You are on page 1of 5

E A R LY W A R N I N G S C O R E S 2010 SNL All rights reserved

The Newborn Early Warning (NEW)


system: development of an at-risk infant
intervention system
The use of early warning systems is widespread but their use in the neonatal age group has
been under-investigated. This article describes the development of a Newborn Early Warning
traffic-light coded observation chart to enable early identification of adverse changes in
physiological parameters. Much work remains to be done but the aim of this initial project is to
allow other maternity units to consider how they can improve the safety of at-risk newborn
infants on their postnatal wards.

Damian Roland t is generally accepted in adult and between chapters in the same book, which
BMed Sci, BMBS, MRCPCH
Academic Clinical Fellow in Paediatric
I paediatric practice, that prior to acute
deterioration and subsequent transfer to
may result in different clinical approaches
(TABLE 1). The absence of, or variance in
Emergency Medicine, Leicester University intensive care, patients often show signs of published normal values illustrates a
deterioration which are either difficulty in establishing response
John Madar unrecognised or not acted upon by nursing parameters for newborns who require
MRCP, FRCPCH, FHEA and medical staff1,2. Early warning scores observation. Early warning criteria should
Consultant Neonatologist, Neonatal Unit,
based on physiological observations (heart not be so brittle as to be over sensitive and
Derriford Hospital, Plymouth
john.madar@phnt.swest.nhs.uk
and respiratory rate etc) which thus devalue the tool.
automatically trigger medical review have The majority of newborn infants are
Glenys Connolly been validated as useful ways of detecting healthy and not at risk of significant
Bsc(Hons), RGN, RSCN deterioration and prompting intervention morbidity. A second group are clearly
Advanced Neonatal Nurse Practitioner, to reduce morbidity in both adult3 and unwell or compromised and declare
Neonatal Unit, Derriford Hospital, Plymouth paediatric4 populations. themselves as justifying enhanced levels of
To our knowledge, no such tools have care. Between these are those well babies
Keywords
been developed or fully evaluated in the whose perinatal circumstances identify
neonatal scoring systems; risk newborn population. A Medline and them as at risk of potentially significant
stratification; observation chart Embase search found no studies directly morbidity. These include, for example,
Key points related to newborn infants. One reason for those babies at risk of infection through
this may be the lack of well established streptococcal carriage, or prolonged
D. Roland, J. Madar, G. Connolly. The normal ranges for biophysical variables. rupture of membranes, or those babies
Newborn Early Warning (NEW) system: Published studies are sparse and not solely born through meconium. In addition are
devel-opment of an at-risk infant confined to the perinatal period5-7. Even the those manifesting behaviour slightly out of
intervention system. Infant 2010; 6(4):
standard textbooks have differences the normal range, but not so far as to
116-20.
1. The NEW observation chart facilitates
Source Heart rate bpm Respiratory rate
observation of babies deemed at risk
and prompts earlier review in those Examination of the Newborn and Neonatal Health. 110-160: 80-90 if 40-60
demonstrating clinical deterioration. A multidimensional approach. Ed Lorna Davies, asleep, 160+ if non-distressed
2. There was an increase in retrievable Sharon McDonald8 distressed
observations from 48% in the Examination of the Newborn. A Practical Guide. 90-140 - resting 40-60
retrospective audit to 72% in the Helen Baston, Heather Durward9 breaths/min
prospective audit. Robertons Textbook of Neonatology. 120-160 usually 35-45
3. The NEW chart threshold criteria Ed Janet M Rennie10
prompted management decisions in
Avery's Diseases of the Newborn. 40-50 newborn,
nine (47.3%) of 19 infants who required
Taesch, Ballard, Gleason11 35-60 thereafter
intervention.
4. The chart was considered beneficial by Advanced Paediatric Life Support manual12 110-160 30-40
a majority of midwives questioned
TABLE 1 Normal ranges for newborn infants heart rate and respiratory rate as published in
about its use.
standard paediatric texts.

116 VOLUME 6 ISSU E 4 2010 infant


E A R LY W A R N I N G S C O R E S

maternal health characteristics, markers of


Prenatal Perinatal Postnatal
perinatal stress (eg poor cord pH values)
Pathological Thick meconium Grunting
or the need for significant resuscitation at
cardiotocograph Venous cord pH <7.1 Abnormal movements birth (TABLE 2). These indicators are based
Scalp pH<7 Ventilatory support >3 Any ongoing concerns on well established physiological principles
Group B Streptococcus risk minutes At the request of reviewing with an incomplete evidence base to back
Premature rupture of Five minute APGAR <8 medical or neonatal staff them up13,14.
membranes (PROM)
Retrospective review
TABLE 2 At-Risk Newborn Infant (ARNI) criteria.
Using the NICU admission records the
overtly identify them as clearly unwell. (around 450 admissions/year). medical notes of term infants over 2.5kg
Such babies fall into a group where the Internal guidelines include those for the who presented to the neonatal unit from
ability to generate a series of structured identification of so called At-Risk either the postnatal wards or the
observations with evolving trends in Newborn Infants or ARNIs. These are transitional care ward over a two year
physiological parameters permits staff of infants who are deemed to be at increased period were identified. These notes were
varying experience to more clearly deter- risk of postnatal morbidity by virtue of examined for general demographic data,
mine health, stability and the potential pre-identified indicators such as adverse whether the infant had been correctly
need for further intervention. The rapidity
with which a baby can become unwell
drives a need for clear pointers towards Affix AT RISK INFANT OBSERVATION CHART
more aggressive intervention. The early Label
identification of an unwell infant may, for Here Neonatal Early Warning (NEW)
example, prompt attention to airway or
Date
breathing support, or the early
administration of antibiotics and prevent Time
Airway/Breathing
significant morbidity and even mortality. (RR HR Temp CNS
The aims of this study were: (.) (x) (o)
To categorise observations on newborn
<90%

85 200 38.2
Sats
O2

infants in order to formulate prompts for


assessment/intervention the early 80 190 38.0
warning score.
Grunting (G) or
O2 Sats 90-94%

To develop a recording tool for observa- 75 180 37.8

Irritable (I)
tions to help generate such a score and 70 170 37.6
prompt appropriate action the
Newborn Early Warning (NEW) chart 65 160 G 37.4
To assess the charts effectiveness in
clinical practice 60 150 37.2

55 140 37.0
Materials and methods
Wakes to feed
Two studies were carried out: 50 130 36.8
A retrospective review of observations on
Pink

45 120 36.6
babies admitted to the neonatal unit to
compare key observations with proposed 40 110 36.4
early warning criteria and determine
whether assessment against these criteria 35 100 36.2
would have altered management. 30 90 36.0
A prospective study of at-risk babies
Jittery
(J)

observed using the NEW chart to deter- 25 80 35.8


mine effectiveness of the chart as a clini-
20 70 35.6
cal tool. Dusky
Derriford hospital is a network neonatal (D) 60
intensive care unit (previously termed a
Indicate any associated features/symptoms present (CNS or airway) using letters.
level 3 unit) within the Peninsula Neonatal
network with around 4,400 deliveries a All observations in Green No action. Continue four hourly observations.
year. Babies are looked after on the One observation in Yellow Contact neonatal team or senior midwife.
Verbal management plan or review to be
postnatal wards, but can be admitted to a implemented. Repeat observations in 30 mins.
15 bedded transitional care ward (TCW) Two observations in Yellow or One in Red Immediate review required.
with their mothers (around 900 Seizures, apnoeas or obvious cyanosis Immediate review required.
admissions/year) or to the neonatal
intensive care unit (NICU) if more unwell FIGURE 1 Revised Newborn Early Warning Observation Chart containing some sample entries.

infant VOLUME 6 ISSU E 4 2010 117


E A R LY W A R N I N G S C O R E S

Total infants
122

Fulfil ARNI criteria Not ARNI criteria


62 (51%) 60 (49%)

Recognised at the time Not recognised


52 (84%) 10 (16%)

Observations recorded Observations recorded Observations not recorded


25 5 (8%) 55 (92%)

NEW triggers activation NEW triggers activation


13 4

NEW chart no action NEW chart no action


12 1

Observations not recorded


27

TABLE 3 Details of term babies admitted to the NNU/TCW from postnatal wards.

identified as an ARNI at birth (TABLE 2) and NEW programme, familiarise staff with the Reason Totals
whether observations had been recorded. NEW chart and the structure of the
Prenatal
A pilot NEW observation chart was proposed study.
CTG 9
developed providing prompts to aid in the NEW charts were made available on the
Scalp pH<7 0
identification of ARNIs and permit the labour suite and postnatal wards. The GBS 6
recording of the observed physiological criteria for using the NEW charts were PROM 29
variables of these infants using symbols, disseminated among the midwives and
Postnatal
highlighting values of concern. The chart posters highlighting the process placed
Meconium 15
was approved by the Hospital Clinical widely around the obstetric and neonatal
Cord pH <7.1 2
Records and Knowledge Service committee. department. Any child who was on a NEW Ventilatory support 1
As well as physiological observations such chart had their observations recorded four APGAR <8 1
as temperature, pulse and respiratory rate, hourly or more frequently if deemed
Postnatal
comments about the infants work of necessary.
Grunting 14
breathing or conscious level were Babies were excluded from the study if Abnormal movements 0
accommodated. Observation values were they were admitted directly to the NICU/ Concern 15
classified into red (significantly abnormal), TCW or fulfilled automatic admission Request 0
amber (abnormal) or green (normal) criteria such as being <37 weeks gestation
Unclear 9
ranges. The values used were an amalgam or <2.5kg.
TCW (child admitted directly
of those found in standard neonatal All NEW charts had an envelope attached
to the transitional care ward
textbooks selected to ensure chart scales so brief details of the infant could be sent to
because of gestational age) 15
were not unwieldy. Values in the charts the study administrator as soon as
Other (infant readmitted at
amber band were in keeping with the upper observations were commenced. All infants
five days of age) 1
range of normal physiological notes were collated when the study was
completed. Ethical approval was granted by Total 117
measurements.
Clinical observations from the group of the local relevant ethical committee. TABLE 4 At-Risk Newborn Infant (ARNI)
An intervention was defined as an infant criteria for enrolled infants: prospective study.
ARNIs were then plotted on the NEW
chart to see whether the pre-identified receiving an investigation (blood test or
but half would have been reviewed earlier
trigger criteria would have prompted CXR), treatment (antibiotics) or transfer
(13/25) by a neonatal doctor or nurse
earlier medical review. to another care environment.
practitioner if their observations had been
Based on the results of this retrospective A questionnaire was sent to all midwives
charted on the NEW chart. Of the babies
audit a revised chart was generated for the to obtain qualitative data on their thoughts
admitted not classified as ARNIs, few had
subsequent prospective study with on the process.
observations recorded (5/55 8%). This
modified trigger values (FIGURE 1). audit was of infants admitted to the NICU
Results
and does not contain data on those infants
Prospective study Retrospective review who were safely discharged home. Based
The results of the retrospective review were The initial audit identified 122 term on this data the decision to conduct a
used to inform an educational programme infants, 51% of these infants fulfilled ARNI prospective study was made.
including presentations and written criteria. Eighty-four per cent were correctly
material. It was aimed at midwifery, identified as such (TABLE 4). Only 48% Prospective study
nursing and medical staff in the maternity (25/52) of those infants recognised as Over a three month period information
unit and designed to raise awareness of the being ARNIs had observations recorded, was collected on 117 infants who had been

118 VOLUME 6 ISSU E 4 2010 infant


E A R LY W A R N I N G S C O R E S

Surname
First Name
Hospital No.
NEWBORN EARLY WARNING NHS Number
OBSERVATION CHART FOR NEWBORN INFANTS D.O.B. Affix patient label here

DATE
TIME

39.0
38.0
TEMP

37.0

36.0
35.0

85
80
75
70
65
RESPIRATION

60
55
50
45
40
35
30
25
20

GRUNTING

200
190
180
170
160
150
HEART RATE

140
130
120
110
100
90
80
70
60

PINK (>94%)
COLOUR
90-94%
(SpO2)
DUSKY/BLUE (<90%)

ACTIVE/WAKES TO FEED
JITTERY/IRRITABLE
NEURO FLOPPY/DIFF TO ROUSE

SEIZURES

RED
SCORE
AMBER

ALL OBSERVATIONS IN WHITE CONTINUE OBERVATIONS 4 HOURLY OR AS REQUESTED.

CONTACT SHO/ANNP/SENIOR MIDWIFE. VERBAL MANAGEMENT PLAN OR REVIEW. REPEAT


RESPONSE ONE IN AMBER OBSERVATIONS IN 30 MINUTES.

TWO IN AMBER OR ONE IN RED IMMEDIATE REVIEW

PHNT NEONATAL UNIT 2009 File alongside other observation charts VERSION 6 March 2010
FIGURE 2 Final version of the NEW observation chart separating out each variable to improve clarity.

infant VOLUME 6 ISSU E 4 2010 119


E A R LY W A R N I N G S C O R E S

recognised as being ARNIs. Based on an interventions on babies who were deemed from Liverpool developed as a result of the
average of 4,600 deliveries per year, unstable by virtue of transgressing the Confidential Enquiry of Maternal and
approximately 10% (468/4600) of predefined criteria. On a pragmatic basis Childhood Health review. A similar chart
deliveries at Derriford hospital result in an the chart identified nearly 50% of those is being used at the Royal Free in
ARNI being born. The breakdown of the infants where intervention was deemed Hampstead, UK (personal communication
specific criteria for this are shown in TABLE clinically appropriate. No direct feedback Vivienne van Someren, 2009). This chart
4. Of 117 identified, only 84 charts were was given about the chart producing separates out the clinical variables,
available for review (71.2%). Nineteen unnecessary intervention apart from the arguably making it easier to determine
infants received an intervention as per the difficulties with the temperature scale. individual trends. No single chart is likely
predefined criteria and in nine this Ultimately however, it is not the chart, or to cover the needs of all units, but
occurred as a result of the NEW chart. the highlighting of a set of observations establishing the principle and providing an
One infant was admitted to the NICU that should prompt intervention, but the effective template may help others develop
directly from the postnatal wards who full clinical evaluation of the baby that similar tools.
developed ABO incompatibility on day 2 subsequently follows. The ability to clearly The NEW observation chart is but one
of life. A chart had been provided for this assess trends in observations may form an component of the systems that need to be
infant although the reasons for this are important part of that evaluation and is in place to ensure optimal care for these
unclear. The chart did not affect the one of the attributes of the observation babies. This work has demonstrated that
infants management. chart. The NEW chart itself is but one such charts can help those looking after
A sample of midwives views on the component of a system of care and cannot such babies target at risk newborn infants
NEW system were obtained via function effectively without the other more effectively.
questionnaire. Notable responses included: elements. Having adequate numbers of
A majority felt the chart was beneficial. staff able to undertake accurate References
Many commented that the chart made observations is a pre-requisite, with clear 1. Franklin C., Mathew J. Developing strategies to
them more aware of the normal parame- arrangements for subsequent prevent inhospital cardiac arrest: analyzing
responses of physicians and nurses in the hours
ters for a newborn. communication of concern and an ability
before the event. Crit Care Med 1994; 22: 244-47.
Around half felt the chart was overcom- to respond effectively to those concerns. 2. Schein R.M., Hazday N., Pena M. et al. Clinical
plicated and suggested changes. It was felt Also of note is the fact that direct entry antecedents to in-hospital cardiopulmonary arrest.
a different style of might be easier to midwifery students may have had very Chest 1990; 98: 1388-92.
interpret. limited exposure to or training in the care 3. Stubbe C.P., Kruger M., Rutherford P., Gemmell L.
Validation of a modified Early Warning Score in
of the newborn baby and little on the
medical admissions. Q J Med 2001; 94: 521-26.
Conclusion recognition of the unwell infant. Hard 4. Duncan H., Hutchison J., Parshuram C. The pediatric
Our study indicated that many infants pressed staff on labour ward and postnatal early warning system score: A severity of illness
achieved at risk criteria, often prompting wards need effective tools to help them in score to predict urgent medical need in hospitalized
intervention in terms of investigations, the identification and observation of these children. J Crit Care 2006; 21(3): 271-78.
vulnerable babies. 5. Montague T., Taylor P., Stockton R., Roy D., Smith E.
anti-microbial management or transfer to
The spectrum of cardiac rate and rhythm in normal
a higher level of care. It is important robust It is vital to address any staff reservations
newborns. Pediatr Cardiol 1982; 2: 33-38.
procedures are instituted to avoid about the format of the chart. In the 6. Rusconi F., Castagneto M., Gagliardi L., Leo G. et al.
unnecessary morbidity and perhaps original version, the temperature scale was Reference values for respiratory rate in the first 3
mortality through inadvertent delay. The felt to be over sensitive, prompting review years of life. Pediatrics 1994; 94: 350-55.
benefits of early identification of instability and potential intervention when 7. Hooker E., Danzl D., Brueggmeyer M., Harper E.
unnecessary. The format of the chart with Respiratory rates in pediatric emergency patients. J
and of necessary intervention are obvious
Emerg Med 1992; 10: 407-10.
and an early warning chart with clear different symbols for each variable was also
8. Davies L., McDonald S., eds. Examination of the
prompts for action is one tool for felt to confuse and produce an Newborn and Neonatal Health. A multidimensional
facilitating this. Our locally designed and overcrowded display which was difficult to approach. Churchill Livingstone/Elsevier. 2008.
implemented chart appears to have had read. These problems were exacerbated by 9. Baston H., Durward H. Examination of the
staff using poor quality photocopies of the Newborn. A Practical Guide. Routledge. 2001.
some success in identifying infants at an
10. Rennie J.M., ed. Robertons Textbook of
earlier stage than would have occurred in original chart, rather than high quality
Neonatology. Fourth Edition 2005. Churchill
their absence. The chart itself may have reproductions. Budgetary constraints also Livingstone/Elsevier. 2005.
been the arbiter of the increased detection compromised the original charts by the use 11. Taesch H.W., Ballard R., Gleason C.A. Averys
rate, but the very exercise of introducing of grey scale rather than colour banding. Diseases of the Newborn 8th edition. Elsevier
the charts, and the educational package Further work and greater numbers are Saunders. 2004.
needed in order to evolve a working model 12. Mackway-Jones, Molyneux E., Phillips B., Wieteska S.,
surrounding this may also have had an
eds. Advanced Paediatric Life Support: The Practical
effect in raising awareness. which is acceptable to all staff and
Approach 4th Edition. Blackwell Publishing. 2005. 7-14.
The true effect of earlier detection on validation of the results in a different 13. Madar J. Clinical risk management in the newborn
longer term morbidity and mortality is clinical setting should take place. As a and neonatal resuscitation. Semin Fetal Neonatal
difficult to define with the small numbers result of feedback a further version of the Med 2005; 10: 45-61.
chart has been designed which in pilot 14. Victory R., Penava D., Da Silva O., Natale R.,
of babies involved in this study. However,
Richardson B. Umbilical cord pH and base excess
intuitively, earlier management might be testing has proved more popular with
values in relation to advers outcome events for
considered a positive outcome, unless midwifery staff (FIGURE 2). This chart is infants delivering at term. Am J Obstet Gynaecol
prompting unnecessary investigations and based on an obstetric early warning system 2004; 191: 2021-28.

120 VOLUME 6 ISSU E 4 2010 infant