- Email: [email protected]
Validation of the National Early Warning Score in the prehospital setting
G Model
ARTICLE IN PRESS
RESUS 6256 1–5
Resuscitation xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Rapid Response Systems
1
Validation of the National Early Warning Score in the prehospital setting夽
2
3
4
Q1
a
5 6
Daniel J. Silcock a,∗ , Alasdair R. Corfield a , Paul A. Gowens b , Kevin D. Rooney a,c
Q2
7
b c
Royal Alexandra Hospital, Paisley, UK Scottish Ambulance Service, UK Institute for Care and Practice Improvement, University of the West of Scotland, UK
8
9 21
a r t i c l e
i n f o
a b s t r a c t
10 11 12 13 14 15
Article history: Received 6 February 2014 Received in revised form 26 November 2014 Accepted 8 December 2014
16
20
Keywords: Prehospital Risk stratification National Early Warning Score
22
1. Introduction
17 18 19
23Q3 24 25 26 27 28 29 30 31 32
Background: Early intervention and response to deranged physiological parameters in the critically ill patient improves outcomes. A National Early Warning Score (NEWS) based on physiological observations has been developed for use throughout the National Health Service (NHS) in the UK. Although a good predictor of mortality and deterioration in inpatients, its performance in the prehospital setting is largely untested. This study aimed to assess the validity of the NEWS in unselected prehospital patients. Methods: All clinical observations taken by emergency ambulance crews transporting patients to a single hospital were collated along with information relating to hospital outcome over a two month period. The performance of the NEWS in identifying the endpoints of 48 h and 30 day mortality, intensive care unit (ICU) admission, and a combined endpoint of 48 h mortality or ICU admission was analysed. Results: 1684 patients were analysed. All three of the primary endpoints and the combined endpoint were associated with higher NEWS scores (p < 0.01 for each). The medium-risk NEWS group was associated with a statistically significant increase in ICU admission (RR = 2.466, 95% CI 1.0–6.09), but not in-hospital mortality relative to the low risk group. The high risk NEWS group had significant increases in 48 h mortality (RR 35.32 [10.08–123.7]), 30 day mortality (RR 6.7 [3.79–11.88]), and ICU admission (5.43 [2.29–12.89]). Similar results were noted when trauma and non-trauma patients were analysed separately. Conclusions: Elevated NEWS among unselected prehospital patients is associated with a higher incidence of adverse outcomes. Calculation of prehospital NEWS may facilitate earlier recognition of deteriorating patients, early involvement of senior Emergency Department staff and appropriate critical care. © 2015 Published by Elsevier Ireland Ltd.
Early intervention and response to deranged physiological parameters improves survival outcomes in both medical and surgical patients.1–3 To this end a number of early warning systems have been developed in recent years4 and have been shown to be good predictors of mortality and deterioration.5 Recent work by the Royal College of Physicians has led to the development of a National Early Warning Score (NEWS)6 which is currently implemented by the National Health Service (NHS) in acute hospitals across the UK and has been successful in identifying patients at risk of deterioration or death.7 More recently there has been adoption
夽 A Spanish translated version of the summary of this article appears as Appendix in the final online version at http://dx.doi.org/10.1016/j.resuscitation.2014.12.029. ∗ Corresponding author. E-mail address: [email protected] (D.J. Silcock).
of early warning scores in some emergency departments8–10 although this has not been without controversy.11 The use of NEWS in the prehospital setting also remains controversial, partly due to lack of evidence.12,13 As the development of this score involved analysis of clinical observations in hospital inpatients, in whom a course of treatment had already been started, uncertainty exists as to the applicability of the NEWS to other settings, particularly the emergency department and prehospital setting, where scores would be derived prior to the institution of any treatment. In this situation the score may contribute to a decision to transfer a patient to hospital14 or may be used as a triage aid, both of which roles differ slightly from the track and trigger (of a clinical review) role for which NEWS was originally intended. The NEWS was based on the earlier ViEWS15 (VitalPAC Early Warning Score) developed in Portsmouth and stratifies patients into risk categories based on observed heart rate, respiratory rate, systolic blood pressure, arterial oxygen saturation, temperature, and conscious level; plus an additional weighting if the patient is
http://dx.doi.org/10.1016/j.resuscitation.2014.12.029 0300-9572/© 2015 Published by Elsevier Ireland Ltd.
Please cite this article in press as: Silcock DJ, et al. Validation of the National Early Warning Score in the prehospital setting. Resuscitation (2015), http://dx.doi.org/10.1016/j.resuscitation.2014.12.029
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
G Model
ARTICLE IN PRESS
RESUS 6256 1–5
D.J. Silcock et al. / Resuscitation xxx (2015) xxx–xxx
2
Table 1 Breakdown of number of presentation by AMPDS category. Traumatic haemorrhages were included under the injury category. Deliberate self-harm is classified under overdose or injury as appropriate. The presentation in eye problem was an ocular foreign body. n
26 10 01 06 12 23 32 31 28 21 05 18 19 25 13 02 11 08
299 182 155 107 81 81 79 73 57 46 38 28 25 25 21 7 5 3
Sick person Chest pain Abdominal pain Breathing difficulty Seizures Overdose Medical nature unknown Subject unconscious Stroke Haemorrhage Back pain Headache Heart problem Psychiatric problem Diabetic problems Allergic reaction Choking Hazardous exposure
Trauma presentations 17 04 30 29 08 27 07 16
n Falls Assault Injury Traffic collision Hazardous exposure Penetrating trauma Burn subject Eye problem
277 40 33 18 3 2 1 1
59
being given oxygen therapy. Patients are then risk stratified based on the resulting aggregate score into low, medium, and high risk groups. Patients with a low aggregate score but who score in the highest category for any single observation are classified as at least medium risk (see Supp. Figs. 1 and 2). This study, based in a large district general hospital in Paisley, on the western edge of the Greater Glasgow metropolitan area, Scotland, aimed to evaluate the performance of the NEWS in identifying patients at risk of death or deterioration in the prehospital setting.
60
2. Methods
51 52 53 54 55 56 57 58
61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82
Data protection approval to analyse and cross-reference patient-identifiable information was obtained from the Caldicott Guardians of both organisations involved (NHS Greater Glasgow and Clyde and Scottish Ambulance Service) prior to commencement of the study. Details of all emergency ambulance crews dispatched with an intention to transfer to the Royal Alexandra Hospital (RAH) were obtained from the Scottish Ambulance Service data warehouse, along with details of demographics, initial patient presenting complaint, and clinical observations obtained from the ambulances’ electronic patient record forms (ePRF). These were matched to a list of patients presenting to the Emergency Department of the RAH to obtain details related to the patients’ hospital admissions. Patients aged less than 16 years and patients known to be pregnant were excluded, along with patients transferred from other hospitals (as these were by definition not from the prehospital setting). NEWS values for each patient encounter were calculated retrospectively from the clinical parameters obtained. This was a retrospective cohort study over a 2-month consecutive period between October 1st and November 30th, 2012 using a convenience sample of consecutive patients. Ambulance diversion protocols were in place to transfer patients with ST-elevation MI direct to the local primary PCI centre, and pregnant women in
0.9 0.8 0.7
Sensivity
Non-trauma presentations
Receiver Operang Characterisc curves for predicon of mortality 1
0.6 0.5
30 day
0.4
7 day
0.3 0.2
48 hour
0.1
24 hour
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1-specificity Fig. 1. Receiver operating characteristic curves for various mortality outcomes.
labour were diverted to the nearby maternity hospital. All other patients, including following major trauma, were transported to the RAH. From the identified records, information regarding length of stay, discharge status, and admission to intensive care and high dependency units was obtained from hospital computer systems. A number of patient outcomes were identified for study – these being: • • • •
Survival to discharge or 30 days. Death within 48 h of hospital admission. ICU admission within 48 h of hospital admission. A composite adverse outcome of death or ICU admission within 48 h.
Receiver operating characteristic curves plotting sensitivity against (1-specificity) were plotted for the outcomes above, and the area under the receiver operating characteristic curve (AUROC) calculated for each. Comparison was made regarding outcomes between the risk strata identified in the original NEWS specification by means of the chi-squared test. All statistical calculation was carried out in StatsDirect 2.7.8 for Windows. 3. Results 11,052 sets of clinical observations were obtained from 6028 unique patients. After exclusions, 1684 complete patient encounters were identified for study (see Supp. Fig. 3). All patients were transported by emergency ambulances staffed either by two paramedics or one paramedic and one emergency ambulance technician. Table 1 shows a description of the patient population studied. 3.1. Discriminative performance of the NEWS Area under the ROC was calculated for a number of mortality outcomes and these are illustrated in Fig. 1. Areas under the curve for 30, 14, 7, 2 and 1-day mortality were 0.740 [95% CI: 0.661–0.819], 0.788 [0.714–0.863], 0.796 [0.704–0.889], 0.871 [0.75–0.98], and 0.855 [0.69–1] respectively. Analysis with the Mann–Whitney test confirmed a difference in median scores between each pair and all were all of statistical significance with p < 0.0001. Similar results were found when looking at ICU admission within the first 48 h of admission, the AUROC for ICU admission being 0.774 (95% CI: 0.657–0.890) and that for the combined endpoint of ICU admission or death within 48 h being 0.815 (0.730–0.990) – see Fig. 2. For the combined endpoint of death in
Please cite this article in press as: Silcock DJ, et al. Validation of the National Early Warning Score in the prehospital setting. Resuscitation (2015), http://dx.doi.org/10.1016/j.resuscitation.2014.12.029
83 84 85 86 87 88 89 90
91 92 93 94 95
96 97 98 99 100 101 102
103
104 105 106 107 108 109 110
111
112 113 114 115 116 117 118 119 120 121 122 123
G Model
ARTICLE IN PRESS
RESUS 6256 1–5
D.J. Silcock et al. / Resuscitation xxx (2015) xxx–xxx
Ulity of NEWS as a screening tool 100
0.9
90
0.8
80
0.7
70
Affected %
Sensivity
48 hour mortality and ICU admission 1
0.6 0.5 48 hour mortality
0.4 0.3
ICU admission
0.2 Combined outcome
0.1 0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1-specificity Fig. 2. ROC curves demonstrating prediction of 48hr mortality or ICU admission within 48 h of presentation, both individually and as a combined endpoint.
125
the Emergency Department or admission directly to ICU from the ED, the AUROC was 0.889 (0.823–0.957).
126
3.2. Outcomes in the various risk strata identified by NEWS
Outcomes for the patients in the various NEWS categories are given in Table 2. When the 2 test was used to compare out128 comes in the groups, there was no significant difference in 30-day 129 or 48-h mortality in the medium risk category compared to the 130 low risk category, although there was a significant increase in ICU 131 admission. The high risk group demonstrated statistically signif132 icant differences in respect of 30-day mortality, 48-h mortality, 133 and ICU admission in the first 48 h, with risk ratios of 6.7, 35 and 134 5.4, respectively. These findings were similar in trauma and non135 Q4 trauma patients (Table 3). 136 127
3.3. Efficiency of the NEWS
144
Fig. 3 shows the likely impact in terms of patients that would be triaged as high risk at a given NEWS score cut-off by plotting the total proportion of patients above a given NEWS cutoff against the proportion of patients in the various adverse outcome groups above a given NEWS cutoff. From this it can be seen that at a NEWS cutoff of 7, the 10% of patients triaged in account for 70% of those that die within 48 h.
145
4. Discussion
138 139 140 141 142 143
146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164
30 day mortality
50 40
48hr mortality
30
ICU admission
20
ICU/48hr mortality
10 Death/ICU from ED
0
137
60
0
0
124
3
Triage is a crucial part of any unplanned care system, and has been developed significantly since its inception during the Napoleonic Wars. Algorithms such as the Manchester Triage System16 provide objective criteria for the allocation of patients to care areas, and can be sensitive enough to detect the signs of critical illness at the point of entry to the emergency department although they may still miss patients with the possibility of deterioration while still in the Emergency Department.17 Early warning systems have been increasingly employed in the inpatient setting throughout the world in recent years and provide a means of identification of potential deterioration. However these scores are only effective in reducing adverse outcomes if an appropriate clinical response involving those experienced in critical care can be available to respond to the changing clinical situation.18 Scores developed for one patient population or healthcare system may not be applicable to other populations or healthcare systems,19 and even within a given healthcare system there may be individuals for whom the standard scoring points and triggers are inappropriate. Modifications may need to be made to the trigger points in
0
10
20
30
40
50
60
70
80
90
100
Included % Fig. 3. Total patients above a given NEWS score (x) vs. the proportion of patients in a given outcome group with that score or above (y). NEWS = 0 is at the top-right. For example NEWS = 7 includes 10.4% total patients but includes 71% of the 48 h mortality group.
certain other groups of patients with chronically deranged vital observations.20,21 Early warning scores are also reliant on clinical observations being performed at an appropriate frequency to be useful as a predictor of deterioration, and this may be limited by other workload.22 There is some evidence, however, that the introduction of clinical observations charts printed with early warning score calculations increases the incidence of respiratory rate being recorded, itself a valuable independent marker of deterioration.23,24 Compared to the inpatient patient population, the adoption of early warning scoring systems has been less enthusiastic in the emergency and prehospital settings11 partly as there is a relative lack of robust evidence supporting their validity in this patient cohort. However usage of aggregate scoring systems such as NEWS is increasing and is developing a role in this respect as a tool to predict the need for hospital admission,25 as well as likely outcomes, particularly in sepsis.26 This has been helped greatly by the development of point-of-care and rapidly available blood analysis, which improves the prognostic accuracy of the score.27,28 No early warning score can replace clinical assessment, and there are many situations where decisions about clinical management should be based on other criteria: the presence of ST-elevation myocardial infarction, major haemorrhage, or multiple trauma being only a few examples. In the absence of such presentations, signs of potentially severe illness or occult injury may occasionally be missed. In this scenario, a physiologically based score may highlight individual patients in need of more urgent care,29 a particular example being patients with suspected sepsis, where timely administration of antibiotics can make a significant difference in outcome. This may form part of a pre-alert protocol30 or indicate specific pre-hospital treatments. Similarly at the other end of the severity spectrum, a low score in combination with an appropriate clinical assessment may safely allow treatment at a location other than an Emergency Department.31 This study does demonstrate that there is a role for the NEWS in the prehospital setting, and its ability to identify high-risk patients should encourage its use at an early stage. In particular, across the whole range of patient presentations, patients with a NEWS of 7 or greater had 11% chance of death or ICU admission within 48 h. Although still low risk, there is a statistically significant rise in ICU admission in the medium risk group, and scoring shows good sensitivity for adverse outcomes. While the use of NEWS should not replace clinical judgement and a normal score does not preclude serious pathology, the unexpected finding of a high score should prompt increased prioritisation.
Please cite this article in press as: Silcock DJ, et al. Validation of the National Early Warning Score in the prehospital setting. Resuscitation (2015), http://dx.doi.org/10.1016/j.resuscitation.2014.12.029
165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209
G Model
ARTICLE IN PRESS
RESUS 6256 1–5
D.J. Silcock et al. / Resuscitation xxx (2015) xxx–xxx
4 Table 2 Risks of each outcome at each NEWS category. N
Medium risk
High risk
Non-trauma 214
Trauma 37
All 251
Non-trauma 135
Trauma 11
All 146
2 0.17 8.386 2.79–25.21 <0.001
19 0.13 6.710 3.79–11.88 <0.001
30 day mortality
n AR RR 95% p
6 0.029 1.443 0.58–3.57 0.425
0 0 0 0.11–6.33 0.84
6 0.02 1.242 0.52–3.00 0.63
17 0.13 6.482 3.46–12.16 <0.001
ICU admission
n AR RR 95% p
7 0.03 2.466 1.0–6.09 0.045
0 0 0.71
7 0.03 2.59 1.06–6.35 0.03
7 0.05 4.900 1.90–12.66 0.002
1 0 – <0.001
8 0.06 5.433 2.29–12.89 <0.001
48 h mortality
n AR RR 95% p
1 0.17 1.523 0.159–14.575 0.72
0 0 – – 0.61
1 0.004 1.726 0.18–16.52 0.63
10 0.07 24.148 6.73–86.65 <0.001
2 0.18 – – <0.001
12 0.09 35.315 10.08–123.7 <0.001
AR: absolute risk, RR: relative risk, 95%: 95% confidence interval of relative risk. There were no ICU admissions or deaths within 48 h in the low risk trauma group therefore it was not possible to calculate relative risks.
Table 3 Sensitivity and specificity for various outcomes at the cutoff scores suggested in the NEWS specification.
30 day mortality
n
Total ≥1 1126
Medium risk 397
High risk 146
Sensitivity
0.976 (0.874–0.999) 0.134 (0.115–0.154) 0.036 0.994
0.547 (0.386–0.702) 0.743 (0.718–0.767) 0.067 0.980
0.404 (0.256– 0.567) 0.906 (0.888–0.921) 0.125 0.978
1 (0.768–1)* 0.132 (0.114–0.152) 0.012 1
0.785 (0.492–0.953) 0.739 (0.714–0.763) 0.031 0.996
0.714 (0.418–0.916) 0.902 (0.884–0.918) 0.074 0.996
0.941 (0.713–0.998) 0.132 (0.113–0.151) 0.014 0.994
0.764 (0.501–0.931) 0.740 (0.715–0.764) 0.038 0.995
0.411 (0.184–0.670) 0.900 (0.882–0.915) 0.051 0.991
0.966 (0.827–0.999) 0.132 (0.115–0.153) 0.025 0.994
0.767 (0.577–0.900) 0.745 (0.720–0.769) 0.067 0.992
0.533 (0.343–0.716) 0.906 (0.888–0.921) 0.11 0.980
Specificity PPV NPV 48 h mortality
Sensitivity Specificity PPV NPV
Admission to ICU
Sensitivity Specificity PPV NPV
Combined ICU/48hr mortality
Sensitivity Specificity PPV NPV
Bracketed numbers indicate 95% confidence interval except as marked * which indicates 97.5% single sided CI. PPV: positive predictive value; NPV: negative predictive value.
210
211 212 213 214 215 216 217 218 219 220 221 222
5. Limitations This study does have some limitations. We have not adjusted for age or sex differences between the low, medium and high risk NEWS groups, and this may account for some of the difference in mortality. However, the absence of age or gender adjustments is a feature of the NEWS that our study aimed to validate. As one of the end-points was survival to discharge and only in-hospital death was considered, the mortality rate, particularly the 30day mortality rate, may be underestimated. Similarly, due to the method of data collection, it was not possible to conclusively identify all patients who were re-admitted following discharge, although there were no deaths among those that could be identified as being repeat attenders. The overall mortality rate was low, as was the
proportion of people with high NEWS scores and the total numbers in these categories was also low, particularly among trauma patients.
6. Conclusion Elevated NEWS among unselected prehospital patients is associated with increased levels of adverse outcomes. Calculation of an early warning score prior to transfer to hospital is straightforward and may be a useful triage tool with potential to facilitate earlier recognition of at-risk or deteriorating patients, possibly allowing earlier involvement of appropriate ED and critical care staff.
Please cite this article in press as: Silcock DJ, et al. Validation of the National Early Warning Score in the prehospital setting. Resuscitation (2015), http://dx.doi.org/10.1016/j.resuscitation.2014.12.029
223 224 225
226
227 228 229 230 231 232 233
G Model RESUS 6256 1–5
ARTICLE IN PRESS D.J. Silcock et al. / Resuscitation xxx (2015) xxx–xxx
234
235
236 Q5
237
238
239 240 241
242 243 Q6 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279
Conflict of interest statement No conflicts of interest to declare. Uncited references 32,33. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.resuscitation. 2014.12.029. References 1. Rivers E, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001;345:1368–77. 2. Dellinger RP, et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock. Intensive Care Med 2012;39:165–228, http://dx.doi.org/10.1007/s00134-012-2769-8. 3. Smith T, et al. Accuracy of an expanded early warning score for patients in general and trauma surgery wards. Br J Surg 2012;99:192–7, http://dx.doi.org/10.1002/bjs.7777. ISSN: 0007-1323. 4. Morgan RJM, Williams F, Wright MM. An early warning score for the early detection of patients with impending illness. Clin Intensive Care 1997;8:100. 5. Goldhill DR, McNarry AF. Physiological abnormalities in early warning scores are related to mortality in adult inpatients. Br J Anaesth 2004;92:882–4, http://dx.doi.org/10.1093/bja/aeh113. 6. Standardising the assessment of acute-illness severity in the NHS. Report of a working party. London: Royal College of Physicians; 2012. Gary B. Have we found the perfect early war7. Smith ning score? A view of ViEWS. Resuscitation 2013;84:707–8, http://dx.doi.org/10.1016/j.resuscitation.2013.04.001. 8. Groarke JD, et al. Use of an admission early warning score to predict patient morbidity and mortality and treatment success. Emerg Med J 2008;25:803–6, http://dx.doi.org/10.1136/emj.2007.051425. 9. Griffiths JR, Kidney EM. Current use of early warning scores UK emergency departments. Emerg Med J 2012;29:65–6, in http://dx.doi.org/10.1136/emermed-2011-200508. 10. Rees JE, Mann C. Use of the patient at risk scores in the emergency department: a preliminary study. Emerg Med J 2004;21:698–9, http://dx.doi.org/10.1136/emj.2003.006197. 11. Timothy J. An early warning? Universal risk scoring in emergency medicine. Roland, Damian and Coats. Emerg Med J 2011;28:263, http://dx.doi.org/10.1136/emj.2010.106104. 12. Roland D, Jahn H. Are early warning scores too early for paramedic practice? J Paramed Pract 2012;4:16. 13. Correspondence between National Ambulance Service Medical Directors Group and Royal College of Physicians; 2013 [unpublished]. 14. Scottish Ambulance. Service Clinical Strategy 2011–2015; 2010. 15. Prytherch DR, et al. ViEWS – towards a national early warning score for detecting adult inpatient deterioration. Resuscitation 2010;81:932–7, http://dx.doi.org/10.1016/j.resuscitation.2010.04.014.
5
16. Manchester Triage. Group Emergency Triage. Blackwell; 2006. 17. Cooke MW, Jinks S. Does the Manchester triage system detect the critically ill? J Accid Emerg Med 1999;16:179–81. 18. McNeill G, Bryden D. Do either early warning systems or emergency response teams improve hospital patient survival? A systematic review. Resuscitation 2013;84:1652–67, http://dx.doi.org/10.1016/j.resuscitation.2013.08.006. 19. Wheeler India, et al. Early warning scores generated in developed healthcare settings are not sufficient at predicting early mortality in Blantyre, Malawi: a prospective cohort study. 3. PLOS ONE 2013;8:e59830, http://dx.doi.org/10.1371/journal.pone.0059830. 20. Eccles Sinan R, et al. CREWS: improving specificity whilst mainsensitivity of the National Early Warning Score in taining patients with chronic hypoxaemia. Resuscitation 2013;85:109–11, http://dx.doi.org/10.1016/j.resuscitation.2013.08.277. 21. O’Driscoll R. Emergency oxygen use. Br Med J 2012;345:e6856. 22. Hands C, et al. Patterns in the recording of vital signs and early warning scores: compliance with a clinical escalation protocol. BMJ Qual Saf 2013;22:719–26, http://dx.doi.org/10.1136/bmjqs-2013-001954. 23. McBride Jackie, et al. Long-term effect of introducing an early warning score on respiratory rate charting on general wards. Resuscitation 2005;65:41–4, http://dx.doi.org/10.1016/j.resuscitation.2004.10.015. 24. Hammond NE, et al. The effect of implementing a modified early warning scoring (MEWS) system on the adequacy of vital sign documentation. Aust Crit Care 2013;26:18–22, http://dx.doi.org/10.1016/j.aucc.2012.05.001. 25. Burch VC, Tarr G, Morroni C. Modified early warning score predicts the need for hospital admission and inhospital mortality. Emerg Med J 2008;25:674–8, http://dx.doi.org/10.1136/emj.2007.057661. 26. Corfield AR, et al. Utility of a single early warning score in patients with sepsis in the emergency department. Emerg Med J 2013. 27. Mohammed MA, et al. Index blood tests and national early warning scores within 24 hours of emergency admission can predict the risk of in-hospital mortality: a model development and validation study. PLOS ONE 2013;8:e64340, http://dx.doi.org/10.1371/journal.pone.0064340. 28. Jarvis S, et al. Combining the National Early Warning Score with an early warning score based on common laboratory test results better discriminates patients at risk of hospital mortality. Rapid response systems and medical emergency teams. Projects: CHMI. Health informatics; 2013. 29. Fullerton JN, et al. Is the Modified Early Warning Score (MEWS) superior to clinician judgement in detecting critical illness in pre-hospital environment? Resuscitation 2012;83:557–62, the http://dx.doi.org/10.1016/j.resuscitation.2012.01.004. SM, Bloch M. An evaluation of a new prehospi30. Booth pre-alert guidance tool. Emerg Med J 2013;30:820–3, tal http://dx.doi.org/10.1136/emermed-2012-201545. 31. Challen K, Walter D. Physiological scoring: an aid to emergency medical services transport decisions? Prehosp Disaster Med 2010;25:320–3. 32. Smith GB, et al. The ability of the National Early Warning Score (NEWS) to discriminate patients at risk of early cardiac arrest, unanticipated intensive care unit admission, and death. Resuscitation 2013;84:465–70, http://dx.doi.org/10.1016/j.resuscitation.2012.12.016. 33. Jo S, et al. Comparison of the trauma and injury severity score and modified early warning score with rapid lactate level (the ViEWS-L score) in blunt trauma patients. Eur J Emerg Med 2013, http://dx.doi.org/10.1097/MEJ.0b013e32836192d6.
Please cite this article in press as: Silcock DJ, et al. Validation of the National Early Warning Score in the prehospital setting. Resuscitation (2015), http://dx.doi.org/10.1016/j.resuscitation.2014.12.029
280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333
ARTICLE IN PRESS
RESUS 6256 1–5
Resuscitation xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Rapid Response Systems
1
Validation of the National Early Warning Score in the prehospital setting夽
2
3
4
Q1
a
5 6
Daniel J. Silcock a,∗ , Alasdair R. Corfield a , Paul A. Gowens b , Kevin D. Rooney a,c
Q2
7
b c
Royal Alexandra Hospital, Paisley, UK Scottish Ambulance Service, UK Institute for Care and Practice Improvement, University of the West of Scotland, UK
8
9 21
a r t i c l e
i n f o
a b s t r a c t
10 11 12 13 14 15
Article history: Received 6 February 2014 Received in revised form 26 November 2014 Accepted 8 December 2014
16
20
Keywords: Prehospital Risk stratification National Early Warning Score
22
1. Introduction
17 18 19
23Q3 24 25 26 27 28 29 30 31 32
Background: Early intervention and response to deranged physiological parameters in the critically ill patient improves outcomes. A National Early Warning Score (NEWS) based on physiological observations has been developed for use throughout the National Health Service (NHS) in the UK. Although a good predictor of mortality and deterioration in inpatients, its performance in the prehospital setting is largely untested. This study aimed to assess the validity of the NEWS in unselected prehospital patients. Methods: All clinical observations taken by emergency ambulance crews transporting patients to a single hospital were collated along with information relating to hospital outcome over a two month period. The performance of the NEWS in identifying the endpoints of 48 h and 30 day mortality, intensive care unit (ICU) admission, and a combined endpoint of 48 h mortality or ICU admission was analysed. Results: 1684 patients were analysed. All three of the primary endpoints and the combined endpoint were associated with higher NEWS scores (p < 0.01 for each). The medium-risk NEWS group was associated with a statistically significant increase in ICU admission (RR = 2.466, 95% CI 1.0–6.09), but not in-hospital mortality relative to the low risk group. The high risk NEWS group had significant increases in 48 h mortality (RR 35.32 [10.08–123.7]), 30 day mortality (RR 6.7 [3.79–11.88]), and ICU admission (5.43 [2.29–12.89]). Similar results were noted when trauma and non-trauma patients were analysed separately. Conclusions: Elevated NEWS among unselected prehospital patients is associated with a higher incidence of adverse outcomes. Calculation of prehospital NEWS may facilitate earlier recognition of deteriorating patients, early involvement of senior Emergency Department staff and appropriate critical care. © 2015 Published by Elsevier Ireland Ltd.
Early intervention and response to deranged physiological parameters improves survival outcomes in both medical and surgical patients.1–3 To this end a number of early warning systems have been developed in recent years4 and have been shown to be good predictors of mortality and deterioration.5 Recent work by the Royal College of Physicians has led to the development of a National Early Warning Score (NEWS)6 which is currently implemented by the National Health Service (NHS) in acute hospitals across the UK and has been successful in identifying patients at risk of deterioration or death.7 More recently there has been adoption
夽 A Spanish translated version of the summary of this article appears as Appendix in the final online version at http://dx.doi.org/10.1016/j.resuscitation.2014.12.029. ∗ Corresponding author. E-mail address: [email protected] (D.J. Silcock).
of early warning scores in some emergency departments8–10 although this has not been without controversy.11 The use of NEWS in the prehospital setting also remains controversial, partly due to lack of evidence.12,13 As the development of this score involved analysis of clinical observations in hospital inpatients, in whom a course of treatment had already been started, uncertainty exists as to the applicability of the NEWS to other settings, particularly the emergency department and prehospital setting, where scores would be derived prior to the institution of any treatment. In this situation the score may contribute to a decision to transfer a patient to hospital14 or may be used as a triage aid, both of which roles differ slightly from the track and trigger (of a clinical review) role for which NEWS was originally intended. The NEWS was based on the earlier ViEWS15 (VitalPAC Early Warning Score) developed in Portsmouth and stratifies patients into risk categories based on observed heart rate, respiratory rate, systolic blood pressure, arterial oxygen saturation, temperature, and conscious level; plus an additional weighting if the patient is
http://dx.doi.org/10.1016/j.resuscitation.2014.12.029 0300-9572/© 2015 Published by Elsevier Ireland Ltd.
Please cite this article in press as: Silcock DJ, et al. Validation of the National Early Warning Score in the prehospital setting. Resuscitation (2015), http://dx.doi.org/10.1016/j.resuscitation.2014.12.029
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
G Model
ARTICLE IN PRESS
RESUS 6256 1–5
D.J. Silcock et al. / Resuscitation xxx (2015) xxx–xxx
2
Table 1 Breakdown of number of presentation by AMPDS category. Traumatic haemorrhages were included under the injury category. Deliberate self-harm is classified under overdose or injury as appropriate. The presentation in eye problem was an ocular foreign body. n
26 10 01 06 12 23 32 31 28 21 05 18 19 25 13 02 11 08
299 182 155 107 81 81 79 73 57 46 38 28 25 25 21 7 5 3
Sick person Chest pain Abdominal pain Breathing difficulty Seizures Overdose Medical nature unknown Subject unconscious Stroke Haemorrhage Back pain Headache Heart problem Psychiatric problem Diabetic problems Allergic reaction Choking Hazardous exposure
Trauma presentations 17 04 30 29 08 27 07 16
n Falls Assault Injury Traffic collision Hazardous exposure Penetrating trauma Burn subject Eye problem
277 40 33 18 3 2 1 1
59
being given oxygen therapy. Patients are then risk stratified based on the resulting aggregate score into low, medium, and high risk groups. Patients with a low aggregate score but who score in the highest category for any single observation are classified as at least medium risk (see Supp. Figs. 1 and 2). This study, based in a large district general hospital in Paisley, on the western edge of the Greater Glasgow metropolitan area, Scotland, aimed to evaluate the performance of the NEWS in identifying patients at risk of death or deterioration in the prehospital setting.
60
2. Methods
51 52 53 54 55 56 57 58
61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82
Data protection approval to analyse and cross-reference patient-identifiable information was obtained from the Caldicott Guardians of both organisations involved (NHS Greater Glasgow and Clyde and Scottish Ambulance Service) prior to commencement of the study. Details of all emergency ambulance crews dispatched with an intention to transfer to the Royal Alexandra Hospital (RAH) were obtained from the Scottish Ambulance Service data warehouse, along with details of demographics, initial patient presenting complaint, and clinical observations obtained from the ambulances’ electronic patient record forms (ePRF). These were matched to a list of patients presenting to the Emergency Department of the RAH to obtain details related to the patients’ hospital admissions. Patients aged less than 16 years and patients known to be pregnant were excluded, along with patients transferred from other hospitals (as these were by definition not from the prehospital setting). NEWS values for each patient encounter were calculated retrospectively from the clinical parameters obtained. This was a retrospective cohort study over a 2-month consecutive period between October 1st and November 30th, 2012 using a convenience sample of consecutive patients. Ambulance diversion protocols were in place to transfer patients with ST-elevation MI direct to the local primary PCI centre, and pregnant women in
0.9 0.8 0.7
Sensivity
Non-trauma presentations
Receiver Operang Characterisc curves for predicon of mortality 1
0.6 0.5
30 day
0.4
7 day
0.3 0.2
48 hour
0.1
24 hour
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1-specificity Fig. 1. Receiver operating characteristic curves for various mortality outcomes.
labour were diverted to the nearby maternity hospital. All other patients, including following major trauma, were transported to the RAH. From the identified records, information regarding length of stay, discharge status, and admission to intensive care and high dependency units was obtained from hospital computer systems. A number of patient outcomes were identified for study – these being: • • • •
Survival to discharge or 30 days. Death within 48 h of hospital admission. ICU admission within 48 h of hospital admission. A composite adverse outcome of death or ICU admission within 48 h.
Receiver operating characteristic curves plotting sensitivity against (1-specificity) were plotted for the outcomes above, and the area under the receiver operating characteristic curve (AUROC) calculated for each. Comparison was made regarding outcomes between the risk strata identified in the original NEWS specification by means of the chi-squared test. All statistical calculation was carried out in StatsDirect 2.7.8 for Windows. 3. Results 11,052 sets of clinical observations were obtained from 6028 unique patients. After exclusions, 1684 complete patient encounters were identified for study (see Supp. Fig. 3). All patients were transported by emergency ambulances staffed either by two paramedics or one paramedic and one emergency ambulance technician. Table 1 shows a description of the patient population studied. 3.1. Discriminative performance of the NEWS Area under the ROC was calculated for a number of mortality outcomes and these are illustrated in Fig. 1. Areas under the curve for 30, 14, 7, 2 and 1-day mortality were 0.740 [95% CI: 0.661–0.819], 0.788 [0.714–0.863], 0.796 [0.704–0.889], 0.871 [0.75–0.98], and 0.855 [0.69–1] respectively. Analysis with the Mann–Whitney test confirmed a difference in median scores between each pair and all were all of statistical significance with p < 0.0001. Similar results were found when looking at ICU admission within the first 48 h of admission, the AUROC for ICU admission being 0.774 (95% CI: 0.657–0.890) and that for the combined endpoint of ICU admission or death within 48 h being 0.815 (0.730–0.990) – see Fig. 2. For the combined endpoint of death in
Please cite this article in press as: Silcock DJ, et al. Validation of the National Early Warning Score in the prehospital setting. Resuscitation (2015), http://dx.doi.org/10.1016/j.resuscitation.2014.12.029
83 84 85 86 87 88 89 90
91 92 93 94 95
96 97 98 99 100 101 102
103
104 105 106 107 108 109 110
111
112 113 114 115 116 117 118 119 120 121 122 123
G Model
ARTICLE IN PRESS
RESUS 6256 1–5
D.J. Silcock et al. / Resuscitation xxx (2015) xxx–xxx
Ulity of NEWS as a screening tool 100
0.9
90
0.8
80
0.7
70
Affected %
Sensivity
48 hour mortality and ICU admission 1
0.6 0.5 48 hour mortality
0.4 0.3
ICU admission
0.2 Combined outcome
0.1 0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1-specificity Fig. 2. ROC curves demonstrating prediction of 48hr mortality or ICU admission within 48 h of presentation, both individually and as a combined endpoint.
125
the Emergency Department or admission directly to ICU from the ED, the AUROC was 0.889 (0.823–0.957).
126
3.2. Outcomes in the various risk strata identified by NEWS
Outcomes for the patients in the various NEWS categories are given in Table 2. When the 2 test was used to compare out128 comes in the groups, there was no significant difference in 30-day 129 or 48-h mortality in the medium risk category compared to the 130 low risk category, although there was a significant increase in ICU 131 admission. The high risk group demonstrated statistically signif132 icant differences in respect of 30-day mortality, 48-h mortality, 133 and ICU admission in the first 48 h, with risk ratios of 6.7, 35 and 134 5.4, respectively. These findings were similar in trauma and non135 Q4 trauma patients (Table 3). 136 127
3.3. Efficiency of the NEWS
144
Fig. 3 shows the likely impact in terms of patients that would be triaged as high risk at a given NEWS score cut-off by plotting the total proportion of patients above a given NEWS cutoff against the proportion of patients in the various adverse outcome groups above a given NEWS cutoff. From this it can be seen that at a NEWS cutoff of 7, the 10% of patients triaged in account for 70% of those that die within 48 h.
145
4. Discussion
138 139 140 141 142 143
146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164
30 day mortality
50 40
48hr mortality
30
ICU admission
20
ICU/48hr mortality
10 Death/ICU from ED
0
137
60
0
0
124
3
Triage is a crucial part of any unplanned care system, and has been developed significantly since its inception during the Napoleonic Wars. Algorithms such as the Manchester Triage System16 provide objective criteria for the allocation of patients to care areas, and can be sensitive enough to detect the signs of critical illness at the point of entry to the emergency department although they may still miss patients with the possibility of deterioration while still in the Emergency Department.17 Early warning systems have been increasingly employed in the inpatient setting throughout the world in recent years and provide a means of identification of potential deterioration. However these scores are only effective in reducing adverse outcomes if an appropriate clinical response involving those experienced in critical care can be available to respond to the changing clinical situation.18 Scores developed for one patient population or healthcare system may not be applicable to other populations or healthcare systems,19 and even within a given healthcare system there may be individuals for whom the standard scoring points and triggers are inappropriate. Modifications may need to be made to the trigger points in
0
10
20
30
40
50
60
70
80
90
100
Included % Fig. 3. Total patients above a given NEWS score (x) vs. the proportion of patients in a given outcome group with that score or above (y). NEWS = 0 is at the top-right. For example NEWS = 7 includes 10.4% total patients but includes 71% of the 48 h mortality group.
certain other groups of patients with chronically deranged vital observations.20,21 Early warning scores are also reliant on clinical observations being performed at an appropriate frequency to be useful as a predictor of deterioration, and this may be limited by other workload.22 There is some evidence, however, that the introduction of clinical observations charts printed with early warning score calculations increases the incidence of respiratory rate being recorded, itself a valuable independent marker of deterioration.23,24 Compared to the inpatient patient population, the adoption of early warning scoring systems has been less enthusiastic in the emergency and prehospital settings11 partly as there is a relative lack of robust evidence supporting their validity in this patient cohort. However usage of aggregate scoring systems such as NEWS is increasing and is developing a role in this respect as a tool to predict the need for hospital admission,25 as well as likely outcomes, particularly in sepsis.26 This has been helped greatly by the development of point-of-care and rapidly available blood analysis, which improves the prognostic accuracy of the score.27,28 No early warning score can replace clinical assessment, and there are many situations where decisions about clinical management should be based on other criteria: the presence of ST-elevation myocardial infarction, major haemorrhage, or multiple trauma being only a few examples. In the absence of such presentations, signs of potentially severe illness or occult injury may occasionally be missed. In this scenario, a physiologically based score may highlight individual patients in need of more urgent care,29 a particular example being patients with suspected sepsis, where timely administration of antibiotics can make a significant difference in outcome. This may form part of a pre-alert protocol30 or indicate specific pre-hospital treatments. Similarly at the other end of the severity spectrum, a low score in combination with an appropriate clinical assessment may safely allow treatment at a location other than an Emergency Department.31 This study does demonstrate that there is a role for the NEWS in the prehospital setting, and its ability to identify high-risk patients should encourage its use at an early stage. In particular, across the whole range of patient presentations, patients with a NEWS of 7 or greater had 11% chance of death or ICU admission within 48 h. Although still low risk, there is a statistically significant rise in ICU admission in the medium risk group, and scoring shows good sensitivity for adverse outcomes. While the use of NEWS should not replace clinical judgement and a normal score does not preclude serious pathology, the unexpected finding of a high score should prompt increased prioritisation.
Please cite this article in press as: Silcock DJ, et al. Validation of the National Early Warning Score in the prehospital setting. Resuscitation (2015), http://dx.doi.org/10.1016/j.resuscitation.2014.12.029
165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209
G Model
ARTICLE IN PRESS
RESUS 6256 1–5
D.J. Silcock et al. / Resuscitation xxx (2015) xxx–xxx
4 Table 2 Risks of each outcome at each NEWS category. N
Medium risk
High risk
Non-trauma 214
Trauma 37
All 251
Non-trauma 135
Trauma 11
All 146
2 0.17 8.386 2.79–25.21 <0.001
19 0.13 6.710 3.79–11.88 <0.001
30 day mortality
n AR RR 95% p
6 0.029 1.443 0.58–3.57 0.425
0 0 0 0.11–6.33 0.84
6 0.02 1.242 0.52–3.00 0.63
17 0.13 6.482 3.46–12.16 <0.001
ICU admission
n AR RR 95% p
7 0.03 2.466 1.0–6.09 0.045
0 0 0.71
7 0.03 2.59 1.06–6.35 0.03
7 0.05 4.900 1.90–12.66 0.002
1 0 – <0.001
8 0.06 5.433 2.29–12.89 <0.001
48 h mortality
n AR RR 95% p
1 0.17 1.523 0.159–14.575 0.72
0 0 – – 0.61
1 0.004 1.726 0.18–16.52 0.63
10 0.07 24.148 6.73–86.65 <0.001
2 0.18 – – <0.001
12 0.09 35.315 10.08–123.7 <0.001
AR: absolute risk, RR: relative risk, 95%: 95% confidence interval of relative risk. There were no ICU admissions or deaths within 48 h in the low risk trauma group therefore it was not possible to calculate relative risks.
Table 3 Sensitivity and specificity for various outcomes at the cutoff scores suggested in the NEWS specification.
30 day mortality
n
Total ≥1 1126
Medium risk 397
High risk 146
Sensitivity
0.976 (0.874–0.999) 0.134 (0.115–0.154) 0.036 0.994
0.547 (0.386–0.702) 0.743 (0.718–0.767) 0.067 0.980
0.404 (0.256– 0.567) 0.906 (0.888–0.921) 0.125 0.978
1 (0.768–1)* 0.132 (0.114–0.152) 0.012 1
0.785 (0.492–0.953) 0.739 (0.714–0.763) 0.031 0.996
0.714 (0.418–0.916) 0.902 (0.884–0.918) 0.074 0.996
0.941 (0.713–0.998) 0.132 (0.113–0.151) 0.014 0.994
0.764 (0.501–0.931) 0.740 (0.715–0.764) 0.038 0.995
0.411 (0.184–0.670) 0.900 (0.882–0.915) 0.051 0.991
0.966 (0.827–0.999) 0.132 (0.115–0.153) 0.025 0.994
0.767 (0.577–0.900) 0.745 (0.720–0.769) 0.067 0.992
0.533 (0.343–0.716) 0.906 (0.888–0.921) 0.11 0.980
Specificity PPV NPV 48 h mortality
Sensitivity Specificity PPV NPV
Admission to ICU
Sensitivity Specificity PPV NPV
Combined ICU/48hr mortality
Sensitivity Specificity PPV NPV
Bracketed numbers indicate 95% confidence interval except as marked * which indicates 97.5% single sided CI. PPV: positive predictive value; NPV: negative predictive value.
210
211 212 213 214 215 216 217 218 219 220 221 222
5. Limitations This study does have some limitations. We have not adjusted for age or sex differences between the low, medium and high risk NEWS groups, and this may account for some of the difference in mortality. However, the absence of age or gender adjustments is a feature of the NEWS that our study aimed to validate. As one of the end-points was survival to discharge and only in-hospital death was considered, the mortality rate, particularly the 30day mortality rate, may be underestimated. Similarly, due to the method of data collection, it was not possible to conclusively identify all patients who were re-admitted following discharge, although there were no deaths among those that could be identified as being repeat attenders. The overall mortality rate was low, as was the
proportion of people with high NEWS scores and the total numbers in these categories was also low, particularly among trauma patients.
6. Conclusion Elevated NEWS among unselected prehospital patients is associated with increased levels of adverse outcomes. Calculation of an early warning score prior to transfer to hospital is straightforward and may be a useful triage tool with potential to facilitate earlier recognition of at-risk or deteriorating patients, possibly allowing earlier involvement of appropriate ED and critical care staff.
Please cite this article in press as: Silcock DJ, et al. Validation of the National Early Warning Score in the prehospital setting. Resuscitation (2015), http://dx.doi.org/10.1016/j.resuscitation.2014.12.029
223 224 225
226
227 228 229 230 231 232 233
G Model RESUS 6256 1–5
ARTICLE IN PRESS D.J. Silcock et al. / Resuscitation xxx (2015) xxx–xxx
234
235
236 Q5
237
238
239 240 241
242 243 Q6 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279
Conflict of interest statement No conflicts of interest to declare. Uncited references 32,33. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.resuscitation. 2014.12.029. References 1. Rivers E, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001;345:1368–77. 2. Dellinger RP, et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock. Intensive Care Med 2012;39:165–228, http://dx.doi.org/10.1007/s00134-012-2769-8. 3. Smith T, et al. Accuracy of an expanded early warning score for patients in general and trauma surgery wards. Br J Surg 2012;99:192–7, http://dx.doi.org/10.1002/bjs.7777. ISSN: 0007-1323. 4. Morgan RJM, Williams F, Wright MM. An early warning score for the early detection of patients with impending illness. Clin Intensive Care 1997;8:100. 5. Goldhill DR, McNarry AF. Physiological abnormalities in early warning scores are related to mortality in adult inpatients. Br J Anaesth 2004;92:882–4, http://dx.doi.org/10.1093/bja/aeh113. 6. Standardising the assessment of acute-illness severity in the NHS. Report of a working party. London: Royal College of Physicians; 2012. Gary B. Have we found the perfect early war7. Smith ning score? A view of ViEWS. Resuscitation 2013;84:707–8, http://dx.doi.org/10.1016/j.resuscitation.2013.04.001. 8. Groarke JD, et al. Use of an admission early warning score to predict patient morbidity and mortality and treatment success. Emerg Med J 2008;25:803–6, http://dx.doi.org/10.1136/emj.2007.051425. 9. Griffiths JR, Kidney EM. Current use of early warning scores UK emergency departments. Emerg Med J 2012;29:65–6, in http://dx.doi.org/10.1136/emermed-2011-200508. 10. Rees JE, Mann C. Use of the patient at risk scores in the emergency department: a preliminary study. Emerg Med J 2004;21:698–9, http://dx.doi.org/10.1136/emj.2003.006197. 11. Timothy J. An early warning? Universal risk scoring in emergency medicine. Roland, Damian and Coats. Emerg Med J 2011;28:263, http://dx.doi.org/10.1136/emj.2010.106104. 12. Roland D, Jahn H. Are early warning scores too early for paramedic practice? J Paramed Pract 2012;4:16. 13. Correspondence between National Ambulance Service Medical Directors Group and Royal College of Physicians; 2013 [unpublished]. 14. Scottish Ambulance. Service Clinical Strategy 2011–2015; 2010. 15. Prytherch DR, et al. ViEWS – towards a national early warning score for detecting adult inpatient deterioration. Resuscitation 2010;81:932–7, http://dx.doi.org/10.1016/j.resuscitation.2010.04.014.
5
16. Manchester Triage. Group Emergency Triage. Blackwell; 2006. 17. Cooke MW, Jinks S. Does the Manchester triage system detect the critically ill? J Accid Emerg Med 1999;16:179–81. 18. McNeill G, Bryden D. Do either early warning systems or emergency response teams improve hospital patient survival? A systematic review. Resuscitation 2013;84:1652–67, http://dx.doi.org/10.1016/j.resuscitation.2013.08.006. 19. Wheeler India, et al. Early warning scores generated in developed healthcare settings are not sufficient at predicting early mortality in Blantyre, Malawi: a prospective cohort study. 3. PLOS ONE 2013;8:e59830, http://dx.doi.org/10.1371/journal.pone.0059830. 20. Eccles Sinan R, et al. CREWS: improving specificity whilst mainsensitivity of the National Early Warning Score in taining patients with chronic hypoxaemia. Resuscitation 2013;85:109–11, http://dx.doi.org/10.1016/j.resuscitation.2013.08.277. 21. O’Driscoll R. Emergency oxygen use. Br Med J 2012;345:e6856. 22. Hands C, et al. Patterns in the recording of vital signs and early warning scores: compliance with a clinical escalation protocol. BMJ Qual Saf 2013;22:719–26, http://dx.doi.org/10.1136/bmjqs-2013-001954. 23. McBride Jackie, et al. Long-term effect of introducing an early warning score on respiratory rate charting on general wards. Resuscitation 2005;65:41–4, http://dx.doi.org/10.1016/j.resuscitation.2004.10.015. 24. Hammond NE, et al. The effect of implementing a modified early warning scoring (MEWS) system on the adequacy of vital sign documentation. Aust Crit Care 2013;26:18–22, http://dx.doi.org/10.1016/j.aucc.2012.05.001. 25. Burch VC, Tarr G, Morroni C. Modified early warning score predicts the need for hospital admission and inhospital mortality. Emerg Med J 2008;25:674–8, http://dx.doi.org/10.1136/emj.2007.057661. 26. Corfield AR, et al. Utility of a single early warning score in patients with sepsis in the emergency department. Emerg Med J 2013. 27. Mohammed MA, et al. Index blood tests and national early warning scores within 24 hours of emergency admission can predict the risk of in-hospital mortality: a model development and validation study. PLOS ONE 2013;8:e64340, http://dx.doi.org/10.1371/journal.pone.0064340. 28. Jarvis S, et al. Combining the National Early Warning Score with an early warning score based on common laboratory test results better discriminates patients at risk of hospital mortality. Rapid response systems and medical emergency teams. Projects: CHMI. Health informatics; 2013. 29. Fullerton JN, et al. Is the Modified Early Warning Score (MEWS) superior to clinician judgement in detecting critical illness in pre-hospital environment? Resuscitation 2012;83:557–62, the http://dx.doi.org/10.1016/j.resuscitation.2012.01.004. SM, Bloch M. An evaluation of a new prehospi30. Booth pre-alert guidance tool. Emerg Med J 2013;30:820–3, tal http://dx.doi.org/10.1136/emermed-2012-201545. 31. Challen K, Walter D. Physiological scoring: an aid to emergency medical services transport decisions? Prehosp Disaster Med 2010;25:320–3. 32. Smith GB, et al. The ability of the National Early Warning Score (NEWS) to discriminate patients at risk of early cardiac arrest, unanticipated intensive care unit admission, and death. Resuscitation 2013;84:465–70, http://dx.doi.org/10.1016/j.resuscitation.2012.12.016. 33. Jo S, et al. Comparison of the trauma and injury severity score and modified early warning score with rapid lactate level (the ViEWS-L score) in blunt trauma patients. Eur J Emerg Med 2013, http://dx.doi.org/10.1097/MEJ.0b013e32836192d6.
Please cite this article in press as: Silcock DJ, et al. Validation of the National Early Warning Score in the prehospital setting. Resuscitation (2015), http://dx.doi.org/10.1016/j.resuscitation.2014.12.029
280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333