Embed Size (px)
Citation preview
Accepted Manuscript
Title: Neonatal Early Warning Tools for recognising andresponding to clinical deterioration in neonates cared for in thematernity setting: a retrospective case control study
Authors: Michelle Paliwoda Karen New RN,RM,PhD FionaBogossian
PII: S0020-7489(16)30076-1DOI: http://dx.doi.org/doi:10.1016/j.ijnurstu.2016.06.006Reference: NS 2767
To appear in:
Received date: 4-11-2015Revised date: 9-6-2016Accepted date: 14-6-2016
Please cite this article as: Paliwoda, M., New, K., Bogossian, F.,Neonatal Early WarningTools for recognising and responding to clinical deterioration in neonates cared for in thematernity setting: a retrospective case control study, International Journal of NursingStudies (2016), http://dx.doi.org/10.1016/j.ijnurstu.2016.06.006
This is a PDF file of an unedited manuscript that has been accepted for publication.As a service to our customers we are providing this early version of the manuscript.The manuscript will undergo copyediting, typesetting, and review of the resulting proofbefore it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers thatapply to the journal pertain.
Page 1 of 22
Accep
ted
Man
uscr
ipt
1
Title 1 2 Neonatal Early Warning Tools for recognising and responding to clinical deterioration in neonates cared for 3 in the maternity setting: a retrospective case control study 4 5 6 Authors 7 8 Michelle Paliwoda 9 Affiliations: The University of Queensland, School of Nursing and Midwifery and The Royal Brisbane & 10 Women’s Hospital. 11 12 Dr. Karen New RN, RM, PhD 13 Affiliations: The University of Queensland, School of Nursing and Midwifery, UQ Centre of Clinical 14 Research and The Royal Brisbane & Women’s Hospital. 15 16 Associate Professor Fiona Bogossian 17 Affiliations: The University of Queensland, School of Nursing and Midwifery, UQ 18 19 20 Contact details 21 22 Michelle Paliwoda (Corresponding author) 23 Registered nurse 24 Grantley Stable Neonatal Nursery 25 L5 Ned Hanlon Building 26 Royal Brisbane and Women’s Hospital 27 Phone: (W) +61 7 3646 7846 28 Email: [email protected] 29 30 Dr Karen New 31 Midwifery Clinical Academic Fellow, RN, RM, PhD 32 UQ School of Nursing, Midwifery and Social Work 33 UQ Centre for Clinical Research | Royal Brisbane & Women's Hospital 34 Room 337, Level 3, Chamberlain Building, The University of Queensland QLD 4072 35 F: +61 7 3346 5599 36 Email: [email protected] 37 38 Associate Professor Fiona Bogossian 39 Director, Research Higher Degrees 40 School of Nursing, Midwifery and Social Work 41 Level 3, Chamberlain Building, The University of Queensland 4072 42 Email: [email protected] 43 44 Abstract 45
Background: All newborns are at risk of deterioration as a result of failing to make the transition to extra 46
uterine life. Signs of deterioration can be subtle and easily missed. It has been postulated that the use of an 47
Early Warning Tool may assist clinicians in recognising and responding to signs of deterioration earlier in 48
neonates, thereby preventing a serious adverse event. 49
50
Objective: To examine whether observations from a Standard Observation Tool, applied to three neonatal 51
Early Warning Tools, would hypothetically trigger an escalation of care more frequently than actual 52
escalation of care using the Standard Observation Tool. 53
Page 2 of 22
Accep
ted
Man
uscr
ipt
2
54
Design: A retrospective case-control study. 55
56
Setting: A maternity unit in a tertiary public hospital in Australia 57
58
Methods: Neonates born in 2013 of greater than or equal to 34+0 weeks gestation, admitted directly to the 59
maternity ward from their birthing location and whose subsequent deterioration required admission to the 60
neonatal unit, were identified as cases from databases of the study hospital. Each case was matched with 61
three controls, inborn during the same period and who did not experience deterioration and neonatal unit 62
admission. Clinical and physiological data recorded on a Standard Observation Tool, from time of admission 63
to the maternity ward, for cases and controls were charted onto each of three Early Warning Tools. The 64
primary outcome was whether the tool ‘triggered an escalation of care’. Descriptive statistics (n, %, Mean 65
and SD) were employed. 66
67
Results: Cases (n=26) comprised late preterm, early term and post term neonates and matched by gestational 68
age group with 3 controls (n=78). Overall, the Standard Observation Tool triggered an escalation of care for 69
92.3% of cases compared to the Early Warning Tools; New South Wales Health 80.8%, United Kingdom 70
Newborn Early Warning Chart 57.7% and The Australian Capital Territory Neonatal Early Warning Score 71
11.5%. Subgroup analysis by gestational age found differences between the tools in hypothetically triggering 72
an escalation of care. 73
74
Conclusions: The Standard Observation Tool triggered an escalation of care more frequently than the Early 75
Warning Tools, which may be as a result of behavioural data captured on the Standard Observation Tool and 76
escalated, which could not be on the Early Warning Tools. Findings demonstrate that a single tool applied to 77
all gestational age ranges may not be effective in identifying early deterioration or may over trigger an 78
escalation of care. Further research is required into the sensitivity and specificity of Early Warning Tools in 79
neonatal sub-populations. 80
81
Keywords: early warning tool, early warning score, newborn, maternity newborn care, neonate82
Page 3 of 22
Accep
ted
Man
uscr
ipt
3
83
Introduction 84
One of the greatest challenges a neonate must overcome is the transition to extrauterine life. All newborns 85
are at risk of deterioration as a result of failing to make this transition. Physiological immaturity, related to 86
gestational age, and the impact of infection on the immunologically immature neonate can alter the success 87
of adaptation (Ygberg & Nilsson, 2011; Graves & Haley, 2013). Signs of deterioration in the newborn can 88
be subtle and easily missed (Satar & Ozlu, 2012). It has been postulated by health authorities in Australia 89
and overseas that the use of an Early Warning Tool may help clinicians identify early signs of deterioration 90
and therefore respond promptly to prevent serious adverse events in acute health care settings (National 91
Patient Safety Agency, 2007; Clinical Excellence Commission, 2013; Paliwoda & New, 2015). 92
93
Early Warning Tools assist clinicians identify early deterioration by using a systematic process of charting 94
patient vital signs against pre-determined vital sign parameters (Australian Commission on Safety and 95
Quality in Health Care, 2012). Early Warning Tools vary in design but are generally coded with varying 96
colours or shades indicating worsening abnormal parameters, which is designed to alert the clinician by way 97
of set action prompts that an escalation of care is required (McLellan & Connor, 2013; Olroyd & Day, 2011; 98
Paliwoda & New, 2015). Other tools are based on scoring systems or a combination of both where if an 99
aggregate number exceeds a predetermined threshold an escalation process determines the clinician’s path of 100
intervention (Australian Commission on Safety and Quality in Health Care, 2012). 101
102
Safety and quality units of health care facilities worldwide have undertaken steps to assist clinicians 103
recognise and respond to clinical deterioration (National Health Service Trust, n.d; Institute for Healthcare 104
Improvement, 2016; An Roinn Slainte Department of Health, 2014). In 2010, the Australian Commission on 105
Safety and Quality Health Service mandated 15 standards to improve patient outcomes in acute health 106
settings of the 15; 10 apply to direct patient care (ACSQHS, 2010). Standard 9: Recognising and Responding 107
to Clinical Deterioration in Acute Health Care applies to all patients including babies cared for in maternity 108
health settings (ACSQHS, 2012). The study hospital has addressed this standard for adult and paediatric 109
clients by implementing Early Warning Tools (Queensland Government, 2012a). However, to date, an Early 110
Warning Tool has not been implemented for neonates. 111
112
The National Consensus Statement of Australia in 2010 recommended six key physiological observations: 113
respiratory rate, oxygen saturations, heart rate, blood pressure, temperature, and level of consciousness, be 114
incorporated into the development of Early Warning Tool to assist in identification of early deterioration 115
(ACSQHC, 2010). Importantly it could be argued that these ‘all population’ observations do not pertain to 116
newborns in the maternity ward, where presently key physiological observations such as blood pressure and 117
oxygen saturations are not routinely undertaken in the care of newborns in all maternity settings (Queensland 118
Government, 2012; King Edward Memorial Hospital, 2014). It could be further argued that newborn specific 119
observable behaviours that are indicative of deterioration, such as poor feeding, ‘spilling’, failing to wake for 120
Page 4 of 22
Accep
ted
Man
uscr
ipt
4
feeds, or falling asleep during feeding; and parental concern would be more applicable (Paliwoda & New, 121
2015; New South Wales Department of Health, 2011). In response to the Standard, a neonatal Early Warning 122
Tool is being developed for rollout across Queensland (Patient Safety Unit, Queensland Department of 123
Health, personal communication, May 29, 2015). While in other states of Australia, individual hospitals have 124
developed, adapted or introduced Early Warning Tools based on the all population key physiological data for 125
determining clinical deterioration (New South Wales Department of Health, 2012; New South Wales 126
Department of Health, 2013). 127
128
An earlier literature review found there is no ‘gold standard’ or validated early warning tool for use in 129
neonates, and studies in the adult, paediatric and neonatal population have demonstrated that the efficacy of 130
Early Warning Tool are mixed (Paliwoda & New, 2015). In light of this finding and the proliferation of 131
neonatal Early Warning Tool across Australia and internationally, we designed a retrospective case-control 132
study to examine and hypothetically compare the performance of three Early Warning Tools for neonatal use 133
in the maternity setting. 134
135
Objective 136
To examine whether observations from a Standard Observation Tool, applied to three neonatal Early 137
Warning Tools, would hypothetically trigger an escalation of care more frequently than actual escalation of 138
care using the Standard Observation Tool. 139
140
Materials and Methods 141
Study Design 142
A retrospective matched case control study. 143
144
Setting 145
A maternity unit in a tertiary public hospital in Australia. 146
147
Participants 148
Following ethical approval from institutional review boards, neonates inborn between January and December 149
2013, of greater than or equal to 34+0 weeks gestation, admitted directly to the maternity ward from their 150
birthing location (birth suite or operating theatre), were identified from databases of the study hospital. These 151
neonates were then categorised into their gestational age groups, i.e., late preterm (34+0-36+6), early term 152
(37+0-38+6), and post term (≥42+0) and from these groups the cases and controls were identified 153
(Supplementary Flowchart 1). The cases were those neonates deemed well and admitted to the maternity 154
ward but who subsequently deteriorated and required admission to the neonatal unit. All late preterm and 155
post term cases were included. A computer software program was used to randomly select a sample from the 156
early term cases, and then to randomly select three matched controls for each case in each age group. The 157
controls were neonates deemed well, admitted to the maternity ward, and subsequently discharged from the 158
Page 5 of 22
Accep
ted
Man
uscr
ipt
5
maternity ward without adverse event.159
Page 6 of 22
Accep
ted
Man
uscr
ipt
6
160
Instrumentation 161
162
Standard Observation Tool (SOT) 163
The study hospital uses a Vigilance and Baby Management Chart (referred to in this study as the Standard 164
Observation Tool), which was developed by the study hospital and to the best to our knowledge has not been 165
validated nor has its clinical effectiveness been examined. This tool is used routinely for all neonates at the 166
study hospital from the time of birth until discharge in all maternity settings (excluding the neonatal units). 167
The policy of the study hospital is that all neonates have observations every 15 minutes for the first two 168
hours in birth suites/centre. This includes oxygenation saturations and blood pressure monitoring prior to 169
discharge to the maternity ward. Once transferred to the maternity ward, newborns must have observations at 170
least eight hourly, although in the event of maternal risk factors, such as Streptococcus Group B colonisation 171
or an intrapartum event, the frequency of observations is increased (Queensland Government, 2012b). 172
173
The Standard Observation Tool allows for the documentation of routine physiological, clinical, and 174
behavioural observations such as respiration rate, heart rate, blood glucose level, method of feeding, volume 175
of feed consumed and documentation of elimination. Oxygenation saturations and blood pressure are not 176
routinely monitored in the maternity ward. The tool also incorporates a comments column, which allows for 177
documentation of objective observable behaviours such as irregular heartbeat, vomiting, distended abdomen, 178
and the degree to which the neonate is unsettled, inconsolable, grimacing (pain), or sleepy. The comments 179
column appears to facilitate the documentation of actions taken and/or escalation of care in response to any 180
abnormal physiological or behavioural observations. 181
182
Early Warning Tools 183
Despite the absence of a published, validated ‘gold standard’ neonatal Early Warning Tool (Paliwoda & 184
New, 2015), a number of Early Warning Tools are being used in neonates. However, we considered several 185
of these as not being suitable for neonates in maternity settings for reasons such as: a wide age range, for 186
example 0-3 months or 0-12 months; specific design for use in paediatric settings, or the set physiological 187
parameters fell outside the range for newborns (Dawson, Omar, Kamlin, Vento, Wong, Cole, Donath, Davis, 188
& Morley, 2010; King Edward Memorial Hospital, 2014, Fyfe, Yiallourou & Horne, 2012; Ching, Lavin & 189
Blair, 2011). Thus, for inclusion in this study the Early Warning Tools needed to be specific for the neonatal 190
age range (0-28 days) and have been implemented in neonatal settings in Australia or the United Kingdom. 191
The Early Warning Tools selected for this study have been evaluated following introduction into the clinical 192
setting (Clinical Excellence Commission, 2013; Roland, Madar, & Connolly, 2010; Australian Capital 193
Territory, 2013), though, to the best of the authors knowledge they have not been validated. The authors 194
were not involved in the construction of any of the Early Warning Tools included in this study, which are 195
summarised below and available in Supplement One. 196
197
Page 7 of 22
Accep
ted
Man
uscr
ipt
7
1) ACT Health ‘Newborn Early Warning Score’ (ACT NEWS) 198
The ‘Newborn Early Warning Score’ was developed by key stakeholders of the Centenary Hospital for 199
Women & Children and Calvary Bruce Private Hospital for Australian Capital Territory Health. It was 200
randomly audited following implementation in the postnatal environment and the results presented at the 201
2014 ‘Managing the Deteriorating Patient Conference’ (Kecskes, Ovington, & Slater, 2014). This tool 202
utilises a colour-coded (yellow, pale blue, light purple, dark purple) single parameter aggregate scoring 203
system which is designed for use in neonates aged between 0-<1 month post term and incorporates 204
subjective clinical assessment of respiratory effort and objective quantifiable data consisting of heart rate, 205
respiratory rate, oxygen saturation monitoring, and temperature. The frequency and duration of observations 206
are informed by the completion of a ‘newborn risk assessment’ that is part of this four-page document. It is 207
intended that this assessment be completed within one-hour of birth. Neonates with no identifiable risk 208
factors are required to have the minimum observations (12th hourly). If the neonate has an aggregate score of 209
greater than or equal to four, the tool provides prompts for escalation of care and clinical review. 210
211 2) New South Wales Health 212
The Clinical Excellence Commission developed the Between the Flags ‘Safety Net’ system in 2010 to enable 213
clinicians to identify signs of clinical deterioration and provide timely intervention to prevent a serious 214
adverse event and improve patient outcomes (Clinical Excellence Commission, 2013). The system 215
incorporates population specific Early Warning Tools, for example, adults (General, and Emergency), 216
paediatrics (under 3 months, 3 to 12 months, 1 to 4 years, 5-11 years, and over 12 years), maternity 217
(Standard Maternity and antenatal), and newborn (standard). The Standard Neonatal Observation Chart 218
(SNR110.014) is a single parameter colour-coded (blue, yellow, red) track and trigger system designed for 219
use in special care units and maternity settings for neonates under one month of corrected age (Clinical 220
Excellence Commission, 2012). The tool incorporates subjective clinical assessment of respiratory effort and 221
behaviour and objective quantifiable data consisting of heart rate, oxygen saturation monitoring, blood 222
pressure, temperature and blood glucose monitoring. When observations fall within one or more of the 223
colour-coded zones, instructions on the overleaf of this four-page tool provide details of the expected 224
escalation of care for each of the colour zones. 225
226
3) Newborn Early Warning Observation Chart (UK NEW) 227
Dr Damian Roland and Dr John Madar of the Plymouth Hospitals National Health Service Trust, United 228
Kingdom developed this Early Warning Tool (Roland, Madar, & Connolly, 2010). The Newborn Early 229
Warning Observation Chart is a single parameter colour-coded system of red, yellow, and green that prompts 230
escalation of care for clinical, physiological and observational data (heart rate, respiration rate, temperature 231
and oxygen saturations). Observations or symptoms that would indicate central nervous system or airway 232
compromise are additional assessable items charted using letters and can also trigger escalation of care. If all 233
observations are in the green zone, no escalation of care is required. If two or more observations are in the 234
yellow zone or one in the red zone, then immediate review is required. The clinical effectiveness and utility 235
Page 8 of 22
Accep
ted
Man
uscr
ipt
8
of this tool has been examined but not rigorously tested to assess reliability and validity (Roland, Madar, & 236
Connolly, 2010). 237
238
Data parameters 239 For this study, physiological observations were considered abnormal if they fell outside the normal reference 240
range of the study hospital protocol. In charting the observations from the Standard Observation Tool onto 241
each of the three other Early Warning Tools, it became apparent, with the exception of respiratory rate, there 242
is a lack of consensus on ‘normal’ reference ranges (Table 1). 243
244
Table 1: Normal physiological reference ranges for the Standard Observation Tool (SOT) and the three Early 245
Warning Tools 246
Observation SOT ACT NEWS NSW Health UK NEW
Respiration
≥30bpm
≤60bpm
≥30bpm
≤60bpm
≥30bpm
≤60bpm
≥ 30bpm
≤ 60bpm
Heart Rate
≥120 bpm
≤160bpm
≥ 90bpm
≤ 160bpm
≥ 110bpm
≤160bpm
≥ 90bpm
≤ 150bpm
Temperature
≥36.5°C
≤ 37.4°C
≥36.5°C
≤ 37.5°C
≥36.5°C
≤ 37.5°C
≥ 36.0 °C
≤ 37.2°C
Blood Glucose level
≤ 2.6mmol/L
≥15mmol/L Not indicated
≤ 3.0mmol/L
≥ 10mmol/L Not indicated
bpm – breaths per minute (respiration) or beats per minute (heart rate); C - Degrees Celsius; mmol/L - 247 Millimols per litre 248 249 “Triggered an escalation of care” 250 For this study, “triggered an escalation of care”, was deemed to be when; a physiological observation fell 251
outside of the normal range and/or, clinician notation of a behavioural observation which was actioned or 252
care escalated as documented on the Standard Observation Tool or in the clinical notes. For the Early 253
Warning Tools, ‘triggered an escalation of care’ was hypothetical but deemed to have occurred when; either 254
one or more charted observations fell in an abnormal zone or, the aggregate score reached the number which 255
triggered an escalation of care for the individual tool. Each Early Warning Tool indicates that in the event of 256
abnormal observation/s, observation frequency should be increased and/or repeated within 30 minutes. 257
Therefore, for this study, escalation of care for the Standard Observation Tool was deemed not to have 258
occurred if observations were not repeated within 30 minutes. 259
260
Data collection 261
Between April and June 2015, the first author retrieved the medical charts of cases and controls, collected 262
and charted the source data onto each of the three Early Warning Tools. The source data were the data 263
Page 9 of 22
Accep
ted
Man
uscr
ipt
9
documented on the Standard Observation Tool. However, the clinical notes were reviewed to confirm 264
whether an escalation of care or action was taken if this was not documented on the Standard Observation 265
Tool. If the source data or documentation in the clinical notes was unclear to the first author, the second 266
author reviewed the medical chart and consensus was obtained. The data were charted from the first recorded 267
observations on admission to the maternity ward until discharge to either the neonatal unit (case) or home 268
(control). 269
270
Data analysis 271
Data were entered into IBM SPSS data editor Version 22 (SPSS Inc., Chicago, IL). Descriptive statistics (n, 272
%, Mean and SD) were used to report differences between neonatal and clinical characteristics for cases and 273
controls and to determine the frequency of cases and controls triggering an escalation of care. 274
275
Results 276
Participant characteristics 277 Neonatal and clinical characteristics for the 26 cases: late preterm (n=8), early term (n=16), post term (n=2) 278
and 78 controls are presented in Table 2. As a surrogate marker for initial wellness, we report the majority of 279
cases had Apgar scores ≥7 at 1 and 5 minutes, as did controls. 280
281 Table 2: Neonatal and clinical characteristics at birth for cases and controls 282
LPT – Late preterm; ET – Early term; PT – Post term 283
284
Performance of the early warning tools 285
The performance of each of the early warning tools is reported according to their actual and hypothetical 286
incidence of triggering an escalation of care. Firstly, we report the summary of results for triggering an 287
escalation of care for cases and controls (Table 3). Followed by the cases that did or did not trigger an 288
escalation of care on the Standard Observation Tool, and whether these cases hypothetically triggered on 289
each Early Warning Tool (Table 4). We also report controls that either had an observation within the normal 290
Characteristics
Cases (n=26) Controls (n=78)
LPT (n=8) n (%)
ET (n=16) n (%)
PT (n=2) n (%)
LPT (n=24) n (%)
ET (n=48) n (%)
PT (n=6) n (%)
Sex Male 4 (50) 8 (50) 1 (50) 8 (33.3) 23 (47.9) 2 (33.3)
Female 4 (50) 8 (50) 1 (50) 16 (66.7) 25 (52.1) 4 (66.7) Weight (g) mean (± SD)
2845 (± 0.28) 3060 (± 0.42) 4845 (± 0.50) 2899 (± 0.31) 3154 (± 0.41) 3429 (± 0.33)
Type of Birth
Vaginal 5 (62.5) 8 (50) 0 (0) 12 (50) 18 (37.5) 5 (83.3) Caesarean
Section 3 (37.5) 8 (50) 2 (100) 12 (50) 30 (62.5) 1 (16.7)
APGAR (1 minute)
≤ 6 1 (12.5) 0 (0) 0 (0) 1 (4.2) 1 (2.1) 0 (0)
≥ 7 7 (87.5) 16 (100) 2 (100) 23 (95.8) 47 (97.9) 6 (100)
APGAR (5 minutes)
≤ 6 1 (12.5) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)
≥ 7 7 (87.5) 16 (100) 2 (100) 24 (100) 48 (100) 6 (100)
Page 10 of 22
Accep
ted
Man
uscr
ipt
10
reference range of the hospital or that did not trigger an escalation of care on the standard observation tool, 291
but hypothetically triggered on at least one Early Warning Tool (Table 5). 292
293
Table 3: Actual and hypothetical incidence of triggering an escalation of care for cases and controls by 294 gestational age group and tools 295
LPT – Late preterm; ET – Early term; PT – Post term; SOT – Standard Observation Tool 296
297
Of the 26 cases who were well and subsequently deteriorated, abnormal observations recorded on the 298
Standard Observation Tool triggered an escalation of care for 24 (92.3%) cases. The observations charted on 299
each of the three Early Warning Tools hypothetically triggered an escalation of care in between 3 to 21 (11.5 300
- 80.8%) of cases, depending on the respective tool (Table 3). Subgroup analysis by gestational age revealed 301
that the New South Wales Health Tool was hypothetically more responsive in triggering an escalation of care 302
for late preterm neonates than the Standard Observation Tool, and for all gestation age groups compared to 303
the other two Early Warning Tools (Table 3). 304
305
The observations of 78 controls were reviewed to ascertain the incidence of observations that did or did not 306
trigger an escalation of care by the Standard Observation Tool and whether these observations would have 307
hypothetically triggered an escalation of care on any of the Early Warning Tools. There were 32 (41.0%) 308
controls which had an observation recorded on the Standard Observation Tool in which the clinician 309
indicated an escalation of care. The number of control observations that hypothetically triggered an 310
escalation of care for each of the Early Warning Tools varied between 1.3 – 69.2% (Table 3). 311
312
Cases triggering an escalation of care 313 314 Cases that did or did not trigger an escalation of care were identified by the standard observation tool and 315
examined using each of the early warning tools (Table 4). 316
317
318
Gestational age group
SOT ACT NEWS
NSW Health
UK NEW
Yes n (%)
No n (%)
Yes n (%)
No n (%)
Yes n (%)
No n (%)
Yes n (%)
No n (%)
Cases (n=26)
LPT (n=8)
6 (77.0) 2 (25.0) 2 (25.0) 6 (75.0) 7 (87.5) 1 (12.5) 5 (62.5) 3 (37.5)
ET (n=16)
16 (100) 0 (0) 1 (6.3) 15 (93.8) 12 (75.0) 4 (25.0) 9 (56.3) 7 (43.8)
PT (n=2) 2 (100) 0 (0) 0 (0) 2 (100) 2 (100) 0 (0) 1 (50.0) 1 (50.0) TOTAL 24 (92.3) 2 (7.7) 3 (11.5) 23 (88.5) 21 (80.8) 5 (19.2) 15 (57.7) 11 (42.3)
Controls (n=78)
LPT (n=24)
8 (33.3) 16 (66.7) 0 (0) 24 (100) 13 (54.2) 11 (45.8) 9 (37.5) 15 (62.5)
ET (n=48)
22 (43.8) 26 (56.3) 1 (2.1) 47 (97.9) 38 (79.2) 10 (20.8) 15 (31.3) 33 (68.8)
PT (n=6) 2 (33.3) 4 (66.6) 0 (0) 6 (100) 3 (50.0) 3 (50.0) 0 (0) 6 (100) TOTAL 32 (41.0) 46 (59.0) 1 (1.3) 77 (98.7) 54 (69.2) 24 (30.8) 24 (30.8) 54 (69.2)
Page 11 of 22
Accep
ted
Man
uscr
ipt
11
Standard Observation Tool 319
Those observations identified by clinicians as outside of normal reference range by the study hospital, 320
triggered an escalation of care in 24 (92.3%) cases using the Standard Observation Tool. The two cases that 321
had no documentation of escalation of care were late preterm newborns, one with a heart rate of 180 beats 322
per minute and the other with a respiration rate of 148 breaths per minute (Table 4). 323
324 The Australian Capital Territory Neonatal Early Warning Score (ACT NEWS) 325
Of the 26 cases, the tool that hypothetically triggered an escalation of care the least was the Australian 326
Capital Territory Neonatal Early Warning Score. The observations triggered an escalation of care for three 327
neonates (11.5%), two late preterm (25.0%) and one early term (6.3%) (Table 3). Of these neonates, one 328
(Case 5) had an observation that entered the dark purple zone which hypothetically triggered a medical 329
emergency team call and the other two (Cases 8 and 22), had abnormal observations that when combined, 330
resulted in an aggregate score of ≥4 thereby hypothetically triggering an escalation of care (Table 4). The 331
remaining late preterm (n=6), early term (n=15), and post term (n=2) cases in which the abnormal 332
observations did not trigger an escalation of care were a result of the pre-determined aggregate score for 333
escalation not being reached. In addition, this tool does not measure a number of physiological (e.g. blood 334
glucose level), clinical (e.g. vomiting) and behavioural observations (e.g. sleepy and not feeding), resulting 335
in the tool potentially not identifying subtle signs of deterioration for 11 of the cases (Table 4). 336
337
New South Wales Health Early Warning Tool (NSW Health) 338
Of the three Early Warning Tools, the New South Wales Health Early Warning Tool hypothetically triggered 339
an escalation of care for the majority of cases (n=21; 80.8%). This tool hypothetically triggered an escalation 340
of care for 100% of post term cases, 75% of early term cases and identified one additional late preterm 341
neonate compared to the Standard Observation Tool (Table 3). The four early term cases (10, 11, 14, and 16) 342
did not trigger because, as for the previous tool, clinical and behavioural observations have not been 343
incorporated into the design of this Early Warning Tool. 344
345 United Kingdom Newborn Early Warning Chart (UK NEW) 346
The United Kingdom Newborn Early Warning Chart hypothetically triggered an escalation of care for just 15 347
(57.7%) of the 26 cases (Tables 3 and 4). Like the Australian Capital Territory Neonatal Early Warning 348
Score tool, this tool does not facilitate blood glucose levels to trigger an escalation of care and as such, the 349
majority of cases that did not trigger across all gestational age groups were for low blood glucose levels 350
(Cases 3, 6, 7, 13, 15, 17, 19 and 26, Table 4). Likewise, physiological observations such as vomiting did 351
not trigger (Cases 10 and 14) and a tolerance for mild hypothermia, accepting axillary temperatures as low as 352
36.0°C, did not trigger an escalation of care for one neonate (Case 20). 353
354
355
356
Page 12 of 22
Accep
ted
Man
uscr
ipt
12
Table 4: Cases that did or did not trigger an escalation of care by Standard Observation Tool and Early 357 Warning Tool 358
GA – Gestational age; LPT – Late preterm; ET – Early term; PT – Post term; GA -gestational age; T-359 Temperature BGL – Blood glucose level; RR – respiration rate; HR – Heart rate; EOC – Escalation of Care; 360 SOT – Standard Observation Tool 361 362
363
364
365
366
367
368
Case (N=26)
GA Observation
Actual trigger of
EOC Hypothetically triggered an EOC
SOT ACT
NEWS NSW
Health UK
NEW
1 LPT T: 37.5C Yes No No Yes
2 LPT Increased respiratory effort Yes No Yes Yes
3 LPT BGL: 2.6mmol/L Yes No Yes No
4 LPT HR: 180 bpm [RR 50 bpm; T: 37.3C] No No Yes Yes
5 LPT RR: 148 bpm [HR: 140 bpm; T:36.6C] No Yes Yes Yes
6 LPT BGL: 1.9mmol/L Yes No Yes No
7 LPT BGL: 2.6mmol/L Yes No Yes No
8 LPT RR 24bpm; Respiratory effort; T:36.4C Yes Yes Yes Yes
9 ET HR: 80 bpm Yes No Yes Yes
10 ET Mucous vomits Yes No No No
11 ET Irregular HR Yes No No Yes
12 ET HR: 98 bpm; T: 36.3C Yes No Yes Yes
13 ET BGL: 2.1mmol/L Yes No Yes No
14 ET Vomiting Yes No No No
15 ET BGL: 1.4mmol/L Yes No Yes No
16 ET Baby sleepy, not feeding Yes No No Yes
17 ET BGL: 2.4mmol/L Yes No Yes No
18 ET Increased respiratory effort Yes No Yes Yes
19 ET BGL: 1.4mmol/L Yes No Yes No
20 ET T: 36.4C Yes No Yes No
21 ET Dusky episode Yes No Yes Yes
22 ET RR: 80 bpm; T: 36.4C Yes Yes Yes Yes
23 ET Respiratory effort Yes No Yes Yes
24 ET Dusky episode Yes No Yes Yes
25 PT T: 38.2C Yes No Yes Yes
26 PT BGL: 2.4mmol/L Yes No Yes No
Total n (%) 24 (92.3) 2 (7.7) 21 (80.8) 15 (57.7)
Page 13 of 22
Accep
ted
Man
uscr
ipt
13
Controls triggering an escalation of care 369
Of the 78 controls, 44 had observations that were within the normal reference range of the study hospital 370
thereby not requiring any escalation of care. The remaining 34 controls had either an observation that ought 371
to have triggered an escalation of care by the Standard Observation Tool (n=22) or an observation that was 372
within the normal reference range of the hospital (n=12). At least one of the Early Warning Tools 373
hypothetically triggered an escalation of care for each of the 34 controls (Table 5). 374
375
Standard observation tool 376
The 22 (64.7%) controls for which there was no escalation of care documented on the Standard Observation 377
Tool (nor in the clinical notes) were for observations such low or high temperatures (n=16, 47.1%), heart 378
rate (n=3, 8.8%), respiratory (n=1, 2.9%), or a multiple of observation (n=2, 5.9%). 379
380
Early warning tools 381
Of the 34 controls, the New South Wales Health Early Warning Tool hypothetically triggered an escalation 382
of care for 27 (79.4%) controls; the United Kingdom Newborn Early Warning Chart for 12 (35.3%) and the 383
Australian Capital Territory Newborn Early Warning Score just one (2.9%). 384
385
While the New South Wales Health Early Warning Tool was responsive to hypothetically triggering an 386
escalation of care for cases it was also responsive for controls, triggering an escalation of care for seven 387
(70%) late preterm controls, 19 (82.6%) early term and one post term control (Table 5). The triggering 388
observations were temperature variations (n=15, 44.1%); blood glucose levels (n=6, 17.6%); heart rate (n=3, 389
8.8%); respiratory (n=1, 2.9%); or a multiple of observations (n=2, 5.9%). The United Kingdom Newborn 390
Early Warning Chart was less responsive to triggering an escalation of care across all gestational age groups: 391
(late preterm (n=3, 30%), early term (n=9, 39.18%) and no post term, and the triggering observations were 392
similar to that of the New South Wales Health Early Warning Tool with the exception of blood glucose 393
levels (Table 5). While the Australian Capital Territory Newborn Early Warning Score hypothetically 394
triggered an escalation of care for just one early term control (Control 22). This set of observations (Control 395
22) hypothetically triggered on all of the Early Warning Tools, but no action was charted on the Standard 396
Observation Tool or in the clinical notes (Table 5). However, given the significance of these observations 397
and that this neonate remained a control infant, one may theorise that the observation parameters were 398
inadvertently charted in the reverse columns. That is, the heart rate was recorded in the respiratory rate 399
column and vice versa. Conversely the Australian Capital Territory Newborn Early Warning tool did not 400
trigger for a temperature of 39.9oC (RR: 42bpm and HR:132bpm) (Control 19), nor was this escalated on the 401
Standard Observation Tool. Likewise given this neonate remained a control and with no documentation in 402
the clinical notes indicating there was a problem, we speculate that this may have also been a documentation 403
error. 404
405
406
Page 14 of 22
Accep
ted
Man
uscr
ipt
14
Table 5: Controls that did or did not trigger an escalation of care on the Standard Observation Tool and by 407 Early Warning Tool 408
409
Controls (n=34)
GA
Observation
Did Not Trigger an EOC
Hypothetically triggered an EOC
SOT ACT NEWS
NSW Health
UK NEW
1 LPT BSL: 2.9mmol/L NRR No Yes No
2 LPT T: 36.4C No No Yes No
3 LPT T: 36.3C No No Yes No
4 LPT BGL: 2.8mmol/L NRR No Yes No
5 LPT T: 37.5C No No No Yes
6 LPT BGL: 2.9mmol/L NRR No Yes No
7 LPT T: 37.3C NRR No No Yes
8 LPT T: 36.4C No No Yes No
9 LPT T: 37.3C NRR No No Yes
10 LPT T: 36.4 No No Yes No
11 ET T: 36.4C; RR: 66bpm No No Yes Yes
12 ET T: 36.3C No No Yes No
13 ET T: 36.4C No No Yes No
14 ET T: 36.2C No No Yes No
15 ET HR 155bpm NRR No No Yes
16 ET BGL 2.9mmol/L NRR No Yes No
17 ET T: 36.4C No No Yes No
18 ET HR: 160bpm NRR No No Yes
19 ET T: 39.9C [RR: 42bpm, HR:132bpm] No Yes Yes Yes
20 ET HR: 105bpm No No Yes No
21 ET RR: 62bpm No No Yes Yes
22 ET HR: 40bpm; RR: 120bpm No Yes Yes Yes
23 ET T: 36.4C No No Yes No
24 ET BGL: 2.6mmol/L NRR No Yes No
25 ET HR: 100bpm No No Yes No
26 ET HR: 166bpm No No Yes Yes
27 ET T: 36.3C No No Yes No
28 ET T: 37.3C NRR No No Yes
29 ET T: 36.4C No No Yes No
30 ET T: 36.3C No No Yes No
31 ET T: 36.3C No No Yes No
32 ET HR: 156bpm NRR No No Yes
33 ET BSL: 2.8mmol/L NRR No Yes No
34 PT T: 36.3C No No Yes No
Total n (%) 22 (64.7%) 2 (5.9) 27 (79.4) 12 (35.3)
410 LPT – Late preterm; ET – Early term; PT – Post term; GA -gestational age; BGL – Blood glucose level; RR 411
– respiration rate; bpm – breaths per minute; HR – Heart rate; bpm – beats per minute; T-Temperature; EOC 412
- escalation of care; NRR – normal reference range at the study hospital; SOT – Standard Observation Tool 413
414
415
Page 15 of 22
Accep
ted
Man
uscr
ipt
15
Discussion 416
The objective of this study was to examine whether observations from the Standard Observation Tool, 417
applied to three neonatal Early Warning Tools, would hypothetically trigger an escalation of care more 418
frequently than escalation of care using a Standard Observation Tool. Overall, for cases, the three neonatal 419
Early Warning Tools used in this study did not trigger an escalation of care more frequently than the 420
Standard Observation Tool. Although by gestational age group, one tool, the New South Wales Health Early 421
Warning Tool hypothetically triggered one additional case than the Standard Observation Tool. These 422
findings show that the design and measurement of observations in an early warning tool affects the 423
performance of the tool. 424
425
For the controls who had an abnormal observation and for whom the Standard Observation Tool escalated 426
care, none of the Early Warning Tools identified all of these abnormal observations. Nevertheless, for those 427
controls who had abnormal observations and the Standard Observation Tool did not escalate care, the New 428
South Wales Health Early Warning Tool hypothetically triggered an escalation of care considerably more 429
frequently than either of the other two tools (Table 5). However, the majority of the observations triggering 430
an escalation of care fell within the zone directing actions to an increase in frequency of observations and 431
allowing the clinician to make judgement as to whether there is a trend suggesting deterioration and the need 432
for a clinical review. Similarly, the majority of observations on the Standard Observation Tool may have 433
been assessed as part of the overall picture of the neonate and in the clinician’s opinion, not in isolation a 434
sign of deterioration. For example, in an instance of a high heart rate, ‘crying’ was noted on the Standard 435
Observation Tool for that observation and the subsequent observation was within normal range. 436
Additionally, for this study, we deemed that an escalation of care did not occur if observations were not 437
repeated within 30 minutes. However, for most abnormal observations, the frequency of observations were 438
increased, often being repeated within one to three hours. This may reflect the reality of current workloads in 439
busy maternity wards. Arguably, tools that are too responsive can potentially have a negative impact by 440
increasing workloads further of both nursing and medical personal when intervention was not required 441
(Cuthbertson & Smith, 2007). 442
443
Design of Early Warning Tools 444
The three Early Warning Tools tested in this study demonstrated mixed results with cases and controls, 445
despite being similar in a number of design features such as incorporating contrasting colours to indicate 446
worsening abnormal observations and providing specific action prompts once the escalation criterion was 447
met. Notably each tool has implemented varying colour combinations to draw attention to worsening 448
observations. However, traditionally, the green (stable), yellow (caution) and red (danger) combination is 449
used in many instances in the health care industry for tracking progress (Parker, n.d), in related health 450
projects and even the ‘traffic light’ system (New South Wales Government, 2015), the rationale for the 451
choice of other colour combinations in these tools are unknown. The results of this study suggest there are 452
three key differences between the tools: the scoring system used, measurement of additional observations 453
Page 16 of 22
Accep
ted
Man
uscr
ipt
16
(blood glucose levels and clinical observations), and differing physiological parameters. These differences 454
may have attributed to the performance of each of the tools. 455
456
Scoring system 457
The tools that hypothetically triggered an escalation of care more frequently in response to observations that 458
entered an abnormal zone, used a single parameter colour-coded system. The Australian Capital Territory 459
Newborn Early Warning Score had a requirement of an aggregate score. This reduced the number of cases 460
being triggered because even though the observations entered the abnormal coloured zone, the escalation 461
criterion is such that further escalation of care was not required. These results are in keeping with literature 462
that suggests that this method experiences a reduced level of responsiveness as a number of observations 463
need to be in abnormal zones prior to an action being triggered (Australian Commission on Safety and 464
Quality in Health Care, 2012). This has implications for clinical practice because, for example, in the early 465
stages of shock, the neonate can maintain a degree of compensation, thereby not displaying obvious, if any, 466
derangement of vital signs (Polin, Fox, & Abman, 2011; Sinniah, Subramaniam, & Soe-Hsiao, 2013; 467
Buonocore, Bracci, & Weindling, 2012). Perhaps, by the time multiple observations fall within the abnormal 468
zone triggering an escalation of care a serious adverse event may have already taken place. 469
Physiological reference ranges 470
There are significant physiological and neurological differences between the gestational age groups (Engle, 471
Tomashek & Wallman, 2007). It could be argued that what may be appropriate reference ranges for a post 472
term neonate, may in fact be indicative of compromise in the late preterm neonate and setting parameters 473
with a broad range may not be sufficient nor effective in capturing deterioration in neonates as a collective 474
cohort (Van Kuiken & Huth, 2013). There was some evidence of this in this study due to the variance in 475
parameter ranges, thereby no escalation of care for cases and an over-escalation for controls within the 476
respective gestational age groups. This highlights the need for standardised physiological parameters for 477
gestational age groups in the neonatal period (Mattson & Smith, 2011; Verklan & Walden, 2015; Takayama, 478
Wang, Uyemoto, Newman, Pantell, 2000) and importantly this may improve the sensitivity and specificity of 479
Early Warning Tools. Currently a one-size-fits all approach may not be suitable for newborns in the 480
maternity setting as reflected by the individual performance of the tools tested in this study, and in this 481
population where there were variations in the set physiological ranges such as heart rate, temperature, and 482
blood glucose levels. Previous studies have suggested that the cut off for parameters is based on clinical 483
intuition and/or historical data rather than on rigorously gathered data (Cuthbertson & Smith, 2007). The 484
effect of gestational age on physiologic parameters, for example, variations in heart rate between gestational 485
age groups has been documented (Fyfe, Yiallourou & Horne, 2012; Van Kuiken & Huth, 2013). 486
487
Heart rate 488
The Autonomic Nervous System (sympathetic and parasympathetic systems) controls cardiac functions such 489
as heart rate, heart rate variability and blood pressure. Lack of control due to immaturity is demonstrated in 490
Page 17 of 22
Accep
ted
Man
uscr
ipt
17
the late preterm neonate by their exhibiting a higher resting heart rate (Fyfe, Yiallourou & Horne, 2012). On 491
the other hand, post term neonates can exhibit a lower base line heart rate (Fyfe, Yiallourou & Horne, 2012). 492
This variability precludes one set of heart rate parameters across gestational ages and hence, it is necessary to 493
standardise heart rate, by gestational age group within the neonatal population (Van Kuiken & Huth, 2013). 494
495
Temperature 496
Newborns do not have the necessary thermoregulatory mechanisms to maintain body temperature and are at 497
risk of temperature instability due to a larger body surface area to weight ratio, reduced subcutaneous fat 498
stores, and an immature sympathetic nervous system which inhibits the neonate from initiating behaviours to 499
rectify being cold, such as ‘shivering’ (Brown & Landers, 2011). The United Kingdom Newborn Early 500
Warning Chart included a normal temperature range dissimilar to the other Early Warning Tool tested and 501
did not trigger an action for the case neonates who had lower temperatures. Moreover, it triggered an action 502
for controls who exhibited temperatures over 37.2ºC, which in clinical practice at the study hospital, is 503
regarded within normal temperature range, not indicative of sepsis and as such, requires no clinical 504
intervention. Therefore, in order for a tool to be effective, standardised parameters should be identified to 505
ensure deviating vital signs of neonates are captured (Cuthbertson, & Smith, 2007; Van Kuiken & Huth, 506
2013). If physiological reference ranges are set too conservatively, this would result in the tool being overly 507
responsive. A consequence of this would be increased frequency of referral for review, increased clinical 508
workload and potentially delayed discharge; all due to the neonate waiting for a clinical review. Equally, if 509
the ranges were set liberally, as demonstrated by The United Kingdom Newborn Early Warning Chart for 510
temperature, this may result in the tool being less responsive resulting in cases being missed. 511
512
Blood Glucose Levels 513
The only Early Warning Tool tested in this study that incorporated blood glucose level monitoring was the 514
New South Wales Early Warning Tool. To date, as with other physiological parameters, there is no defined 515
range for blood glucose levels, even though traditionally there appears to be agreement in the literature that 516
suggests intervention should begin when the blood glucose level is at or below 2.6mmol/L (Tin, 2014). Our 517
results suggest that by incorporating the blood glucose level, the New South Wales tool identified more cases 518
compared to the other Early Warning Tools. On the other hand, this tool had a normal blood glucose level 519
range between ≥3.0 and ≤10mmol/L. This conservative reference range triggered an escalation of care for a 520
number of controls exhibiting blood glucose levels ≥2.8mmol/L, which is considered normal at the study 521
hospital. 522
523
Behavioural observations 524
This study identified a number of observations in the case group that were not physiologically deranged vital 525
signs but clinical observations such as vomiting (with or without mucous), sleepiness and not feeding. 526
Moreover, these clinical observations were not included in the six physiological observations recommended 527
by the Australian Commission on Quality Safety Health Care yet triggered an escalation of care due to the 528
Page 18 of 22
Accep
ted
Man
uscr
ipt
18
concern of the clinician. Emerging research from the United Kingdom suggests that the gut instinct or 529
clinician intuition or concern regarding the patient is almost as effective as assessment of vital signs alone in 530
identifying the need for admission to the acute health setting (D. Roland, personal communication, 531
December 18, 2014). Therefore, it is vital that the design of an Early Warning Tool incorporates the ability 532
for the clinician to notate and escalate their concerns when required. 533
534
Clinical observable behaviours 535
The Standard Observation Tool incorporates a free text space that allows documentation of changes or 536
concerns with the newborns behaviour. For example, an irregular heartbeat, urine output, bowel motions, 537
vomiting, distended abdomen, the degree of being unsettled (inconsolable), grimacing (pain), and/or 538
sleepiness. The only early warning tool to incorporate elements of behaviour was the United Kingdom Tool, 539
which allowed documentation of behaviours and incorporated escalation processes depending on the 540
coloured zone the observation entered. 541
542
Strengths and limitations 543
A strength of this study is that a Neonatal Early Warning Tool designed in the United Kingdom and two 544
designed by different health authorities in Australia were compared to a Standard Observation Tool, which 545
has traditionally been used in the maternity ward for all neonates post birth at the study hospital. We 546
compared tools based on single parameter colour coded track and trigger system and an aggregate score 547
system and to the best of our knowledge, this is the first study of its kind in neonates in maternity settings. 548
549
While the data applied to the Early Warning Tools was from a retrospective data source, the data itself was 550
captured at the time of being documented. Even though both retrospective and prospective data may suffer 551
biases (Pannucci & Wilkins, 2010), to minimise bias due to seasonal variations and staff changes we 552
sampled all late preterm and post term neonates and a random selection of early term neonates, deemed well 553
who deteriorated in the maternity setting over a 12-month period. However, a limitation of this study and a 554
weakness of studies relying on retrospective data is that it can be incomplete and the validity of the data is 555
difficult to verify. Therefore, we were unable to verify several sets of observations that we surmise were 556
recorded in the wrong column. Due to a lack of documentation and the passing of time nor could we verify if 557
an escalation of care did or did not take place for both cases and controls with abnormal observations. 558
559
Conclusion 560
This study compared three Early Warning Tools designed for use for neonates cared for in the maternity 561
setting. Although the concept of an early warning tool is viewed as a positive step in the safe care and 562
management of neonates, the results of this study demonstrate that overall, the three Early Warning Tools 563
tested did not trigger an escalation of care more frequently than that of the Standard Observation Tool for 564
either cases or controls. Subgroup analysis by gestational age revealed differences between the tools in 565
frequency of triggering an escalation of care. However, the findings demonstrate that the design and 566
Page 19 of 22
Accep
ted
Man
uscr
ipt
19
measurement of observations in an early warning tool and the lack of standardised physiological reference 567
ranges affects the performance of the tools and it can be argued that one early warning tool ‘does not fit all’. 568
Consequently, several tools for specific gestational age groups of neonates may need to be developed if Early 569
Warning Tools are to be effective in detecting and triggering an escalation of care for early deterioration in 570
the newborn. Further research is needed into the normal physiological ranges of the gestational age groups 571
and the effectiveness of Early Warning Tools in these neonatal sub-populations. 572
573
574
575
Page 20 of 22
Accep
ted
Man
uscr
ipt
20
576
References 577 578 An Roinn Slainte Department of Heatlh., 2014. National Early Warning Score: National Clinical Guideline 579
No. 1. National Clinical Effectiveness Committee. Retrieved from: http://health.gov.ie/wp-580 content/uploads/2015/01/NEWSFull-ReportAugust2014.pdf [accessed 20.2.16]. 581
582 Australian Capital Territory (ACT) Health/Calvary Health., n.d. General Observation Chart Neonatal 0-<1-583
month post term. 584 585 Australian Commission on Safety and Quality in Health Care., 2010. National Consensus Statement: 586
Essential elements for recognising and responding to clinical deterioration. Sydney: ACSQHC. 587 Retrieved from: 588 http://www.safetyandquality.gov.au/wp-content/uploads/2012/01/NSQHS-Standards-Fact-Sheet-589 Standard-9.pdf [accessed 14.02.16]. 590 591
Australian Commission on Safety and Quality in Health Care., 2012. Recognising and responding to clinical 592 deterioration: use of observation charts to identify clinical deterioration. Sydney: ACSQHC. 593 Retrieved from: 594 http://www.safetyandquality.gov.au/wp-content/uploads/2012/02/UsingObservationCharts-595
20091.pdf [accessed 14.02.16]. 596 597 Babbie, E., 2010. The practice of social research (12th ed.). USA: Wadsworth Cengage Learning. 598 599 Brown, V., & Landers, S., 2011. Heat Balance. In Gardner, S., Carter, B., Enzman-Hines, M., & Hernandez, 600
J (Ed.). 2011. Merenstein and Gardner’s Handbook of Neonatal Intensive Care (pp113-133). USA: 601 Elsevier. 602
603 Buonocore, G., Bracci, R., & Weindling, M (Eds)., 2012. Neonatology: A practical approach to neonatal 604
management. Italy: Springer 605 606 Ching, E., Lavin, S., & Blair, H., 2011. Electrophysiology. In Watson, S., & Gorski, K (Ed). Invasive 607
Cardiology: A manual for cath lab personnel. (3rd Ed.)(pp.303-323). USA: Jones & Bartlett Learning 608 609 Clinical Excellence Commission., 2013. Between the Flags Program: Interim Evaluation Report: D12/17695. 610
Retrieved: http://www.cec.health.nsw.gov.au/__data/assets/pdf_file/0004/258151/btf-program-interim-611 evaluation-report-april-2013-v2.pdf [accessed 14.02.16]. 612
613 Cuthbertson, B., & Smith, G., 2007. A Warning on early-warning scores!. British Journal of Anaesthesia, 614
98(6), 704-6 615 616 Engle, W., Tomashek, K., & Wallman, K., 2007. “Late-Preterm” infants: A population at risk. Pediatrics. 617
120(6), 1390-1401. doi: 10.1542/peds.2007-2952 618 619 Fyfe, K., Yiallourou, S., & Horne, R., 2012. Cardiovascular Consequences of Preterm Birth in the First Year 620
of Life. In J. Morrison (Ed.). Preterm Birth - Mother and Child. (pp. 319-340). doi: 10.5772/26252. 621 Retrieved from: 622 http://www.intechopen.com/books/preterm-birth-mother-and-child/cardiovascular-consequences-of-623 preterm-birth-in-the-first-year-of-life [accessed 12.01.16]. 624
Page 21 of 22
Accep
ted
Man
uscr
ipt
21
625 626 Graves, B., & Haley, M., 2013. Newborn Transition. Journal of Midwifery and Womens Health. 58(6), 662-627
70. 628 629 Institute for Healthcare Improvement., 2016. Improvement Stories: Early Warning Systems: Scorecards That 630
Save Lives. Retrieved from: 631 http://www.ihi.org/resources/pages/improvementstories/earlywarningsystemsscorecardsthatsavelives.as632
px [accessed 25.02.16]. 633 634 635 King Edward Memorial Hospital., 2014. Clinical Guidelines: Routine care of the neonate in the ward. 636
Retrieved from: 637 http://kemh.health.wa.gov.au/development/manuals/O&G_guidelines/sectionb/10/b10.2.2.pdf 638 [accessed 28.10.15]. 639
640 Mattson, S., & Smith, J. (Eds)., 2011. Core Curriculum for Maternal-Newborn Nursing. (4th Edn). USA: 641
Saunders 642 643 McLellan, M., & Connor, J., 2013. The cardiac children’s hosptial early warning score (C-CHEWS). Journal 644
of Pediatric Nursing. 28(2), 171-8 645 646 National Patient Safety Agency., 2007. Recognising and responding appropriately to early signs of 647
deterioration in hospitalised patients. (Reference number 0683). Retrieved from: 648 http://www.nrls.npsa.nhs.uk/EasySiteWeb/getresource.axd?AssetID=60151 [accessed 15.02.16]. 649
650 New South Wales Department of Health., 2011. Clinical Practice Guidelines: Recognition of a sick baby or 651
child in the emergency department. Retrieved from: 652 http://www0.health.nsw.gov.au/policies/pd/2011/pdf/PD2011_038.pdf [accessed 28.10.15]. 653
654 New South Wales Department of Health., 2013. Recognition and Management of patients who are clinically 655
deteriorating: (PD2013_049). Retrieved from: 656 http://www0.health.nsw.gov.au/policies/pd/2013/pdf/PD2013_049.pdf [accessed 28.10.15]. 657
658 New South Wales Department of Health., 2012. Standard Newborn Observation – Special Care 659
Nursery/Postnatal ward Under 1 month (Corrected). 660 661 New South Wales Government., 2015. Traffic lights and road markings. Department of Roads and Maritime 662
Services. Retrieved from: 663 http://www.rms.nsw.gov.au/roads/safety-rules/road-rules/traffic-lights.html [accessed 2.02.16]. 664
665 666
Olroyd, C., & Day, A., 2011. The use of pediatric early warning scores in the emergency department. 667 Journal of Emergency Nursing. 37(4), 374-6 668 doi: 10.1016/j.jen.2011.03.007 669
670 Paliwoda, M., & New, K., 2015. Early warning tools in maternity and neonatal settings: A literature 671
review. Journal of Neonatal, paediatric, and Child Health Nursing. 18(1), 19-23. 672 673
Page 22 of 22
Accep
ted
Man
uscr
ipt
22
Pannucci, C., & Wilkins, E., 2010. Identifying and avoiding bias in research. Plastic Reconstructive 674 Surgery. 126(2), 619-25 675
676 Parker, R., n.d. The Meaning of Colours. Retrieved from: 677
https://resources.oncourse.iu.edu/access/content/user/rreagan/Filemanager_Public_Files/meaningofc678 olors.htm [accessed 20.02.16]. 679
680 Polin., R., Fox. W., & Abman, S., 2011. Fetal and Neonatal Physiology. (4th Edn). USA: Elsevier 681 682 Queensland Government., 2012a. Patient safety: from learning to action. Retrieved from: 683
https://www.health.qld.gov.au/psu/reports/docs/lta5.pdf [accessed 28.10.15]. 684 685 Queensland Government., 2012b. 21002/Procedure: Assessment and observation of the neonate. Queensland 686
Health, Royal Brisbane and Women's Hospital. 687 688 Queensland Government. (2012c). MR A 12890: Vigilance and Baby management chart. Queensland Health, Royal 689
Brisbane and Women's Hospital. 690 691 Roland, D., Madar, J., & Connolly., 2010. The newborn early warning (NEW) system: development of an at-692
risk infant intervention system. Infant. 6(4), 116-20 693 694 Satar, M., & Ozlu, F., 2012. Neonatal sepsis: a continuing disease burden. Turkish Journal of Pediatrics. 695
54(5), 449-57. 696 697 Sinniah, D., Subramaniam, T., & Soe-Hsiao, M., 2013. Shock in the neonate. International e-Journal of 698
Science, Medicine & Education. 7(2). 17-28 699 700 Song, J., Chung, K., 2010. Observational studies: cohort and case-control studies. Plastic and Reconstructive 701
Surgery. 126(6), 2234-2242. doi: 10.1097/PRS.ob013e3181144abc 702 703 Takayama, J., Wang, T., Uyemoto, J., Newman, T., Pantell, R., 2000. Body temperature of newborns: What 704
is normal? Clinical Pediatrics. 39(9). 503-510). 705 706 Tin, W., 2014. Defining neonatal hypoglycaemia: A continuing debate. Seminars in Fetal and Neonatal 707
Medicine. 19(1). 27-32. doi:10.1016/j.siny.2013.09.003 708 709 710 Van Kuiken, D., & Huth, M., 2013. What is ‘normal?’ Evaluating vital signs. Pediatric Nursing. 39(5). 711 712 Verklan, M., & Walden, M., 2015. Core Curriculum for Neonatal Intensive Care Nursing. USA: Elsevier 713 714 Ygberg, S & Nilsson, A., 2012. The developing immune system – from foetus to toddler. ACTA 715
Paediatricia. 101 (2). 120-127 716
