Third National Audit of
The Royal College of Anaesthetists. British Journal of
Anaesthesia 2009; 102: 179–90.
3. Bolam v Friern Hospital Management Committee 1957 2 All ER
118.
4. Bolitho v City Hackney Health Authority. 1998 AC 232.
5. Cook T, Bland L, Mihai R, Scott S. Litigation related to
anaesthesia: an analysis of claims against the NHS in England
1995–2007. Anaesthesia 2009; 64: 706–18.
6. Obstetric Anaesthetists’ Association. Information for Mothers.
www.labourpains.com (accessed 15/07/2019).
7. Association of Anaesthetists. AAGBI: consent for anaesthesia
2017. Anaesthesia 2017; 72: 93–105.
8. Montgomery v Lanarkshire Health Board [2015] UKSC 11,
2015All ER(D) 113 (Mar).
9. Obstetric Anaesthetists’ Association. Regional anaesthetic for
Caesarean Section – information card. 2012. www.labourpa
ins.com/assets/_managed/cms/files/InfoforMothers/REGIONAL-
ANAESTHESIA/Feb%2015%20Regional%20Anaesthesia%20for
%20CS%20English-1.pdf (accessed 15/07/2019).
10. Royal College of Anaesthetists. Risk and probability. https://
www.rcoa.ac.uk/patientinfo/risk-and-probability (accessed 15/
07/2019).
11. St George’s Healthcare NHS Trust v S; R v Collins, ex parte S
199844 BMLR 160 (CA).
12. Mental Capacity Act 2005. Section 24 – 26. 2005 https://www.le
gislation.gov.uk/ukpga/2005/9/part/1/crossheading/advance-
decisions-to-refuse-treatment (accessed 15/07/2019).
13. Re MB 1997EWCA Civ 3039.
7
Anaesthesia 2019
14. Russell IF. Postpartum neurological problems. In: Russell IF,
Lyons G, eds. Clinical problems in obstetric anaesthesia.
London, UK: Chapman and Hall, 1997: 150–3.
15. Bromage PR. Neurological complications of regional anesthesia
for obstetrics. In: Shnider SM, Levinson G, eds. Anesthesia for
obstetrics, 3rd edn. Baltimore, MD: Williams and Wilkins, 1993:
433–53.
16. Holdcroft A, Gibberd FB, Hargrove RL, Hawkins DF, Della-
portas CI. Neurological complications associated with
pregnancy. British Journal of Anaesthesia 1995; 75:
522–6.
17. Tournier A, Doremieux AC, Drumez E, et al. Lower limb
neurologic deficit after vaginal delivery; a prospective
observational study. International Journal of Obstetric
Anesthesia 2019. https://doi.org/10.1016/j.ijoa.2019.09.003
[Epub ahead of print].
18. Saifuddin A, Burnett SJ, White J. The variation of position of the
conus medullaris in an adult population. An MRI study. Spine
1998; 23: 1452–6.
19. Margarido CB, Mikhael R, Arzola C, Balki M, Carvalho J. The
intercristal line determined by palpation is not a reliable
anatomical landmark for neuraxial anesthesia. Canadian
Journal of Anesthesia 2011; 58: 262–5.
20. Locks Gde F, Almeida MC, Pereira AA. Use of the
ultrasound to determine the level of lumbar puncture in
pregnant women. Revista Brasileira de Anestesiologia 2010;
60: 13–19.
8 © 2019 Association of Anaesthetists
McCombe and Bogod | Litigation after obstetric nerve injury
21. Broadbent CR, Maxwell WB, Ferrie R, Wilson DJ, Gawne-Cain
M, Russel R. Ability of anaesthetists to identify a marked lumbar
interspace. Anaesthesia 2000; 55: 1106–26.
22. Reynolds F. Logic in the safe practice of spinal anaesthesia.
Anaesthesia 2000; 55: 1045–6.
23. Bogod DG. Keeping in the Reynold’s Zone. International
Journal of Obstetric Anaesthesia 2014; 23: 201–3.
24. Maltby JR, Hutter CD, Clayton KC. The Wooley and Roe Case.
British Journal of Anaesthesia 2000; 84: 121–6.
25. Sutcliffe v Aintree Hospitals NHS Trust 2008EWCA Civ 179.
26. Mother paralysed after being injected with cleaning fluid will
win up to 5 million in damages Daily Mail. 2008. https://www.da
ilymail.co.uk/news/article-517296/Mother-paralysed-injected-
cleaning-fluid-childbirth-win-5 m-damages.html (accessed
27/03/2019).
27. Bogod DG. The sting in the tail: antiseptics and the neuraxis
revisited. Anaesthesia 2012; 67: 1305–9.
28. Reason J. Human error: models and management. British
Journal of Anaesthesia 2000; 320: 768–70.
29. Association of Anaesthetists Obstetric Anaesthetists Association,
Regional Anaesthesia UK, Association of Paediatric Anaesthetists
of Great Britain and Ireland. Safety guideline: skin antisepsis
for central neuraxial blockade. Anaesthesia 2014; 69: 1279–86.
30. NHS England. Patient safety alert – risk of death or severe harm
due to inadvertent injection of skin preparation solution. 2015.
https://www.england.nhs.uk/2015/05/psa-skin-prep-solution
(accessed 27/03/2019).
Anaesthesia 2020, 75, 549 doi:10.1111/anae.15024

Correction
This article corrects the Original Article, by O. Dransart-Ray
e, E. Roldi, L. Zieleskiewicz, P. G. Gulnot, F. Mojoli, S. Mongodi, B
Bouhemad, ‘Lung ultrasound for early diagnosis of postoperative need for ventilatory support: a prospective observational
study’ [1].

The Author Guidelines clearly state that any funding obtained and potential competing interests need to be declared including
fees for consultancy and teaching. It has been brought to the Editor in Chief’s attention that neither F. Mojoli nor S. Mongodi
provided such declarations. This correction rectifies this error. Failure to declare conflicts of interest is regarded as a serious
matter, as outlined by the Committee on Publication Ethics (COPE). These omissions have now been added into the
Acknowledgements.

‘Acknowledgements
The study was registered with ClinicalTrials.gov (NCT03289975) and the National Commission for Information Technology and
Civil Liberties ((2013-03-No 1938757 V0). LZ has received fees from General Electrics for ultrasound teaching. SM has received
fees for lectures from General Electrics. FM has received fees for lectures from General Electrics and Hamilton Medical. No other
competing interests declared.’

The online version of the paper has been updated with these corrections.

Reference
1. Dransart-Ray
e O, Roldi E, Zieleskiewicz L, et al. Lung ultrasound for early diagnosis of postoperative need for ventilatory support: a
prospective observational study. Anaesthesia 2020; 75: 202–9.

© 2020 Association of Anaesthetists 549

Anaesthesia 2020, 75, 550–559
Correspondence
Improving early warning scores – more data, better
validation, the same response
We thank Oglesby et al. for their interesting editorial [1]
accompanying our recent paper [2]. We value their
interpretation that a single standardised early warning score
(EWS) may not be applicable to all types of patients and are
grateful for their qualified support of the concept of future
population-specific EWS. We are also in agreement that
evidence-based models are preferable to consensus-based
scores and that external validation is imperative for all EWS.
We have a major issue with their criticism that we failed
to externally validate our score in a dataset separate from
that used to develop the score. Multiple external validation
techniques were undertaken and have been available in
electronic format in the online supplement since 3 July 2019
[2]. Each centre independently collected four separate
datasets. Although these datasets were ultimately merged
into one combined database, we retained the ability to use
separate databases for external validation purposes.
Multiple predictive performance and validation techniques
were utilised and these are summarised in the supporting
information which accompanies the online version of our
paper (Appendix 3: Table A3.4.) [3].
Debray et al. [4] define external validation as using new
participant-level data, external to those used for model
development, to examine whether the model’s predictions
are reliable in individuals from potential population(s) for
clinical use. Following that definition, we purposely held
back from using some centres to fit/develop the model so
that we could then use them to externally validate a model
fit for other centres. We even reported all possible
combinations of doing this, which were a sensible cut of the
dataset to avoid accusations of potential bias.
Given the revised evidence-based physiological
weightings, novel sub-division of oxygen therapy and
potential clinical utility of our app [5] in discriminating
Responses to previously published articles or letters may be made using the commenting feature on the journal website. You
may then be invited to submit a full letter. Correspondence on new topics should be submitted via Editorial Manager — accessed
via ‘Submit an Article’ on the journal homepage.
550
between ‘hard’ and ‘soft’ escalation threshold NEWS scores
(≥ 5 or ≥ 7), we believe ‘Improving EWS – more data, better
validation, a step in the right direction’ may have been a
more accurate choice of editorial title.
J. H. Mackay
J. W. Brand
Y. D. Chiu
S. S. Villar
Royal Papworth Hospital NHS Foundation Trust,
Cambridge, UK
Email: jonmackay@doctors.org.uk
No competing interests declared.
References
1. Oglesby KJ, Sterne JAC, Gibbison B. Improving early warning
scores – more data, better validation, the same response.
Anaesthesia 2020; 75: 149–51.
2. Chiu Y, Villar SS, Brand JW, et al. Logistic early warning scores to
predict death, cardiac arrest or unplanned intensive care unit re-
admission after cardiac surgery. Anaesthesia 2020; 75: 162–70.
3. Chiu Y, Villar SS, Brand JW, et al. Logistic early warning scores
to predict death, cardiac arrest or unplanned intensive care
unit re-admission after cardiac surgery. Anaesthesia 2019.
Supplementary Information: https://onlinelibrary.wiley.com/
action/downloadSupplement?doi=10.1111%2Fanae.14755&f
ile=anae14755-sup-0001-AppendixA1-A5.docx (accessed 27/
08/2019).
4. Debray TP, Vergouwe Y, Koffijberg H, et al. A new framework to
enhance the interpretation of external validation studies of
clinical prediction models. Journal of Clinical Epidemiology
2015; 68: 279–89.
5. Chiu YD, Villar SS, Mackay JH. Logistic early warning score app
for cardiac surgical patients. 2019. https://yidachiu.shinyapps.
io/vitalpac_log_ews_app (accessed 27/08/2019).
doi:10.1111/anae.14878
© 2020 Association of Anaesthetists
Correspondence Anaesthesia 2020, 75, 550–559

Improving early warning scores – more data, better

validation, the same response: a reply
We thank Mackay et al. [1] for their comments on our
editorial [2], which accompanied the article by Chiu et al. [3]
and agree with the definition of external validation that they
provide. In broad terms, internal validation is concerned
with reproducibility of a prediction model, whereas external
validation is concerned with the transportability of model
predictions to other settings and populations. Randomly
selected patients whose data are not included in the
training data used to develop the prediction model and
used for validation are similar to, and subject to, the same
sampling processes as patients in the development dataset.
Internal validation based on such data will address
overfitting due to selection of predictor variables and
parametrisation of their associations with the outcome, but
does not address transportability to different settings.
We applaud Mackay et al. for conducting a range of
validation exercises; given the importance of validation we
encourage authors and editors to include results of such
exercises in the main text of published papers rather than in
supplementary material. From our reading of
manuscript and supplementary material, it appears that no
population from a different source was used to externally
validate their model. The populations used for validation
were subsets of data from the same four hospitals used in
model development (some excluded from
development data set), either: (1) using three hospitals
other than Papworth; (2) using data from all four hospitals in
2017; and (3) using the temporally ‘last 10%’ of data from
each patient.
These discussions illustrate that there is a continuum
within the dichotomy ‘internal validation’ and ‘external
validation’ [4].
K. J. Oglesby
University Hospitals Bristol NHS Foundation Trust,
Bristol, UK
J. A. C. Sterne
B. Gibbison
University of Bristol,
Bristol, UK
Email: ben.gibbison@bristol.ac.uk
No competing interests declared.
References
1. Mackay JH, Brand JW, Chiu Y-D, Villar SS. Improving early
warning scores – more data, better validation, the same
response. Anaesthesia 2020; 75: 550.
the 2. Oglesby KJ, Sterne JAC, Gibbison B. Improving early warning
scores - more data, better validation, the same response.
Anaesthesia 2020; 75: 149–51.
3. Chiu Y-D, Villar SS, Brand JW, et al. Logistic early warning scores
to predict death, cardiac arrest or unplanned intensive care unit
re-admission after cardiac surgery. Anaesthesia 2020; 75: 162–
70.
the 4. Justice AC, Covinsky KE, Berlin JA. Assessing the generalizability
of prognostic information. Annals of Internal Medicine 1999;
130: 515–24.
doi:10.1111/anae.14899

© 2020 Association of Anaesthetists 551

Anaesthesia 2020, 75, 550–559 Correspondence

Sleep disturbances and residual neuromuscular blockade:

future research possibilities
We read with great interest the article by Christensson et al. pulmonary complications and critical respiratory events. In
[1] investigating how partial neuromuscular blockade this context, considering and understanding how
interferes with the regulation of breathing in patients with comorbidities may influence this can help, not only to
obstructive sleep apnoea (OSA). This was a physiological improve therapeutic approaches to this condition, but also
study performed in 10 unsedated male volunteers. These to prevent respiratory complications requiring tracheal re-
findings are very important, given the high prevalence of intubation and unplanned admission to critical care. Sleep
OSA and how it can influence the use of neuromuscular disturbances are particularly hazardous, can impair
blocking drugs and reversal agents. Additionally, it breathing control and, as the study itself demonstrates, may
highlights the importance of implementing 1 month of alter the effect of conditions such as residual neuromuscular
nightly home continuous positive airways pressure (CPAP) blockade. Broader studies, focusing on OSA and other sleep
treatment in OSA patients before major surgery. disorders, would increase the understanding of breathing
However, we should like to make some observations regulation changes in patients with residual neuromuscular
regarding the methodology used by Christensson et al. and blockade, including the dynamics of these conditions.
suggest how future studies might be designed in order to
improve knowledge in this area, as well as to extend the Acknowledgements
discussion to include other factors that may also influence ~o Fundo de
Our studies are supported by the Associacßa
postoperative breathing regulation in patients. Although  Pesquisa (AFIP). ST and MA also received
Incentivo a
the methodology was sound, of the 10 patients recruited, support from the Conselho Nacional de Desenvolvimento
only three completed the entire protocol and it would be Cientıfico e Tecnol
ogico (CNPq). No competing interests
interesting to replicate the protocol using a much larger declared.
sample. In a cohort study of 2646 patients with OSA, for
example, it was shown that patients who had not been
M. Medina
treated pre-operatively with CPAP (n=1465) were at
S. Tufik
increased risk for postoperative cardiopulmonary
M. L. Andersen
complications [2]. Universidade Federal de S~
ao Paulo,
In addition, according to the study by Christensson S~
ao Paulo, Brazil
et al., the patients were recruited and diagnosed by Email: ml.andersen12@gmail.com
specialists in sleep medicine using the Emblettaâ (Embla,
References
Broomfield, CO, USA) portable diagnostic system; In-lab 1. Christensson E, Ebberyd A, H ardemark Cedborg A, et al.
polysomnography might have been used instead as it takes Hypoxic ventilatory response after rocuronium-induced partial
into account more variables and is more effective in neuromuscular blockade in men with obstructive sleep apnoea.
Anaesthesia 2020; 75: 338–47.
diagnosing sleep disturbances than the Embletta. In 2. Abdelsattar ZM, Hendren S, Wong SL, Campbell DA Jr,
addition to obstructive apnoea syndrome, central apnoea Ramachandran SK. The impact of untreated obstructive sleep
apnea on cardiopulmonary complications in general and
syndrome and hypoventilation/hypoxaemia associated with
vascular surgery: a cohort study. Sleep 2015; 38: 1205–10.
sleep [3] may also be related to postoperative hypoxaemia 3. American Academy of Sleep Medicine. International
and might consequently interfere with breathing regulation Classification of Sleep Disorders. 3rd ed. Darien, IL: American
Academy of Sleep Medicine.
in postoperative patients with residual neuromuscular
blockade. doi:10.1111/anae.14912
Residual neuromuscular blockade is an important
postoperative condition and is largely associated with

552 © 2020 Association of Anaesthetists

Correspondence Anaesthesia 2020, 75, 550–559

Sleep disturbances and residual neuromuscular blockade:

future research possibilities: a reply
We thank Medina et al. [1] for their interest in our recent postoperative respiratory complications in obstructive
article [2]. We agree that it would be interesting to include a sleep apnoea patients are lacking and further research is to
larger sample of patients with obstructive sleep apnoea, be welcomed.
particularly after they have received an initial 3 months of
continuous positive airway pressure (CPAP) treatment.
E. Christensson
However, there were a large number of patients that did not
M. J. Fagerlund
continue with CPAP for a variety of reasons. A power Karolinska University Hospital and Karolinska Institutet,
analysis was performed before the start of the study Stockholm, Sweden
and indicated that a sample size of 10 patients was Email: eva.christensson@sll.se
sufficient. The methodology included advanced respiratory
measurements, taken over 4 h.
We chose to do the diagnostics with a portable home No competing interests declared.
â
monitoring system, (Embletta , Embla, Broomfield, CO,
USA), which is the standard method in Europe. The References
1. Medina M, Tufik S, Andersen ML. Sleep disturbances and
recordings were manually scored by physicians specialising residual neuromuscular blockade: future research possibilities.
in sleep medicine, including differentiation between Anaesthesia 2020; 75: 552.
2. Christensson E, Ebberyd A, H ardemark Cedborg A, et al.
obstructive sleep apnoea and central apnoea.
Hypoxic ventilatory response after rocuronium-induced partial
We fully agree with Medina et al. that partial neuro- neuromuscular blockade in men with obstructive sleep apnoea.
muscular blockade is a risk factor for pulmonary Anaesthesia 2020; 75: 338–47.
complications and other critical respiratory events. Studies
targeting the effects of residual neuromuscular blockade on doi:10.1111/anae.14930

© 2020 Association of Anaesthetists 553

Anaesthesia 2020, 75, 550–559
Future directions in regional anaesthesia: not just for the
cognoscenti
We thank Turbitt et al. for their recent editorial on
rationalising teaching of block diversity [1]. As regional
anaesthesia fellows at a large London teaching hospital,
we are both regional anaesthesia learners and,
increasingly, regional anaesthesia teachers. We applaud
the approach and, whilst the exact choice of blocks will
doubtless generate debate, we believe the overall strategy
of selecting a small number of blocks, and driving
competence among as many consultants and trainees as
possible, is compelling.
However, the scale of the challenge should not be
underestimated. We carried out a snapshot survey of
anaesthetic consultants in seven acute Trusts covering
10 hospitals in London. Of 139 anaesthetic consultants
at these Trusts, only 15 (10.8%) stated they could
teach all seven of the plan A blocks under ultrasound
guidance. Twenty-nine (20.9%) consultants felt they
were not confident they could teach any of them and
a further 12 (8.6%) felt they could teach only one of
them. Only 24 (17.3%) consultants felt confident they
could teach the erector spinae block (the newest of
the blocks on the list). Training anaesthetic trainees
and consultants in performing these blocks, in our
554
Correspondence
region at least, is likely to be challenging on the basis
of these results.
Our results are not meant to undermine the ‘plan A
blocks’ approach, but they do reflect the size of the problem
faced. We believe experts in the field, as well as relevant
societies and associations, should consider strategies to
drive wider competence and make regional anaesthetic and
analgesic approaches available for more patients.
T. W. Ashken
M. H. W. Thompson
University College Hospital,
London, UK
Email: tobyashken@gmail.com
No competing interests declared.
Reference
1. Turbitt LR, Mariano ER, El-Boghdadly K. Future directions in
regional anaesthesia: not just for the cognoscenti. Anaesthesia
2020; 75: 293–7.
doi:10.1111/anae.14938
© 2020 Association of Anaesthetists
Correspondence
Future directions in regional anaesthesia: a reply
Turbitt et al. [1] and Ashken and Thompson [2] raise two
valid points regarding the current shape of regional
anaesthesia teaching and training. Turbitt et al. highlight
that with the advent of ultrasound-guided techniques, the
plethora of blocks available and rise of the ‘expert
regionalist’, it is quite easy for many other anaesthetists to
steer away from practising regional anaesthesia at an early
stage of their career. In an effort to solve this problem the
article goes on to list basic level (plan A) peripheral nerve
blocks which they feel all anaesthetists should be able to
perform. Ashken and Thomson, while supporting the
concept, stress (via a survey) the difficulty in teaching even
basic blocks. In effect, they are flagging the fact that there is
a postcode lottery for the provision, teaching and training of
regional anaesthetic techniques that exists throughout the
UK.
Although not all anaesthetists should be expected to
perform all nerve blocks available (such as the plan B blocks
described by Turbitt et al.), Regional Anaesthesia UK (RA-
UK) feels all anaesthetists should be able to perform basic
blocks as a requirement for achieving a CCT in anaesthesia.
Ashken and Thompson raise the important point that the
exact list of basic blocks will cause debate. A list of basic
regional anaesthesia techniques should be pragmatic,
evidenced-based and safe. Regional Anaesthesia-UK feels
that the list as suggested by Turbitt et al. is appropriate, or
even that put forward by Shonfeld and Harrop-Griffiths
when they described ‘desert island blocks’ [3].
Regional Anaesthesia-UK recommends the following
basic level peripheral regional techniques: interscalene
brachial plexus; axillary brachial plexus; rectus sheath;
femoral nerve; and sciatic (popliteal level) nerve. These
blocks have an excellent track record, are easy to teach and
the anaesthetist that can perform all five can effectively
anaesthetise the majority of the body. How do we achieve
this basic level with the existing limitations in training? In the
first instance, the importance of regional anaesthesia needs
to be appreciated by governing bodies. Regional
Anaesthesia-UK has been working with the Royal College of
Anaesthetists on the curriculum redesign. Schools of
Anaesthesia need to create training rotations so all trainees
© 2020 Association of Anaesthetists
Anaesthesia 2020, 75, 550–559
are exposed to the aforementioned basic blocks (as well as
opportunities for advanced techniques) in an effort to
combat the existing postcode lottery.
In addition to on-the-job training, all anaesthetists
should attend at least one regional anaesthesia course. The
selection of course will depend on the individual
anaesthetist’s needs, but generally speaking the content
can include a combination of model scanning, cadaveric
anatomy and cadaveric needling. Regional Anaesthesia-UK
now highlights the variation offered by courses through the
RA-UK course approval process.
Lastly, for training and career grade anaesthetists,
RA-UK now offers a link network where anaesthetists
interested in seeing blocks in action can contact experts
around the country to arrange a visit. This is especially
important when a new regional anaesthetic technique is
introduced.
B. Fox
Queen Elizabeth Hospital King’s Lynn NHS Foundation
Trust, Kings Lynn, UK
Email: benjaminfox@doctors.org.uk
A. Pawa
On behalf of Regional Anaesthesia UK (RA-UK)
Guy’s and St. Thomas’ Hospitals,
London, UK
No competing interests declared.
References
1. Turbitt LR, Mariano ER, El-Boghdadly K. Future directions in
regional anaesthesia: not just for the cognoscenti. Anaesthesia
2020; 75: 293–7.
2. Ashken TW, Thompson MHW. Future directions in regional
anaesthesia: not just for the cognoscenti: a reply. Anaesthesia
2020; 75: 554.
3. Shonfeld A, Harrop-Griffiths W. Chapter 42: Regional
Anaesthesia. In: Allman KG, Wilson IH, ed. Oxford Handbook of
Anaesthesia, Oxford: Oxford University Press, 2016; 1904.
doi:10.1111/anae.14975
555
Anaesthesia 2020, 75, 550–559 Correspondence

Utilisation of ‘verbalisation’ to reduce the complications of

tracheal intubation
We read with interest the study in which Jelacic et al. [1]
demonstrate the benefit of a pre-induction aviation-style
computerised checklist in reducing non-routine induction
events and missing critical pre-induction actions. With the
increasing recognition of the advantages of utilising
checklists to act as prompts in order to avoid missing critical
steps, the Difficult Airway Society (DAS), in collaboration
with the Intensive Care Society, the Royal College of
Anaesthetists and the Faculty of Intensive Care Medicine
have recently launched a tracheal intubation checklist for
the preparation of tracheal intubation in the critically ill adult
patient. Checklist-based approaches to maximise patient
safety at induction are to be commended and should be
considered for all patients requiring tracheal intubation.
Although rare, the National Audit Project of the Royal
College of Anaesthetists on major complications of airway
management (NAP4) and accidental awareness during
general anaesthesia (NAP5) demonstrate that catastrophic
complications occur during and after tracheal intubation.
These events include: failure to perform timely front-of-neck
access in the occurrence of a ‘can’t intubate can’t oxygenate’
(CICO) event; failure to oxygenate caused by unrecognised
oesophageal intubation; accidental endobronchial
intubation; and accidental awareness during general
anaesthesia (AAGA) caused by a failure to turn on the
vaporiser following successful intubation.
In many reported cases, better planning,
communication and situational awareness could have
mitigated or prevented patient harm, lending further
credence to the benefit of utilising checklists and prompts
that ensure all necessary steps are reliably performed.
As well as written checklists and aviation-style
checklists, verbalisation of important steps (also known as
risk-triggered commentary) has been demonstrated to
reduce human error when complex tasks are performed
[2]. The act of verbalisation facilitates the cognitive
control processes that are necessary for effective retrieval
and activation of working memory to be reinforced. In
addition, if the operator is part of a team it allows the
coworkers to know that the task has been satisfactorily
completed. Consequently, it is used extensively by the
airline and rail industries to reduce human error [2,3] and
is especially recommended for single pilots [3].
Therefore, we propose that all of the aforementioned pre-
intubation checklists could, and should, be improved by
the addition of a further five standardised statements to
be verbalised during and immediately after tracheal
intubation in order to ensure safe placement of the

Table 1 The rationale and statements to be verbalised at each stage of tracheal intubation.
Stage
of intubation Statement to be verbalised Rationale
Initial Grade of Cormack–Lehane view • To prompt situational awareness of difficulties with tracheal intubation
laryngoscopy (especially unanticipated)
• To prompt consideration of alternative strategies, which in CICO
events could facilitate faster declaration and subsequent transition to
emergency front-of-neck access
After intubation Distance of black line on the • To prompt mental processing and need for remedial action if the
and before tracheal tube to the lips distance stated is inappropriate for that patient. This is vital, as
removal of the endobronchial intubation has been reported as the most common
laryngoscope untoward event associated with tracheal intubation [4].
Following ‘Positive capnography’ • To prompt awareness that in the event of no trace, wrong place must
intubation be the initial diagnosis
• Other signs also occur in oesophageal intubation and may lead to
confirmation bias in the event of an oesophageal intubation and
therefore must not be used [5]
‘Cricoid off’ • To ask the anaesthetic assistant to remove cricoid pressure, once
satisfactory tracheal tube placement is confirmed
‘Vaporiser on’ or ‘Intravenous • To mitigate the risk of AAGA associated with successful tracheal
sedating anaesthetic drug being intubation it is essential to ensure adequate anaesthetic agent is being
administered’ administered
CICO, can’t intubate can’t oxygenate; AAGA, accidental awareness during general anaesthesia.

556 © 2020 Association of Anaesthetists

Correspondence
tracheal tube and to reduce the occurrence of induction-
related AAGA. These statements and their rationale are
summarised in Table 1.
We congratulate Jelacic et al. and DAS for their efforts
to enhance safety and pre-induction preparation through
the use of checklists. Moreover, in line with other safety-
conscious industries, we suggest the anaesthetic community
considers adopting the practice of ‘verbalisation’, especially
during the period of tracheal intubation.
B. Nourallah
N. Levy
West Suffolk Hospital,
Bury St Edmunds, UK
Email: nicholas.levy@wsh.nhs.uk
No conflicts of interest declared.
Utilising the multidisciplinary concepts of peri-operative
medicine to improve the outcome of the parturient with
diabetes
We read with interest the editorial by Dennis and Sheridan
discussing the need to incorporate the practice of peri-
operative medicine in the management of pregnant women
[1]. In 2015, pre-existing diabetes and gestational diabetes
complicated approximate 5% of pregnancies in the UK [2].
These parturients represent a high-risk group as maternal
hyperglycaemia is associated with an increased incidence
of miscarriage, sepsis and pre-eclampsia and, as gestational
diabetes is associated with obesity and increased maternal
age, they are at further risk of complications [2].
Furthermore, neonates of parturients with diabetes are at
greater risk of hypoglycaemia, macrosomia, birth injuries,
congenital abnormalities and peri-natal mortality [3].
In 2015, in an effort to improve outcomes for both infant
and mother, the National Institute for Health and Care
Excellence (NICE) published clinical guidance on ‘Diabetes
in pregnancy: management from preconception to the
postnatal period’ [2]. Among other recommendations, the
guidance recommended that capillary blood glucose
1 1
should be maintained between 4 mmol.l and 7 mmol.l
during delivery, and that intravenous (i.v.) dextrose and i.v.
© 2020 Association of Anaesthetists
Anaesthesia 2020, 75, 550–559
References
1. Jelacic S, Bowdle A, Nair BG, et al. The effects of an aviation-style
computerised pre-induction anaesthesia checklist on pre-
anaesthetic set-up and non-routine events. Anaesthesia 2019;
74: 1138–46.
2. O’Connell PJ, Lawton F, Mills AM, Klockner K. Improving signal
passed at danger management in New Zealand rail operations:
combining stabilised approach procedures with risk-triggered
commentary driving. Proceedings of the Institution of
Mechanical Engineers, Part F: Journal of Rail and Rapid Transit
2017; 231: 1070–7.
3. Civil Aviation Authority. Best practice in pilot monitoring.
https://www.caa.co.uk/Safety-initiatives-and-resources/Safety-
projects/Monitoring-matters/Best-practice-in-pilot-monitoring/
(accessed 08/12/2019).
4. McCoy EP, Russell WJ, Webb RK. Accidental bronchial
intubation: an analysis of AIMS incident reports from 1988 to
1994 inclusive. Anaesthesia 1997; 52: 24–31.
5. Jafferji D, Morris R, Levy N. Reducing the risk of confirmation bias
in unrecognised oesophageal intubation. British Journal of
Anaesthesia 2019; 122: 66–8.
doi:10.1111/anae.15001
insulin infusions should be used for women with diabetes
whose capillary plasma glucose is not maintained in this
range [2]. This stringent target range is based on the
rationale that hyperglycaemia during the peripartum period
increases the risk of neonatal hypoglycaemia. The literature
cited by NICE supporting this notion was mostly published
more than 20 years ago, before the advent of improved
glycaemic control during pregnancy [3]. More recent data,
including a systematic review [4], the prospective trial by
Yamamoto et al. [5] and a retrospective multicentre study
[6], suggest that it is chronic exposure of the fetus to
hyperglycaemia during the whole of the pregnancy that is
the main cause of neonatal hypoglycaemia, rather than the
immediate glycaemic control in the few hours just before
delivery.
As the intrapartum use of i.v. insulin is associated with
hypoglycaemia [7], it is necessary to ensure that parturients
are not allowed to develop hypoglycaemia and thus we
recommend a lowest acceptable intrapartum glucose value
1
of 5 mmol.l [3]. Moreover, as the use of hypotonic
dextrose is associated with maternal and neonatal
557
Correspondence
tracheal tube and to reduce the occurrence of induction-
related AAGA. These statements and their rationale are
summarised in Table 1.
We congratulate Jelacic et al. and DAS for their efforts
to enhance safety and pre-induction preparation through
the use of checklists. Moreover, in line with other safety-
conscious industries, we suggest the anaesthetic community
considers adopting the practice of ‘verbalisation’, especially
during the period of tracheal intubation.
B. Nourallah
N. Levy
West Suffolk Hospital,
Bury St Edmunds, UK
Email: nicholas.levy@wsh.nhs.uk
No conflicts of interest declared.
Utilising the multidisciplinary concepts of peri-operative
medicine to improve the outcome of the parturient with
diabetes
We read with interest the editorial by Dennis and Sheridan
discussing the need to incorporate the practice of peri-
operative medicine in the management of pregnant women
[1]. In 2015, pre-existing diabetes and gestational diabetes
complicated approximate 5% of pregnancies in the UK [2].
These parturients represent a high-risk group as maternal
hyperglycaemia is associated with an increased incidence
of miscarriage, sepsis and pre-eclampsia and, as gestational
diabetes is associated with obesity and increased maternal
age, they are at further risk of complications [2].
Furthermore, neonates of parturients with diabetes are at
greater risk of hypoglycaemia, macrosomia, birth injuries,
congenital abnormalities and peri-natal mortality [3].
In 2015, in an effort to improve outcomes for both infant
and mother, the National Institute for Health and Care
Excellence (NICE) published clinical guidance on ‘Diabetes
in pregnancy: management from preconception to the
postnatal period’ [2]. Among other recommendations, the
guidance recommended that capillary blood glucose
1 1
should be maintained between 4 mmol.l and 7 mmol.l
during delivery, and that intravenous (i.v.) dextrose and i.v.
© 2020 Association of Anaesthetists
Anaesthesia 2020, 75, 550–559
References
1. Jelacic S, Bowdle A, Nair BG, et al. The effects of an aviation-style
computerised pre-induction anaesthesia checklist on pre-
anaesthetic set-up and non-routine events. Anaesthesia 2019;
74: 1138–46.
2. O’Connell PJ, Lawton F, Mills AM, Klockner K. Improving signal
passed at danger management in New Zealand rail operations:
combining stabilised approach procedures with risk-triggered
commentary driving. Proceedings of the Institution of
Mechanical Engineers, Part F: Journal of Rail and Rapid Transit
2017; 231: 1070–7.
3. Civil Aviation Authority. Best practice in pilot monitoring.
https://www.caa.co.uk/Safety-initiatives-and-resources/Safety-
projects/Monitoring-matters/Best-practice-in-pilot-monitoring/
(accessed 08/12/2019).
4. McCoy EP, Russell WJ, Webb RK. Accidental bronchial
intubation: an analysis of AIMS incident reports from 1988 to
1994 inclusive. Anaesthesia 1997; 52: 24–31.
5. Jafferji D, Morris R, Levy N. Reducing the risk of confirmation bias
in unrecognised oesophageal intubation. British Journal of
Anaesthesia 2019; 122: 66–8.
doi:10.1111/anae.15001
insulin infusions should be used for women with diabetes
whose capillary plasma glucose is not maintained in this
range [2]. This stringent target range is based on the
rationale that hyperglycaemia during the peripartum period
increases the risk of neonatal hypoglycaemia. The literature
cited by NICE supporting this notion was mostly published
more than 20 years ago, before the advent of improved
glycaemic control during pregnancy [3]. More recent data,
including a systematic review [4], the prospective trial by
Yamamoto et al. [5] and a retrospective multicentre study
[6], suggest that it is chronic exposure of the fetus to
hyperglycaemia during the whole of the pregnancy that is
the main cause of neonatal hypoglycaemia, rather than the
immediate glycaemic control in the few hours just before
delivery.
As the intrapartum use of i.v. insulin is associated with
hypoglycaemia [7], it is necessary to ensure that parturients
are not allowed to develop hypoglycaemia and thus we
recommend a lowest acceptable intrapartum glucose value
1
of 5 mmol.l [3]. Moreover, as the use of hypotonic
dextrose is associated with maternal and neonatal
557
Anaesthesia 2020, 75, 550–559
hyponatraemia [3], we recommend the NICE guidance on
i.v. fluid therapy in children and young people is adhered to
[8], and only solutions with a sodium concentration of
1 1
between 131 mmol.l and 154 mmol.l are administered
to this vulnerable population.
Adopting the PARRCEL approach advocated by Dennis
and Sheridan lends itself neatly to promote improvements
in care to the parturient/potential parturient with concurrent
diabetes [1].
During Pre-pregnancy counselling, patients can be
educated on adopting a healthier lifestyle which can reduce
their risk of developing gestational diabetes. Antenatal care
involves optimising management of pre-existing diabetes
or newly diagnosed gestational diabetes. Improved
glycaemic control reduces the risk of morbidity and
mortality to mother and fetus.
Risk stratification identifies those at increase risk of
complications. Parturients with poorly controlled diabetes
should be managed in a multidisciplinary clinic.
Resuscitation involves early recognition and treatment of
the critically unwell. For the parturient with diabetes, over
and above the usual complications, this includes
hypoglycaemia, hyponatraemia and diabetic ketoacidosis.
Involvement of multidisciplinary team members such as
the peri-operative physician allows for Collaborative
decision making. Care plans should be tailored to their
individual needs, and always involve the parturient in this
process. This is especially pertinent in the era post
‘Montgomery’.
The Enhanced recovery and rehabilitation pathway
component should place emphasis on avoiding maternal
hypoglycaemia, as insulin requirements will decrease
rapidly post delivery. Upon discharge, linkage to
community support ensures there is continuity of care with
ongoing input from primary care and community nurses in
order to prevent long-term secondary complications.
Thus, in addition to updating existing NICE guidance,
the skills of anaesthetists in driving the concepts of peri-
operative medicine (patient-centred, pathway focused,
558
Correspondence
shared decision making, multidisciplinary and integrated
care) can be utilised to improve the care of pregnant women
with diabetes.
Y. Yap
A. Modi
N. Levy
West Suffolk Hospital,
Bury St Edmunds, UK
Email: nicholas.levy@wsh.nhs.uk
No competing interests declared.
References
1. Dennis AT, Sheridan N. Sex, suffering and silence – why peri-
operative medicine must prioritise pregnant women.
Anaesthesia 2019; 74: 1504–8.
2. National Institute for Health and Care Excellence. Diabetes in
pregnancy: management from preconception to the postnatal
period. NG3. 2015. https://www.nice.org.uk/guidance/ng3
(accessed 27/11/2019).
3. Modi A, Levy N, Hall GM. Controversies in the peripartum
management of diabetes. Anaesthesia 2016; 71: 750–5.
4. Yamamoto JM, Benham J, Mohammed K, et al. Intrapartum
glycaemic control and neonatal hypoglycaemia in pregnancies
complicated by diabetes: a systematic review. Diabetic Medicine
2018; 35: 173–83.
5. Yamamoto JM, Corcoy R, Donovan LE, et al. Maternal glycaemic
control and risk of neonatal hypoglycaemia in Type 1 diabetes
pregnancy: a secondary analysis of the CONCEPTT trial. Diabetic
Medicine 2019; 36: 1046–53.
6. Yamamoto JM, Donovan LE, Mohammad K, Wood SL. Severe
neonatal hypoglycaemia and intrapartum glycaemic control in
pregnancies complicated by type 1, type 2 and gestational
diabetes. Diabetic Medicine 2020; 37: 138–46.
7. Dashora U, Rafique S, Tharayil G, Jones S, Castro E, Sathiskumar
P. The feasibility and impact of implementing NICE guidance on
diabetes control during delivery. British Journal of Diabetes
2017; 17: 100–6.
8. National Institute for Health and Care Excellence. Intravenous
fluid therapy in children and young people in hospital. NG29.
2015. https://www.nice.org.uk/guidance/ng29 (accessed 10/
11/2019).
doi:10.1111/anae.14962
© 2020 Association of Anaesthetists
Correspondence Anaesthesia 2020, 75, 550–559

Ongoing role for starches in anaesthesia

We read with interest the recent editorial by Charlesworth
and Shelton [1], and feel they provided a measured
argument in the debate regarding the role of starches and
gelatins in peri-operative management, as well as the
caution that should be exercised with their use. The use of
starches as an intravenous fluid therapy was achieved by
taking advantage of the similarity between natural human
tissue glycogen and naturally occurring amylopectin. This
was originally derived from waxy maize, although later
generations have utilised potatoes, despite potential
variations in pharmacological and clinical equivalence.
Having performed a brief literature review of the
humble potato in the anaesthetic literature, we feel that their
role should not be overlooked.
Potatoes have a vital role as an educational aid for
neuraxial anaesthesia [2] and have been part of a slightly
gruesome opportunistic study investigating the gastric transit
times of food following a shotgun-induced gastric fistula [3]. In
a rigorously conducted crossover study providing insight into
the gastric absorption of diazepam, potatoes were combined
with a diet of frankfurters and a glass of milk compared with a
strict control group that had no intake (with the obvious
exceptions of coffee, tea, cola and tobacco) [4]. It has also
featured in its ‘Mr. Potato Head’ version as a potentially
threatening stimulus for Macaques undergoing anaesthesia [5].
Despite the varied success the potato has reached in
increasingly questioned in the future of intravenous fluid
therapy, we are glad it will not be lost from our practice
forever.
L. Kidd
T. Cope
North Bristol NHS Trust,
Bristol, UK
Email: lawrence.kidd@nbt.nhs.uk
No conflicts of interest declared.
References
1. Charlesworth M, Shelton CL. Should intravenous gelatins have a
role in contemporary peri-operative and critical care?
Anaesthesia 2020; 75: 266–9.
2. Bexton MD. A model for instruction in epidural technique.
Anaesthesia 1981; 36: 426–7.
3. Beaumont W, Martin AS. Experiments and observations on the
gastric juice, and the physiology of digestion. Lilly, Wait: F. P.
Allen, 1834.
4. Korttila K, Matttla MJ, Linnoila M. Prolonged recovery after
diazepam sedation: The influence of food, charcoal ingestion
and injection rate on the effects of intravenous diazepam. British
Journal of Anaesthesia 1976; 48: 333–40.
5. Coleman K, Robertson ND, Dissen GA, et al. Isoflurane
anesthesia has long-term consequences on motor and
behavioral development in infant rhesus macaques.
Anesthesiology 2017; 126: 74–84.

numerous facets of anaesthesia, we were delighted to see

that this month’s Anaesthesia arrived in a compostable doi:10.1111/anae.15016
potato starch wrapping. Although the potato may be

© 2020 Association of Anaesthetists 559