British Journal of Anaesthesia, 121 (4): 768e775 (2018)

doi: 10.1016/j.bja.2018.06.018
Advance Access Publication Date: 25 July 2018
Special Article

NEUROSCIENCE AND NEUROANAESTHESIA

Targeted temperature management in patients with

intracerebral haemorrhage, subarachnoid
haemorrhage, or acute ischaemic stroke: consensus
recommendations
P. J. D. Andrews1,*, V. Verma2, M. Healy2, A. Lavinio3, C. Curtis4, U. Reddy5,
J. Andrzejowski6, A. Foulkes7 and S. Canestrini8
1
Centre for Clinical Brain Sciences, University of Edinburgh, Edinburgh, UK, 2Royal London Hospital,
London, UK, 3Neurosciences and Trauma Critical Care Unit, Addenbrooke’s Hospital, Cambridge,
UK, 4University College London Hospital, London, UK, 5National Hospital for Neurology and Neurosurgery,
University College London Hospitals NHS Foundation Trust, London, UK, 6Sheffield Teaching Hospitals NHS
Foundation Trust, Sheffield, UK, 7The Walton Centre for Neurology and Neurosurgery, Liverpool, UK and
8
King’s College Hospital NHS Foundation Trust, London, UK

*Corresponding author. E-mail: P.Andrews@ed.ac.uk

Abstract
Background: A modified Delphi approach was used to identify a consensus on practical recommendations for the use of
non-pharmacological targeted temperature management in patients with intracerebral haemorrhage, subarachnoid
haemorrhage, or acute ischaemic stroke with non-infectious fever (assumed neurogenic fever).
Methods: Nine experts in the management of neurogenic fever participated in the process, involving the completion of
online questionnaires, face-to-face discussions, and summary reviews, to consolidate a consensus on targeted tem-
perature management.
Results: The panel’s recommendations are based on a balance of existing evidence and practical considerations. With
this in mind, they highlight the importance of managing neurogenic fever using a single protocol for targeted temper-
ature management. Targeted temperature management should be initiated if the patient temperature increases above
37.5
C, once an appropriate workup for infection has been undertaken. This helps prevent prophylactic targeted tem-
perature management use and ensures infection is addressed appropriately. When neurogenic fever is detected, targeted
temperature management should be initiated rapidly if antipyretic agents fail to control the temperature within 1 h, and
should then be maintained for as long as there is potential for secondary brain damage. The recommended target
temperature for targeted temperature management is 36.5e37.5
C. The use of advanced targeted temperature man-
agement methods that enable continuous, or near continuous, temperature measurement and precise temperature
control is recommended.
Conclusions: Given the limited heterogeneous evidence currently available on targeted temperature management use in
patients with neurogenic fever and intracerebral haemorrhage, subarachnoid haemorrhage, or acute ischaemic stroke, a

Editorial decision: 2 July 2018; Accepted: 2 July 2018

© 2018 British Journal of Anaesthesia. Published by Elsevier Ltd. All rights reserved.
For Permissions, please email: permissions@elsevier.com

768
 Targeted temperature management in neurogenic fever - 769

Delphi approach was appropriate to gather an expert consensus. To aid in the development of future investigations, the
panel provides recommendations for data gathering.

Keywords: stroke; subarachnoid hemorrhage; intracerebral hemorrhage; Delphi technique

Targeted temperature management (TTM) is the process of Methods

controlling the core body temperature at a specific level. It can
A modified Delphi consensus approach was used, which
be used to achieve hypothermia (TTMhypo) or maintain normal
involved a combination of online questionnaires, a face-to-
body temperature (TTMnorm). TTM has been used in several
face meeting, and post-meeting reviews. The process con-
clinical situations, such as out-of-hospital cardiac arrest,
sisted of two rounds of a Delphi questionnaire (questions are
traumatic brain injury (TBI), and cerebral vascular accidents,
in supplementary data) plus a final validation stage, as shown
in an attempt to reduce neurological damage and enhance
in Table 1. Rounds 1 and 2 were conducted at a face-to-face
functional outcomes.1 The evidence base for TTM use in pa-
meeting held at the De Vere Grand Connaught Rooms in
tients with intracerebral haemorrhage (ICH), subarachnoid
London on July 5, 2017. P.J.D.A. acted as Chair, with an inde-
haemorrhage (SAH), or acute ischaemic stroke (AIS) is limited
pendent Delphi facilitator moderating the meeting. After the
and difficult to interpret given the range of TTM methods used,
initial meeting, the outcomes report (Round 3) and manuscript
the different target temperatures used, the heterogeneity of
validations were conducted asynchronously, with documents
the patient groups, and the presence or absence of neurogenic
shared by e-mail and feedback collected from each participant
or infectious fever. A similar amount of heterogeneity exists in
independently by the facilitator. The agreed cut-off of for the
the limited number of clinical guidelines published in France
consensus was 70% of experts in agreement; this was in
and the USA.
keeping with recent consensus initiatives in this field.12
Fever is common in critically ill patients with neurological
conditions.2,3 In those with AIS or TBI, fever can contribute to
secondary brain injury, and is associated with poorer func- Participants
tional outcomes and higher morbidity and mortality.4e6 Fever
A total of nine experts in the management of neurogenic fever
has an infectious cause in about half of all cases.2,3,7,8 It has
participated in the consensus process. The participants were
also been shown to have a strong independent association
selected on the basis of their clinical role, and their experience
with poor outcomes.7,9,10 Most evidence on fever prevention in
of managing patients with ICH, SAH, and AIS; managing fever
patients in critical care is observational in nature, so the spe-
in these patients; and using TTM. The nine participants were
cific role fever plays in causing secondary brain injury is un-
drawn from leading intensive and neurocritical care groups in
clear. The Impact of Fever Prevention in Brain Injured Patients
the UK. Five participants attended Rounds 1 and 2. Of these
study (NCT02996266) is a US study that may provide some
five participants, one felt that they had insufficient breadth of
answers to this question, as it is designed to assess the impact
relevant expertise to respond to the questions, and therefore,
of fever prevention on fever burden and short- and long-term
withdrew from voting to avoid bias. This participant did,
neurological outcomes in brain-injured patients.11
however, engage in the discussions and provided insight into
The goal of this modified Delphi consensusd‘establishing
infection-related issues. Nine participants were involved in
consensus in health outcomes (ECHO) for TTM in patients with
the final manuscript validation.
neurogenic fever’dwas to identify common expert practice
recommendations for the use of non-pharmacological TTM in
patients with ICH, SAH, or AIS who develop fever. The panel of Rounds 1 and 2 questions
experts were drawn from UK centres and reflect UK-specific
Statements and questions for each round were prepared by
practice, although their recommendations may be extrapo-
the facilitator in consultation with P.J.D.A. and delivered by
lated to other high-income countries with similar healthcare
SurveyMonkey® to each attendee’s e-mail address for them to
systems.
complete anonymously online and without collaboration

Table 1 Delphi process

Step Format Description n

Round 1 Face-to-face meeting; SurveyMonkey Questionnaire completed anonymously; 4

questionnaire 25 statements/questions
Round 2 Face-to-face meeting; SurveyMonkey Revised questionnaire completed anonymously; 4
questionnaire 14 revised/new consensus statements
Round 3 Independent, asynchronous review Consensus summary document for review 9
via e-mail and comment
Manuscript Independent, asynchronous review Final manuscript review and validation 9
validation via e-mail
 770 - Andrews et al.

whilst at the meeting. Statements and questions were
informed by a literature search; the search yielded very few
publications relevant to the specific topics under discussion,
and these were not shared with the expert panel.
Round 1 comprised 25 statements and questions related to
the clinical use of TTM for neurogenic fever in patients with
ICH, SAH, or AIS. These had been created in consultation with
the meeting Chair. The majority of Round 1 questions were in
a multiple-choice format, with two free-text questions
designed to elicit information and one ranking question. All
questions were mandatory, and included a comment box
where participants could provide additional comments or
insights.
Pooled responses to the Round 1 questions were displayed
on screen to the whole group, and the results and comments
were discussed. All responses were reviewed and discussed
regardless of the level of consensus. Where consensus (70%
agreement) was achieved, the discussion focused on im-
provements in the phrasing or scope of the initial statement to
arrive at a final statement that clearly captured the consensus
views of all experts. Where consensus was not reached, a
detailed facilitated discussion was undertaken to identify the
reasons for the lack of agreement. From these discussions, 14
revised or new statements or questions were identified, which
the participants addressed and voted on in Round 2.
Round 3 and final validation
The responses from the meeting were captured in a summary
document that showed how the consensus evolved at the
meeting. For Round 3, this summary was distributed to all
participants via e-mail, with meeting attendees asked to
confirm the accuracy of the discussion and non-attendees
asked to add their opinions to the document. The additional
comments from non-attendees were collated and reviewed.
Areas requiring additional discussion were identified, and the
process for addressing these was guided by P.J.D.A. A manu-
script was created, structuring the recommendations, adding
additional narrative, and providing context. This manuscript
was distributed to all participants for review and final
validation.
Results
The results of the final consensus agreements are presented in
Table 2. The debate and considerations behind these agree-
ments are captured in the discussion section to provide a
broader context, in which they can be properly reviewed.
A detailed debate on the definition of neurogenic fever was
not undertaken; however, the group established that, for the
purposes of this Delphi consensus, neurogenic fever equated
with non-infectious fever.
Discussion
This consensus discussion approach was necessary because of
the scant and heterogeneous nature of existing published ev-
idence. The modified Delphi approach sought to combine the
advantages of a traditional Delphi process (specifically, the
structured information flow and anonymous submission of

Table 2 Summary of recommendations. AIS, acute ischaemic stroke; ICH, intracerebral haemorrhage; MOHS, modified Oxford
handicap scale; SAH, subarachnoid haemorrhage; TTM, targeted temperature management

Topic Level of consensus

Core-temperature measurement 100% Round 2

Core-temperature measurement is important to enable the effective identification, treatment, and
monitoring of neurogenic fever.
Core temperature should be measured continuously, or at a minimum hourly, in patients with ICH, 100% Round 1
SAH, or AIS.
Neurogenic fever and TTM 100% Round 1
As neurogenic fever is associated with poor patient outcomes amongst patients with ICH, SAH, or AIS,
treating it is important.
TTMnorm is appropriate for the treatment of neurogenic fever in adult patients with ICH, SAH, and AIS. 75% ICH; 100% SAH,
AIS Round 1
From a practical perspective, it is appropriate to have a single protocol for TTM of neurogenic fever in 100% Round 2
ICH, SAH, and AIS.
When to use TTM 100% Round 1
TTM should be used reactively in patients with ICH, SAH, or AIS in response to neurogenic fever.
TTM should be initiated if the patient’s temperature increases to 37.5
C and infection is excluded. >75% Round 3
TTM should be initiated as rapidly as possibly once fever is detected and if pharmacological treatment 100% Round 2
has not controlled the temperature within 1 h of administration.
How to use TTM 100% Round 2
The target temperature for patients with ICH, SAH, or AIS who develop neurogenic fever is
37.0
C±0.5
C.
TTM should be maintained for as long as there is potential for secondary brain damage. 100% Round 2
The use of an advanced TTM method, enabling precise temperature control, is required to maintain 100% Round 1
the temperature effectively.
Shivering 100% Round 1
Shivering should be managed during TTM.
Outcomes assessment 100% Round 2
Data on outcomes should be collected for patients with ICH, SAH, or AIS
(i) MOHS assessment is recommended at 1 month.
(ii) MOHS and modified Rankin scale assessment should be attempted at 6 months.
 Targeted temperature management in neurogenic fever
opinion) with those of the nominal group technique (specif-
ically, the ability to actively discuss the responses to the
questions, leading to further voting). The focus was on gaining
a consensus and expert insight into the usual practice in the
UK regarding TTM, which, in this case, relates to TTMnorm,
with the recommendations focusing on non-pharmacological
TTM approaches.
Core-temperature measurement
(i) Core-temperature measurement is important to enable
effective identification, treatment, and monitoring of
neurogenic fever.
(ii) Core temperature should be measured continuously, or at
a minimum hourly, in patients with ICH, SAH, or AIS.
There was no clear agreement amongst the participants on
the best site at which to measure core temperature. The lack of
a final consensus related to the practical challenges that can
exist locally, if the patient was awake or comatose, or how the
core-temperature measurements were interpreted at specific
sites, especially if these sites are not standard practice.
Guidelines from a French expert panel, whilst recom-
mending a core-temperature measurement for TTM in ICUs,
failed to indicate specific sites.12 Oesophageal or bladder
probes are recommended for temperature measurement
during TTM after a neurological injury by recent guidelines
from the US Neurocritical Care Society (NCS), which also
suggest a preference for continuous monitoring.5
The Delphi participants’ recommendations for sites of
temperature measurement were split between two non-gold
standard sites: the oesophagus and the rectum. The panel
felt that whatever site is selected, it should enable continuous,
or near continuous, monitoring. However, specific clinical
situations may make continuous monitoring impractical and,
in such cases, a minimum of hourly measurements is advised.
Neurogenic fever and targeted temperature
management
(i) As neurogenic fever is associated with poor patient out-
comes amongst patients with ICH, SAH, or AIS, treating it is
important.
(ii) TTMnorm is appropriate for the management of neurogenic
fever in adult patients with ICH, SAH, or AIS.
Fever is commonly reported in ICH, SAH, and AIS, although
incidence rates vary between studies and the cause of fever is
often not clearly defined. In a large retrospective US study of
fever in patients with brain injuries, 51% were classified as
having fever, with fever incidences of 60% for TBI, 54% for
aneurysmal SAH, 50% for ICH, and 37% for AIS.3
All panel members accepted that poor outcomes are associ-
ated with neurogenic fever in these patients, given the weight of
published evidence, for example, by Scaravilli and colleagues6
and Kramer and colleagues9 on SAH, Leira and colleagues10 on
ICH, and Rincon and colleagues3 and Greer and colleagues4 on all
strokes and TBI. Fever can increase the brain’s metabolic de-
mand, exacerbating ischaemic injury and leading to cerebral
vessel vasodilation, ultimately worsening the intracranial
pressure (ICP)13; therefore, it should be avoided or treated.
A major question regarding the use of TTM in febrile patients
with brain injury is the type of TTM that should be used: TTMhypo
(maintenance of temperature below 37
C) or TTMnorm (mainte-
nance of temperature at 37
C). From the literature, it is difficult
- 771
to discern the evidence for reactive TTMnorm to manage neuro-
genic fever from the data published on the use of prophylactic
TTMnorm or TTMhypo for neuroprotection in TBI and
stroke.5,12,14,15 Overall, the panel felt that the evidence for
TTMhypo in patients with neurogenic fever in ICH, SAH, and AIS
was poor, and that, on balance, hypothermia could be associated
with negative outcomes. The panel, therefore, recommended
TTMnorm for these patients, and hence, the focus of this paper.
There was no detailed discussion of the criteria for TTMhypo.
(iii) From a practical perspective, it is appropriate to have a
single protocol for TTM of neurogenic fever in ICH, SAH,
and AIS.
Although the panel acknowledged the existence of some
differences between the three conditions, they felt that there
was enough similarity to allow recommendations to be
grouped in a single protocol, especially as it was felt imprac-
tical to have more than one. The panel did not consider loca-
tion of care (ICU or ward) for these patients important,
appreciating that such decisions would be dictated by local
practice, service availability, and patient needs.
When to use targeted temperature management
(i) TTM should be used reactively in patients with ICH, SAH, or
AIS in response to neurogenic fever.
The panel felt that TTM should only be used once a patient
develops fever, as there is limited evidence to support the use
of prophylactic TTM after ICH, SAH, or AIS regardless of core
temperature. Prophylactic TTM also carries additional risks,
such as masking fever associated with infection, and should
therefore be avoided.
(ii) TTM should be initiated if the patient temperature in-
creases to 37.5
C and infection is excluded.
The temperature threshold for TTM was discussed in detail
at the meeting and in subsequent review rounds. The
threshold debate centred on balancing the need to act as soon
as temperature increase is detected, against mitigating the
risk of prophylactic TTM use.
In healthy individuals, core temperature is not constant but
fluctuates by up to 0.5
C around the ‘normal’ 37
C average
temperature.16 Although the term fever is commonly under-
stood, no universally accepted definition exists, and it is often
used interchangeably with the terms pyrexia or hyperthermia.2,17
Similarly, there is no accepted temperature threshold that de-
fines the occurrence of fever, with between 5 and 14 different
thresholds reported in various studies identified by a literature
review.2 A threshold of 38.3
C is often cited. This equates to
101
F and is based on US recommendations on fever control.17
The majority of the panel felt that >37.5
C was appropriate
and in line with the range agreed for normothermia (see later),
although a minority preferred 38
C. The panel also recog-
nised the urgency of expediting the identification and treat-
ment of potentially life-threatening infection as an important
step before the use of TTM. Fever should be treated early,
regardless of its cause.
(iii) TTM should be initiated as rapidly as possible once fever is
detected if pharmacological treatment has not controlled
temperature within 1 h of administration.
The panel thought that the evidence for the effectiveness of
antipyretic agents, such as paracetamol, in controlling fever in
 772 - Andrews et al.
patients with ICH, SAH, or AIS was limited. However, they
accepted that, as these agents are widely used as the first
option for fever control, it was appropriate to try them, but act
swiftly if they yield no benefit within 1 h. The panel advised a
more aggressive treatment of temperature if the fever was
associated with seizures.
Paracetamol is widely used in ICU patients to treat both
fever and pain. A 4 g day1 dose is often administered as
standard, but reports suggest this is frequently ineffective for
fever control in patients with brain injury.1 Other studies using
a higher dose (6 g day1) have shown small but important re-
ductions in temperature.1,18,19 Many of these studies admin-
istered paracetamol to all patients, not only those with fever,
so it is difficult to draw firm conclusions on its use in fever.
There is little evidence on functional outcomes in the use of
paracetamol in patients with acute stroke.20e22 However, one
study suggested that stroke patients with fever on admission
experience functional outcome benefits from paracetamol
use.23
The alternative antipyretic agent, ibuprofen, shows no
greater efficacy than paracetamol, and can be associated with
an increased bleeding risk.18,24 When NSAIDs are used, one
small randomised study on neurosurgical ICU patients sug-
gested that temperature control is significantly better with a
low-dose continuous infusion of diclofenac than with inter-
mittent bolus doses of NSAIDs.25
The panel debated the relative benefit of high-dose para-
cetamol compared with the risk of possible side-effects. The
panel also discussed the possibility of augmented fever control
by combing paracetamol with external cooling via fans. No
specific recommendations emerged from these discussions.
How to use TTM
(i) The target temperature for patients with ICH, SAH, or AIS
who develop neurogenic fever is 37.0
C±0.5
C.
The specific definition of normothermia recommended by
the panel is 37.0
C±0.5
C, with a target temperature plus range
believed to offer clear parameters within which to control
temperature. Strict control of core temperature to avoid fluc-
tuations may be beneficial. Evidence from a post hoc analysis of
the Eurotherm3235 Trial on TTM to 32e35
C in patients with
TBI showed that temperature fluctuations during the first 48 h
cooling period was associated with worse neurological out-
comes only in the control group. Patients in the control group
were managed to normothermia and had more episodes of
temperature in excess of 38
C than those in the hypothermia
group.26
(ii) TTM should be maintained for as long as there is potential
for secondary brain damage.
The panel debated various durations of TTM in an attempt
to provide specific guidance, but no consensus was achieved.
The suggested durations included 48 h, 72 h, 3e5 days, or
based on need. Overall, the group felt that it was difficult to
provide categorical recommendations, especially as clinical
situations, such as whether the patient is awake or comatose,
could impact the decision. The panel also recognised that
clinicians must be vigilant for the possibility of infection
occurring during the period of TTM.
The NCS TTM guidelines also addressed the question of
TTM duration by comparing the benefits of a pre-specified
duration of TTM with a goal-based approach for patients
with TBI. Their conditional recommendation, based on low-
quality evidence, was to suggest longer duration of TTM for
severe TBI patients should ICP control be the goal.5
(iii) The use of an advanced TTM method enabling precise
temperature control is required to maintain temperature
effectively.
The panel felt that accurate and consistent temperature
control required advanced TTM methods, although there was
no detailed debate on what specific advanced methods were
preferred.
The two main non-pharmacological advanced TTM
methods currently in use are surface cooling and endovascular
cooling. Surface-cooling methods include air-circulating
blankets, water-circulating blankets, and hydrogel-coated
water-circulating pads, whilst endovascular cooling uses i.v.
heat exchange catheters. Endovascular cooling can be asso-
ciated with additional risks similar to those found with inva-
sive central vascular access.1 The recent NCS guidelines made
a strong recommendation for the use of intravascular cathe-
ters or gel pads if such catheters are not available to maintain
constant temperature.5
The panel also considered inadvertent overcooling, when the
core temperature decreases below the normothermic range of
37.0
C±0.5
C. This can occur when less advanced TTM methods
are used. In such a situation, rewarming to normothermia
should be slow and controlled at a rate of 0.25
C per hour.
Shivering
(i) Shivering should be managed during TTM.
All panellists recommended that shivering should be
managed, as it can cause cerebral and metabolic stress, and
negate the beneficial effects of TTM. However, there was no
clear agreement on the method(s) that should be used to
manage shivering. The approaches to shivering control sug-
gested by the panel included pharmacological and non-
pharmacological methods, either alone or in combination:
sedation, counter-warming (not all members of the panel were
aware of counter-warming, so felt unable to comment on its
value), neuromuscular block, and a combination of the above.
Shivering is a physiological heat-generating mechanism
that occurs to maintain core temperature at the set-point
temperature of the hypothalamus.24 As about 20% of the
thermoregulatory drive for shivering comes from the skin,
counter-warming of the skin on the head, hands, arms, or face
(use of gloves, hats, socks, and blankets) can reduce shivering
without pharmacological intervention, whilst allowing the
core temperature to be controlled.27
Paracetamol, magnesium, sedatives (such as dexmedeto-
midine or buspirone), opioids (such as pethidine, tramadol,
and naloxone), neuromuscular blocking agents, and alpha-2
adrenergic agonists (such as clonidine and dexmedetomi-
dine) have all been used to treat shivering pharmacologi-
cally.5,27 Decisions may be guided by the level of
consciousness of the patient (e.g. paracetamol and clonidine
may be more appropriate in patients who are awake) or the
risks of pharmacological interventions (e.g. clonidine can
cause bradycardia and hypotension).
Shivering is well recognised as a problem with TTM. Like
the Delphi panel, the NCS also suggested the prompt treat-
ment of shivering during TTM in brain-injury patients, but
specifically suggested a stepwise approach, starting with non-
 Targeted temperature management in neurogenic fever - 773

Table 3 Summary of recommendations and guidelines on the use of TTM. Shading shows recommendations in line with this paper.
AHA, American Heart Association; AIS, acute ischaemic stroke; ASA, American Stroke Association; aSAH, aneurysmal subarachnoid
haemorrhage; DCI, delayed cerebral ischaemia; ESO, European Stroke Organisation; GCP, good clinical practice; ICH, intracerebral
haemorrhage; ICP, intracranial pressure; NCS, Neurocritical Care Society; RCT, randomised controlled trial; SAH, subarachnoid hae-
morrhage; SFAR, Socie  te
 Française d’Anesthe
sie Re
animation (French Society of Anaesthesia and Intensive Care Medicine); SRLF,
 te
Socie  de Re
animation de Langue Française (French Intensive Care Society); TTM, targeted temperature management

Paper and Condition or situation

guideline
Intracerebral Subarachnoid haemorrhage Acute ischaemic stroke Patients in
haemorrhage neurocritical care

Cariou and Consider hypothermia Consider TTM (to an unspecified Consider prophylactic
colleagues12 (35e37
C) in temperature) in comatose normothermia during
(2017); comatose patients patients with aneurysmal SAH the early phase of severe
SRLF/SFAR with spontaneous to lower the ICP or to improve ischaemic stroke.
ICH. the neurological outcome.
Madden and Recommend using
collleagues5 controlled
(2017); NCS normothermia to
reduce fever
burden in patients
with fever
refractory to
conventional
therapy.
Steiner and Increased temperature should be
collleagues32 treated medically and
(2013); ESO physically (GCP).
Steiner and There is insufficient
collleagues33 evidence from RCTs
(2014); ESO to make strong
recommendations on
whether, when, and
for whom preventive
or early fever
treatment should be
given after acute ICH.
Ntaios and In patients with AIS and
colleagues22 hyperthermia, we
(2015); ESO cannot make any
recommendation for
treating hyperthermia to
improve the functional
outcome or survival.
Connolly and Aggressive control of fever to
collleagues28 normothermia using standard
(2012); AHA/ASA or advanced temperature-
modulating systems is
reasonable in the acute phase of
aSAH.
Diringer and During the period of risk for DCI,
Bleck29 (2011); control of fever is desirable; the
NCS intensity should reflect the
individual patient’s relative risk
of ischaemia; surface cooling or
intravascular devices are more
effective, and should be used
when antipyretics fail in cases
where fever control is highly
desirable.
Hemphill and Treatment of fever
collleagues30 after ICH may be
(2015); AHA/ASA reasonable.
Jauch and Sources of hyperthermia
collleagues31 (temperature >38
C)
(2013) AHA/ASA should be identified and
treated, and antipyretic
medications should be
administered to lower
the temperature in
hyperthermic patients
with stroke.
 774 - Andrews et al.
sedating interventions, such as counter-warming, before the
use of sedatives or neuromuscular block.5 The NCS also rec-
ommends the use of a scale, such as the bedside shivering
assessment scale, to help characterise shivering and support
decision-making.5
Outcomes assessment
(i) Data on outcomes should be collected for patients with ICH,
SAH, or AIS:
(a) The modified Oxford handicap scale (MOHS) assess-
ment is recommended at 1 month.
(b) The MOHS and modified Rankin scale assessment
should be attempted at 6 months.
Given the lack of consistent evidence on TTM use in pa-
tients with ICH, SAH, or AIS, the collection of outcome data at
specific time points could support better audit and guide
further recommendations on TTM use. These recommenda-
tions should, however, balance the value of the data against
the practicality of data collection.
Conclusions
This consensus project was based on the shared desire to
develop a series of practical recommendations based on
expert opinion to support clinicians less experienced in using
TTM to manage neurogenic fever in ICH, SAH, and AIS pa-
tients. The modified Delphi approach utilised was designed to
overcome some of the gaps in published evidence on TTM in
these specific situations and to allow the recommendations to
be practically focused.
Other expert panels have considered similar, although not
identical, questions using population, intervention, compari-
son, outcomes and grades of recommendation, assessment,
development, and evaluation methodologies, and have come
up with differing opinions. A range of guidelines for the
management of patients with ICH, SAH, or AIS, or the use of
TTM in neurological patients, exists, but these provide no, or
conflicting, recommendations for TTM because of limitations
of currently available data. Table 3 summarises the key
statements on control of fever or use of TTM from these
guidelines.5,12,22,28e33 The 2017 NCS guidelines on the use of
TTM in patients in neurocritical care, for example, recom-
mend using controlled normothermia to reduce fever in those
whose fever is refractory to conventional therapy,5 whilst the
French 2017 guidelines on TTM recommend prophylactic
normothermia during the early phase of severe ischaemic
stroke, hypothermia (35e37
C) in comatose patients with
spontaneous ICH to lower ICP, and TTM (to an unspecified
temperature) in comatose patients with aneurysmal SAH to
lower ICP or improve the neurological outcome.12 This un-
derscores the challenge of interpreting and extrapolating data
to specific clinical situations. Looking more widely at general
guidelines on the management of ICH, SAH, and AIS, a simi-
larly mixed picture emerges (Table 3). On balance, reactive
TTMnorm appears a rational and consistent recommendation
for the management of neurogenic fever in patients with ICH,
SAH, or AIS given the range of views.
The consensus statements presented here focus on rec-
ommendations for a specific group of patients. They do not
contain detailed recommendations on the level of care or the
situation for comatose vs non-comatose patients. Many of
these statements may be influenced by local service structure
and practice, and the panel was keen to ensure that the rec-
ommendations could be used in any care setting, yet not be so
complex or prescriptive as to deter adoption.
The Delphi process has some drawbacks. Although the
online questionnaire was completed anonymously and inde-
pendently, the subsequent group discussion could allow social
bias in responses. Several panellists could not attend the
meeting, so their input was gathered on the meeting summary
document. This approach may have led to an unequal
weighting of opinions, with greater weight given to those who
attended in person.
Given the heterogeneity of published evidence and clinical
guidelines on the use of targeted temperature management to
treat fever in patients with intracerebral haemorrhage, sub-
arachnoid haemorrhage, or acute ischaemic stroke, all at-
tendees felt that the agreed recommendations would provide
clinical guidance for the development of local protocols for the
benefit of clinicians and patients. The next challenge will be to
assess their use and the impact on patient outcomes.
Authors’ contributions
Meeting participation: P.J.D.A., V.V., M.H., A.L., C.C.
Delphi participation: P.J.D.A., V.V., M.H., A.L.
Review of meeting report: all authors.
Review of manuscript: all authors.
Acknowledgements
The authors would like to acknowledge the support of K.
McKillen in facilitating the Delphi meeting and Hayward
Medical Communications in preparing the manuscript.
Declaration of interest
P.J.D.A. receives consultancy and speaker fees from C. R. Bard,
Inc. and Bard Medical Division, and speaker fees and Centre of
Excellence payment from Integra Neurosciences. The views
and opinions reported are those of the authors and not
necessarily those of C. R. Bard, Inc. and Bard Medical Division.
Funding
C. R. Bard, Inc. and Bard Medical Division. Bard is now a
wholly-owned subsidary of BD (Becton, Dickinson and
Company).
Appendix A. Supplementary data
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.bja.2018.06.018.
References
1. Doyle JF, Schortgen F. Should we treat pyrexia? And how
do we do it? Crit Care 2016; 20: 303
2. Niven DJ, Laupland KB. Pyrexia: aetiology in the ICU. Crit
Care 2016; 20: 247
3. Rincon F, Hunter K, Schorr C, Dellinger RP, Zanotti-
Cavazzoni S. The epidemiology of spontaneous fever and
hypothermia on admission of brain injury patients to
intensive care units: a multicenter cohort study.
J Neurosurg 2014; 121: 950e60
 Targeted temperature management in neurogenic fever
4. Greer DM, Funk SE, Reaven NL, Ouzounelli M, Uman GC.
Neurologic injury: a comprehensive meta-analysis. Stroke
2008; 39: 3029e36
5. Madden LK, Hill M, May TL, et al. The implementation of
targeted temperature management: an evidence-based
guideline from the Neurocritical Care Society. Neurocrit
Care 2017; 27: 468e87
6. Scaravilli V, Tinchero G, Citerio G. Fever management in
SAH. Neurocrit Care 2011; 15: 287e94
7. Douds GL, Tadzong B, Agarwal AD, Krishnamurthy S,
Lehman EB, Cockroft KM. Influence of fever and hospital-
acquired infection on the incidence of delayed neurolog-
ical deficit and poor outcome after aneurysmal sub-
arachnoid hemorrhage. Neurol Res Int 2012; 2012, 479865.
https://doi.org/10.1155/2012/479865
8. Meier K, Lee K. Neurogenic fever. J Intensive Care Med 2017;
32: 124e9
9. Kramer CL, Pegoli M, Mandrekar J, Lanzino G,
Rabinstein AA. Refining the association of fever with
functional outcome in aneurysmal subarachnoid hemor-
rhage. Neurocrit Care 2017; 26: 41e7
10. Leira R, Davalos A, Silva Y, et al. Early neurologic deteri-
oration in intracerebral hemorrhage: predictors and
associated factors. Neurology 2004; 63: 461e7
11. ClinicalTrials.gov. Impact of Fever Prevention in Brain
Injured Patients (INTREPID): NCT02996266. Available from:
https://clinicaltrials.gov/ct2/show/record/NCT02996266?
view¼record [accessed 30 April 2018].
12. Cariou A, Payen J-F, Asehnoune K, et al. Targeted tem-
perature management in the ICU: guidelines from a
French expert panel. Ann Intensive Care 2017; 7: 70
13. Zoerle T, Carbonara M, Zanier ER, et al. Rethinking neu-
roprotection in severe traumatic brain injury: toward
bedside neuroprotection. Front Neurol 2017; 8: 354
14. Lewis SR, Evans DJ, Butler AR, Schofield-Robinson OJ,
Alderson P. Hypothermia for traumatic brain injury.
Cochrane Database Syst Rev 2017; 9: CD001048
15. Madden LK, DeVon HA. A systematic review of the effects
of body temperature on outcome after adult traumatic
brain injury. J Neurosci Nurs 2015; 47: 190e203
16. Kelly G. Body temperature variability (part 1): a review of
the history of body temperature and its variability due to
site selection, biological rhythms, fitness, and aging.
Altern Med Rev 2006; 11: 278e93
17. Walter EJ, Hanna-Jumma S, Carraretto M, Forni L. The
pathophysiological basis and consequences of fever. Crit
Care 2016; 20: 200
18. Dippel DWJ, van Breda EJ, van der Worp HB, et al. Effect of
paracetamol (acetaminophen) and ibuprofen on body
temperature in acute ischemic stroke PISA, a phase II
double-blind, randomized, placebo-controlled trial
ISRCTN98608690. BMC Cardiovasc Disord 2003; 3: 2
19. Saxena MK, Taylor C, Billot L, et al. The effect of para-
cetamol on core body temperature in acute traumatic
brain injury: a randomised, controlled clinical trial. PloS
One 2015; 10. e0144740
- 775
20. de Ridder IR, den Hertog HM, van Gemert HMA, et al. PAIS
2 (Paracetamol Acetaminophen in Stroke 2): results of a
randomized, double-blind placebo-controlled clinical
trial. Stroke 2017; 48: 977e82
21. Fang J, Chen C, Cheng H, Wang R, Ma L. Effect of
paracetamol (acetaminophen) on body temperature in
acute stroke: a meta-analysis. Am J Emerg Med 2017; 35:
1530e5
22. Ntaios G, Dziedzic T, Michel P, et al. European Stroke
Organisation (ESO) guidelines for the management of
temperature in patients with acute ischemic stroke. Int J
Stroke 2015; 10: 941e9
23. den Hertog HM, van der Worp HB, van Gemert HM, et al.
The Paracetamol (Acetaminophen) in Stroke (PAIS) trial: a
multicentre, randomised, placebo-controlled, phase III
trial. Lancet Neurol 2009; 8: 434e40
24. Thompson HJ. Evidence-base for fever interventions
following stroke. Stroke 2015; 46: e98e100
25. Cormio M, Citerio G. Continuous low dose diclofenac so-
dium infusion to control fever in neurosurgical critical
care. Neurocrit Care 2007; 6: 82e9
26. Abu-Arafeh A, Rodriguez A, Paterson RL, Andrews PJ.
Temperature variability in a modern targeted tempera-
ture management trial. Crit Care Med 2018 Feb; 46(2):
223e8. https://doi.org/10.1097/CCM.0000000000002832
[Epub ahead of print]
27. van der Worp HB, Macleod MR, Bath PM, et al. EuroHYP-1:
European multicenter, randomized, phase III clinical trial
of therapeutic hypothermia plus best medical treatment
vs. best medical treatment alone for acute ischemic
stroke. Int J Stroke 2014; 9: 642e5
28. Connolly ES, Rabinstein AA, Carhuapoma JR, et al.
Guidelines for the management of aneurysmal sub-
arachnoid hemorrhage. Stroke 2012; 43: 1711e37
29. Diringer MN, Bleck TP. Critical care management of pa-
tients following aneurysmal subarachnoid hemorrhage:
recommendations from the Neurocritical Care Society’s
multidisciplinary consensus conference. Neurocrit Care
2011; 15: 211e40
30. Hemphill JC, Greenberg SM, Anderson CS, et al. Guidelines
for the management of spontaneous intracerebral hem-
orrhage. Stroke 2015; 46: 2032e60
31. Jauch EC, Saver JL, Adams HP, et al. Guidelines for the
early management of patients with acute ischemic stroke:
a guideline for healthcare professionals from the Amer-
ican Heart Association/American Stroke Association.
Stroke 2013; 44: 870e947
32. Steiner T, Juvela S, Unterberg A, Jung C, Forsting M,
Rinkel G. European Stroke Organization guidelines for the
management of intracranial aneurysms and subarach-
noid haemorrhage. Cerebrovasc Dis 2013; 35: 93e112
33. Steiner T, Al-Shahi Salman R, Beer R, et al. European
Stroke Organisation (ESO) guidelines for the management
of spontaneous intracerebral hemorrhage. Int J Stroke
2014; 9: 840e55
Handling editor: H.C. Hemmings Jr