- Email: [email protected]
Factors affecting response to National Early Warning Score (NEWS)
Resuscitation 90 (2015) 85–90
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Rapid Response Systems
Factors affecting response to National Early Warning Score (NEWS)夽 Ivana Kolic a,∗ , Smiley Crane a , Suzanne McCartney a , Zane Perkins b , Alex Taylor a a b
Acute Medical Unit, The Queen Elizabeth Hospital, Lewisham & Greenwich NHS Trust, London, United Kingdom William Harvey Hospital, East Kent Hospitals University NHS Trust, Ashford, United Kingdom
a r t i c l e
i n f o
Article history: Received 9 August 2014 Received in revised form 28 January 2015 Accepted 8 February 2015 Keywords: National Early Warning Score Critical illness Patient safety Clinical response Consistency
a b s t r a c t Introduction: The NEWS is a physiological score, which prescribes an appropriate response for the deteriorating patient in need of urgent medical care. However, it has been suggested that compliance with early warning scoring systems for identifying patient deterioration may vary out of hours. We aimed to (1) assess the scoring accuracy and the adequacy of the prescribed clinical responses to NEWS and (2) assess whether responses were affected by time of day, day of week and score severity. Methods: We performed a prospective observational study of 370 adult patients admitted to an acute medical ward in a London District General Hospital. Patient characteristics, NEW score, time of day, day of week and clinical response data were collected for the first 24 h of admission. Patients with less than a 12 h hospital stay were excluded. We analysed data with univariate and multivariate logistic regression. Results: In 70 patients (18.9%) the NEW score was calculated incorrectly. There was a worsening of the clinical response with increasing NEW score. An appropriate clinical response to the NEWS was observed in 274 patients (74.1%). Patients admitted on the weekend were more likely to receive an inadequate response, compared to patients admitted during the week (p < 0.0001). After adjusting for confounders, increasing NEWS score remained significantly associated with an inadequate clinical response. Furthermore, our results demonstrate a small increase in inadequate NEWS responses at night, however this was not clinically or statistically significant. Conclusion: The high rate of incorrectly calculated NEW scores has implications for the prescribed actions. Clinical response to NEWS score triggers is significantly worse at weekends, highlighting an important patient safety concern. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Hospitalised patients who develop acute illness often exhibit preceding abnormalities in vital signs.1,2 Early Warning Scores (EWS) aim to identify these changes and allow an opportunity for early intervention and timely treatment.3,4 As a result, NICE Guidelines recommend that EWS should be used to monitor all adult patients in acute hospital settings.5 A variety of EWS have been used across the United Kingdom (UK), with the NEWS having been shown to be better than 33 other EWS.6 The NEWS is a scoring system for the prevention and early identification of patients who develop or present with acute illness.7–9 However, the effectiveness of EWS is dependent on appropriate implementation,10 compliance
夽 A Spanish translated version of the abstract of this article appears as Appendix in the final online version at http://dx.doi.org/10.1016/j.resuscitation.2015.02.009. ∗ Corresponding author at: Acute Medical Unit, The Queen Elizabeth Hospital, Lewisham & Greenwich NHS Trust, Stadium Road, London SE18 4QH, United Kingdom. E-mail address: [email protected] (I. Kolic). http://dx.doi.org/10.1016/j.resuscitation.2015.02.009 0300-9572/© 2015 Elsevier Ireland Ltd. All rights reserved.
and an effective clinical response.5,11,12 It has been suggested that compliance with track-and-trigger systems for identifying patient deterioration may vary out of hours.13 A principle of the NHS is ‘equality of treatment or clinical outcome regardless of the day of the week’ as outlined in The Foster Report.14 However, it is recognised that mortality is higher in patients admitted to hospital out of hours.15–17 The overall aim of this study was to assess the association between the appropriate use of NEWS and out-of hours activity. Our first aim was to assess the scoring accuracy and the adequacy of the prescribed clinical responses to NEWS. Secondly, our aim was to assess whether responses were affected by time of day, day of week and score severity as a possible explanation for the increased mortality at these times. 2. Methodology We conducted an observational study in patients (n = 370) presenting to the Acute Medical Unit (AMU) from 1 October 2013 to 15 October 2013 and from 9th December 2013 to 22nd December
86
I. Kolic et al. / Resuscitation 90 (2015) 85–90
Fig. 1. The NEWS Scoring System. Figure published with permission from the Royal College of Physicians.9
2013. The protocol was reviewed by the Local Clinical Effectiveness Department and met National Institute of Health Research criteria for service evaluation. The study was conducted in Queen Elizabeth Hospital (QEH), a National Health Service (NHS) Trust District General hospital in London. Eligibility included adult patients on two acute medical wards in QEH. Exclusion criteria involved patients with a less than 12 h inpatient stay. Data was collected prospectively. Information was collected on patient baseline characteristics, documented scores allocated to each physiological parameter, documented total scores, recalculated scores (manually calculated from the documented vital signs on the observation charts), time to subsequent observations and the adequacy of clinical responses. Points were allocated according to basic clinical observations including pulse rate, respiratory rate, blood pressure, oxygen saturation, and level of
consciousness (see Fig. 1). Time of day and day of week was noted. For NEWS categories 3 (score 5–6) and Category 4 (score ≥ 7), we looked into the reasons for an inadequate clinical response. 2.1. Outcomes Two outcomes were scoring error and adequacy of the clinical response. For a clinical response to be adequate we required the prescribed actions to be carried out according to level of score as set out by the Royal College of Physicians report ‘Standardising the assessment of acute-illness severity in the NHS 2012’.9 This involves the correct frequency of observations and the correct action (Fig. 2). The correct action requires both the appropriate urgency and competency of the clinical responder and the appropriate clinical environment. Furthermore, we collected data on
Fig. 2. Clinical response to NEWS triggers. Figure published with permission from the Royal College of Physicians.
I. Kolic et al. / Resuscitation 90 (2015) 85–90
87
Table 1 Univariate analysis of factors associated with a NEWS scoring error. Factor
Correct score
Age
Incorrect score
77 (18–102)
Univariate analysis Crude OR (95% CI)
p-Value
77 (23–97)
1.01 (0.99–1.02)
0.160
Time of day
Day (9 am–9 pm) Night (9 pm–9 am)
149 (79.7) 151 (82.5)
38 (20.3) 32 (17.5)
0.83 (0.49–1.40)
0.487
Day of week
Weekday Weekend
237 (80.3) 63 (84.0)
58 (19.7) 12 (16.0)
0.78 (0.39–1.54)
0.471
NEW score
NEWS 0 NEWS 1–4 NEWS 5–6 NEWS 7
131 (87.9) 156 (78.0) 9 (69.2) 4 (50.0)
18 (12.1) 44 (22.0) 4 (30.8) 4 (50.0)
1.0 2.05 (1.13–3.72) 3.24 (0.90–11.60) 7.28 (1.67–31.68)
0.018 0.072 0.008
Data is presented as median (range) or number (percent). OR, Odds Ratio; CI, Confidence Interval; NEW score, National Early Warning Score. Binary logistic regression analysis with score error as the indicator dependent variable.
patient mortality, in order to assess whether inadequate NEWS responses were associated with worse outcomes. This outcome data was collected for the first cohort of patients. 2.2. Definitions The NEWS includes the following prescribed recommendations; a score of 0 (Category 1) requires minimum 12 hourly observation recording; a score of 1–4 (Category 2) requires minimum 4–6 hourly observation recording and the registered nurse to be informed; a score of 5 or 6, or 3 in one parameter, (Category 3) requires minimum 1 hourly observation recording, the registered nurse to immediately inform the medical team, urgent assessment by a competent clinical responder and clinical care in an environment with monitoring facilities; a score of 7 or more (Category 4) requires continuous monitoring of observations, a registered nurse to immediately inform the Specialist Registrar or more senior physician, emergency assessment by a clinical team competent in critical care with advanced airway skills and the consideration of transfer to a level 2 or 3 clinical environment.9 We defined there to be a scoring error if the score we calculated from the documented observations on the observation charts was different to the documented score on the NEWS chart. Patients were considered adults if their age was >16 years. We defined daytime to include time from 09:00 to 21:00.
Numerical data are reported as median with interquartile range (IQR) and categorical data as frequency (n) and percent (%). The Mann–Whitney U test was used to compare numerical data and Fisher’s Exact test was used to compare categorical data. For comparisons of multiple groups a Kruskal–Wallis test was used to compare numerical data and a chi-squared test was used to compare categorical data. A multivariable logistic regression model was developed to compare the effects of our variables on the adequacy of clinical responses. Patient characteristics significantly associated with inadequate clinical outcomes (p < 0.1) were included in the model. Statistical significance was set as a two tailed p-value of <0.05.
3. Results A total of 370 patients were included during the study period (Table 1). Median age was 77 years (range: 18–102). One hundred eighty seven patients (50.5%) were admitted during the day and 183 patients (49.5%) were admitted during the night. Two hundred ninety five patients (79.7%) were admitted on weekdays and 75 (20.3%) on the weekend.
3.1. Scoring error 2.3. Statistical analysis Statistical analysis was performed using SPSS version 20 (SPSS, Chicago, IL). Normal-quartile plots were used to test for normality.
In 70 patients (18.9%) the NEW score was calculated incorrectly. There was a significant increase in scoring errors as the NEW score increase (Table 1).
Table 2 Univariate and multivariate analysis of factors affecting the adequacy of the clinical response to NEWS score. Factor
Adequate response
Age
Inadequate response
Univariate analysis
Multivariate analysis
Crude OR (95% CI)
p-Value
Adjusted OR (95% CI)
p-Value
78 (18–102)
75 (21–97)
1.01 (0.99–1.02)
0.358
Time of day
Day (9am–9pm) Night (9pm–9am)
142 (75.9) 132 (72.1)
45 (24.1) 51 (27.9)
1.22 (0.77–1.94)
0.404
Day of week
Weekday Weekend
234 (79.3) 40 (53.3)
61 (20.7) 35 (46.7)
3.36 (1.97–5.73)
<0.0001
4.15 (2.24–7.69)
<0.0001
NEW score
NEWS 0 NEWS 1–4 NEWS 5–6 NEWS 7
137 (92.0) 135 (67.5) 0a 2 (25.0)
12 (8.0) 65 (32.5) 13 (100.0) 6 (75.0)
1.0 5.50 (2.84–10.64) 297 (16.64–5302) 34.25 (6.22–188.6)
<0.0001 <0.0001 <0.0001
1.0 6.13 (3.08–12.16) 177 (20.72–1510) 40.64 (7.04–234.7)
<0.0001 <0.0001 <0.0001
Data is presented as median (range) or number (percent). OR, Odds Ratio; CI, Confidence Interval; NEW score, National Early Warning Score. Binary logistic regression analysis with adequate response as the indicator dependent variable. a A contingency factor of 0.5 was used to enable logistic regression calculations.
88
I. Kolic et al. / Resuscitation 90 (2015) 85–90
Fig. 3. Graph to show the variation of adequate clinical responses with NEWS category.
3.2. Clinical response An appropriate clinical response to NEWS was observed in 274 patients (74.1%) with the remaining 96 patients (25.9%) receiving an inadequate response. There was a significant worsening of the clinical response with increasing NEW score (Table 2, Fig. 3). All 13 patients with a NEWS of 5 or 6 received an inadequate clinical response. The action taken was inappropriate in 12 of these patients and the frequency of repeat observations was inappropriate in 12 patients. Six patients (75%) with a NEWS ≥ 7 received an inadequate clinical response. Both the action taken and the frequency of repeat observations were incorrect for all six patients. The day of admission was also significantly associated with an inadequate clinical response (Table 2). Patients admitted on the weekend were more likely to receive an inadequate response than patients admitted during the week (p < 0.0001). After adjusting for confounders, day of admission (OR 4.15 (95% CI: 2.24–7.69)) and increasing NEWS remained significantly associated with an inadequate clinical response. Furthermore, our results demonstrate a small increase in inadequate NEWS responses at night, however this was not clinically or statistically significant ((night 53.1% vs. day 46.9%, p = 0.404)). 3.3. Outcome Complete outcome data was available for the first cohort of 199 patients. Eight patients who had an adequate response to the score died (6%) compared to six patients (8.5%) who had an inadequate clinical response (p = 0.573) (relative risk 1.35 (95% CI 0.49–3.74)). 4. Discussion This study shows that clinical response to NEWS is significantly worse on weekends. Our key finding is that there is a different standard of care at these times, which has implications for patient safety out of hours. Furthermore we showed scoring errors occurred more frequently with higher NEWS. EWS have been shown to predict mortality and length of hospital stay.18,19 To the best of our knowledge, there have been few studies looking at scoring errors and compliance with the NEWS. Even with the use of EWS, there are problems with sub-optimal recording of observations and clinical response to EWS.20 In some
cases there is little documentation of responses and intervention7 and in others the frequency of observations is incomplete.1 Gordon & Beckett have shown an accuracy in EWS documentation in 21% of patients,21 compared to our scoring accuracy of 81.1%. Shearer et al. showed an inadequacy in the clinical response to early warning scores in 42% of patients (n = 570),22 which is almost double that of the rate of inadequate responses in our study (25.9%). The MERIT study analysed 246 cardiac arrests and identified that in 44% of cases staff did not activate an emergency response despite having fulfilled the EWS criteria.20 Deficiencies in the overnight use of EWS have been shown by Gordon & Beckett, however no comparison was made with daytime, and limited only to patients already causing clinical concern (scores >4).21 Similar findings have been shown that weekends are associated with a different level of care. Peberdy et al. demonstrated that when critical events such as cardiac arrests occur out of hours patient outcomes are worse.23 Studies have demonstrated that mortality can be 10% higher in emergency admissions on weekends compared to weekdays.24 Specifically, mortality is higher for patients suffering from specific conditions, namely myocardial infarctions25 and upper gastrointestinal haemorrhage.26 In one study, 23 of the 100 leading causes of death had significantly increased mortality on weekends.17 In contrast to our study, compliance with EWS has been shown to vary with time of day. Peak compliance has been shown to occur during routine nursing observations.13 Nighttime work has been shown to be associated with a lower frequency of vital sign monitoring.27 Beckett et al.28 demonstrated that it was most common for delays in investigations and initiation of treatment to occur at night. However, a study on an intensive care unit has shown that the nighttime presence of an Intensivist on site did not improve patient outcomes.29 The primary aim of our study was to investigate if incorrect responses to NEWS occurred more frequently out of hours. There is evidence to show that the correct use of early warning scores improve medical care and patient outcomes.3 Our results show a trend towards increased mortality in people who had an incorrect response; eight patients who has an adequate response to the score died (6%) compared to six patients (8.5%) who had an inadequate clinical response. Complete outcome data was available for the first cohort of 199 patients. Clinically there was a significant difference but the results did not reach statistical significance (p = 0.573), demonstrating a type 2 error. The most likely explanation for this is the small sample size. Whether this difference would be significant with a larger sample size is unclear. Our above findings have several possible explanations. Numerous studies speculate whether the differences in care are as a result of lower staffing out of hours.17,24 Staffing levels have been shown to be lower on weekends,30 which results in a greater workload per healthcare provider.31 During times of high workload in the intensive care setting it was shown that patients had higher mortality rates.30 Often a large proportion of staff working on weekends are more junior and less experienced than those who work on weekdays and are often temporary staff and not familiar with the patients or workplace.17 Primary care services are less available out of hours24 and patients often experience difficulty in accessing these services at night or over the weekend.15 Although this has been largely thought not to be the case, Meynaar et al. have shown that patients tend to be more sick out of hours which may result in greater workload for staff.32 Staff have been interviewed as to the potential reasons for failing to activate rapid response systems and found key reasons to include that they felt the clinical situation was under control on the ward, that they had instigated other investigations or actions and that they feared negative reactions from colleagues about activating the system.22 The Foster Report
I. Kolic et al. / Resuscitation 90 (2015) 85–90
highlighted that if patients are admitted at the weekend they are less likely to receive prompt treatment and more likely to die and a possible explanation is the reduced number of senior doctors.14 In view of these concerns, we agree with the proposed seven day working week as set out by the Academy of Medical Royal Colleges report ‘Seven Day Consultant Present Care’,33 with particular focus on early and increased consultant level decision making. Increasing the availability of diagnostic services out of hours may contribute to improving care out of hours. Furthermore, we could recommend the use of automated observation calculation as is being used across trusts in the United Kingdom, with the aim of decreasing scoring. Our results showed no difference in frequency of vital sign recording for higher NEW scores. Contrary to this, Hands et al.27 demonstrated more frequent observation recording with sicker patients, based on their early warning scores. These findings raise some concern. This may be explained by a lack of staff available to provide the necessary increase in recording observations. Another possible explanation is that the frequency prescribed by the EWS imposes an unachievable routine, owing to staffing levels, the availability of medical diagnostic equipment and the business of the clinical shifts.27 The clinical workload may be as an indirect effect of reduced staffing levels, or due to heavy workload and sick patients. The key strength of our study is that we are directly assessing clinical practice. Furthermore, strength is the prospective nature of the study and the completeness of data collection from ward observation charts. This allows us to highlight specific areas of care that need attention. 4.1. Limitations Firstly, this is an observational study and does not demonstrate the effect of NEWS utilisation on patient outcome. We only looked at factors affecting scoring error and compliance with NEWS. The small sample size is a limitation. Our study looks at a snapshot of compliance with the NEWS for each patient, and only looks at patients in acute medical wards, which impacts on the generalisability of our results. However the factors found were not specific to acute medical wards and would be important to other areas of hospital. None of the above limitations detract from the fact that the illustrated factors affect clinical responses to NEWS. Patients should receive the same standard of clinical care at any time of day or day of the week that they are in hospital. Attempts to establish a 24/7 inpatient service has been restricted due to limited resources and restrictions on working hours for doctors in training.15 Aylin et al. suggest that hospitals need to re-evaluate the level of services they provide out of hours.24 Our study findings may be a favourable factor towards the seven day working week as set out by the Academy of Medical Royal Colleges report ‘Seven Day Consultant Present Care’.33 This may contribute towards achieving parity for inpatient care throughout the week.34 4.2. Conclusion We have shown that clinical responses to EWS were significantly worse on weekend days. This may reflect differences in quality of care out of hours and highlights an important patient safety concern. This in turn may contribute to the increased mortality seen at these times. Authors’ contributions IK and AT involved in study conception and design; IK, SM and SC involved in data collection; ZP and IK involved in data analysis and interpretation; IK, SC and ZP involved in drafting of manuscript; IK, SM, SC, ZP and AT involved in critical revision.
89
Conflicts of interest statement None declared. 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. 2015.02.009. References 1. Smith GB, Prytherch DR, Schmidt P, et al. Hospital-wide physiological surveillance – a new approach to the early identification and management of the sick patient. Resuscitation 2006;71:19–28. 2. Kause J, Smith G, Prytherch D, et al. A comparison of antecedents to cardiac arrests, deaths and emergency intensive care admissions in Australia and New Zealand, and the United Kingdom – the ACADEMIA study. Resuscitation 2004;62:275–82. 3. Alam N, Hobbelink EL, van Tienhoven AJ, et al. The impact of the use of the Early Warning Score (EWS) on patient outcomes: a systematic review. Resuscitation 2014;85:587–94. 4. Jansen JO, Cuthbertson BH. Detecting critical illness outside the ICU: the role of track and trigger systems. Curr Opin Crit Care 2010;16:184–90. 5. NIfHaCE. NICE clinical guideline 50 Acutely ill patients in hospital: recognition of and response to acute illness in adults in hospital. London: National Institute for Health and Clinical Excellence; 2007. 6. Smith GB, Prytherch DR, Meredith P, 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. 7. Hodgetts TJ, Kenward G, Vlachonikolis IG, et al. The identification of risk factors for cardiac arrest and formulation of activation criteria to alert a medical emergency team. Resuscitation 2002;54:125–31. 8. Jones DA, DeVita MA, Bellomo R. Rapid-response teams. N Engl J Med 2011;365:139–46. 9. Staff RCoPoL. National Early Warning Score (NEWS): standardising the assessment of acute-illness severity in the NHS. Royal College of Physicians of London; 2012. 10. Ludikhuize J, Borgert M, Binnekade J, et al. Standardized measurement of the Modified Early Warning Score results in enhanced implementation of a Rapid Response System: a quasi-experimental study. Resuscitation 2014;85:676–82. 11. Martin I, Mason M, Mason D, et al. Emergency admissions: a journey in the right direction? NCEPOD; 2007. 12. Force RCoPoLAMT, Royal College of Physicians, Williams B. Acute medical care: the right person, in the right setting – first time. Royal College of Physicians; 2007. 13. Jones D, Bates S, Warrillow S, et al. Circadian pattern of activation of the medical emergency team in a teaching hospital. Crit Care 2005;9:R303–6. 14. Intelligence DF. Inside your hospital: Dr Foster Hospital Guide 2001–2011. London: Dr Foster Limited; 2011. 15. Maggs F, Mallet M. Mortality in out-of-hours emergency medical admissions – more than just a weekend effect. J R Coll Physicians Edinb 2010;40:115–8. 16. Laupland KB, Shahpori R, Kirkpatrick AW, et al. Hospital mortality among adults admitted to and discharged from intensive care on weekends and evenings. J Crit Care 2008;23:317–24. 17. Bell CM, Redelmeier DA. Mortality among patients admitted to hospitals on weekends as compared with weekdays. N Engl J Med 2001;345:663–8. 18. Gao H, McDonnell A, Harrison DA, et al. Systematic review and evaluation of physiological track and trigger warning systems for identifying at-risk patients on the ward. Intensive Care Med 2007;33:667–79. 19. Groarke JD, Gallagher J, Stack J, 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. 20. Hillman K, Chen J, Cretikos M, et al. Introduction of the medical emergency team (MET) system: a cluster-randomised controlled trial. Lancet 2005;365:2091–7. 21. Gordon CF, Beckett DJ. Significant deficiencies in the overnight use of a Standardised Early Warning Scoring system in a teaching hospital. Scott Med J 2011;56:15–8. 22. Shearer B, Marshall S, Buist MD, et al. What stops hospital clinical staff from following protocols? An analysis of the incidence and factors behind the failure of bedside clinical staff to activate the rapid response system in a multi-campus Australian metropolitan healthcare service. BMJ Qual Saf 2012;21:569–75. 23. Peberdy MA, Ornato JP, Larkin GL, et al. Survival from in-hospital cardiac arrest during nights and weekends. JAMA 2008;299:785–92. 24. Aylin P, Yunus A, Bottle A, et al. Weekend mortality for emergency admissions. A large, multicentre study. Qual Saf Health Care 2010;19:213–7. 25. Kostis WJ, Demissie K, Marcella SW, et al. Weekend versus weekday admission and mortality from myocardial infarction. N Engl J Med 2007;356:1099–109. 26. Ananthakrishnan AN, McGinley EL, Saeian K. Outcomes of weekend admissions for upper gastrointestinal hemorrhage: a nationwide analysis. Clin Gastroenterol Hepatol 2009;7:296–3020.
90
I. Kolic et al. / Resuscitation 90 (2015) 85–90
27. Hands C, Reid E, Meredith P, 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. 28. Beckett DJ, Gordon CF, Paterson R, et al. Improvement in out-of-hours outcomes following the implementation of Hospital at Night. QJM 2009;102:539–46. 29. Kerlin MP, Small DS, Cooney E, et al. A randomized trial of nighttime physician staffing in an intensive care unit. N Engl J Med 2013;368:2201–9. 30. Tarnow L, Rossing P, Nielsen FS, et al. Cardiovascular morbidity and early mortality cluster in parents of type 1 diabetic patients with diabetic nephropathy. Diabetes Care 2000;23:30–3.
31. Goddard L. Vaccinations (or lack of them) for student doctors. N Z Med J 2009;122:128. 32. Meynaar IA, van der Spoel JI, Rommes JH, et al. Off hour admission to an intensivist-led ICU is not associated with increased mortality. Crit Care 2009;13:R84. 33. Academy of Medical Royal Colleges. Seven day consultant present care: implementation considerations. Academy of Medical Royal Colleges; 2013. 34. Temple J. Time for training. A review of the impact of the European working time directive on the quality of training. Med Educ Engl 2010:9–10 (Nil volume).
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Rapid Response Systems
Factors affecting response to National Early Warning Score (NEWS)夽 Ivana Kolic a,∗ , Smiley Crane a , Suzanne McCartney a , Zane Perkins b , Alex Taylor a a b
Acute Medical Unit, The Queen Elizabeth Hospital, Lewisham & Greenwich NHS Trust, London, United Kingdom William Harvey Hospital, East Kent Hospitals University NHS Trust, Ashford, United Kingdom
a r t i c l e
i n f o
Article history: Received 9 August 2014 Received in revised form 28 January 2015 Accepted 8 February 2015 Keywords: National Early Warning Score Critical illness Patient safety Clinical response Consistency
a b s t r a c t Introduction: The NEWS is a physiological score, which prescribes an appropriate response for the deteriorating patient in need of urgent medical care. However, it has been suggested that compliance with early warning scoring systems for identifying patient deterioration may vary out of hours. We aimed to (1) assess the scoring accuracy and the adequacy of the prescribed clinical responses to NEWS and (2) assess whether responses were affected by time of day, day of week and score severity. Methods: We performed a prospective observational study of 370 adult patients admitted to an acute medical ward in a London District General Hospital. Patient characteristics, NEW score, time of day, day of week and clinical response data were collected for the first 24 h of admission. Patients with less than a 12 h hospital stay were excluded. We analysed data with univariate and multivariate logistic regression. Results: In 70 patients (18.9%) the NEW score was calculated incorrectly. There was a worsening of the clinical response with increasing NEW score. An appropriate clinical response to the NEWS was observed in 274 patients (74.1%). Patients admitted on the weekend were more likely to receive an inadequate response, compared to patients admitted during the week (p < 0.0001). After adjusting for confounders, increasing NEWS score remained significantly associated with an inadequate clinical response. Furthermore, our results demonstrate a small increase in inadequate NEWS responses at night, however this was not clinically or statistically significant. Conclusion: The high rate of incorrectly calculated NEW scores has implications for the prescribed actions. Clinical response to NEWS score triggers is significantly worse at weekends, highlighting an important patient safety concern. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Hospitalised patients who develop acute illness often exhibit preceding abnormalities in vital signs.1,2 Early Warning Scores (EWS) aim to identify these changes and allow an opportunity for early intervention and timely treatment.3,4 As a result, NICE Guidelines recommend that EWS should be used to monitor all adult patients in acute hospital settings.5 A variety of EWS have been used across the United Kingdom (UK), with the NEWS having been shown to be better than 33 other EWS.6 The NEWS is a scoring system for the prevention and early identification of patients who develop or present with acute illness.7–9 However, the effectiveness of EWS is dependent on appropriate implementation,10 compliance
夽 A Spanish translated version of the abstract of this article appears as Appendix in the final online version at http://dx.doi.org/10.1016/j.resuscitation.2015.02.009. ∗ Corresponding author at: Acute Medical Unit, The Queen Elizabeth Hospital, Lewisham & Greenwich NHS Trust, Stadium Road, London SE18 4QH, United Kingdom. E-mail address: [email protected] (I. Kolic). http://dx.doi.org/10.1016/j.resuscitation.2015.02.009 0300-9572/© 2015 Elsevier Ireland Ltd. All rights reserved.
and an effective clinical response.5,11,12 It has been suggested that compliance with track-and-trigger systems for identifying patient deterioration may vary out of hours.13 A principle of the NHS is ‘equality of treatment or clinical outcome regardless of the day of the week’ as outlined in The Foster Report.14 However, it is recognised that mortality is higher in patients admitted to hospital out of hours.15–17 The overall aim of this study was to assess the association between the appropriate use of NEWS and out-of hours activity. Our first aim was to assess the scoring accuracy and the adequacy of the prescribed clinical responses to NEWS. Secondly, our aim was to assess whether responses were affected by time of day, day of week and score severity as a possible explanation for the increased mortality at these times. 2. Methodology We conducted an observational study in patients (n = 370) presenting to the Acute Medical Unit (AMU) from 1 October 2013 to 15 October 2013 and from 9th December 2013 to 22nd December
86
I. Kolic et al. / Resuscitation 90 (2015) 85–90
Fig. 1. The NEWS Scoring System. Figure published with permission from the Royal College of Physicians.9
2013. The protocol was reviewed by the Local Clinical Effectiveness Department and met National Institute of Health Research criteria for service evaluation. The study was conducted in Queen Elizabeth Hospital (QEH), a National Health Service (NHS) Trust District General hospital in London. Eligibility included adult patients on two acute medical wards in QEH. Exclusion criteria involved patients with a less than 12 h inpatient stay. Data was collected prospectively. Information was collected on patient baseline characteristics, documented scores allocated to each physiological parameter, documented total scores, recalculated scores (manually calculated from the documented vital signs on the observation charts), time to subsequent observations and the adequacy of clinical responses. Points were allocated according to basic clinical observations including pulse rate, respiratory rate, blood pressure, oxygen saturation, and level of
consciousness (see Fig. 1). Time of day and day of week was noted. For NEWS categories 3 (score 5–6) and Category 4 (score ≥ 7), we looked into the reasons for an inadequate clinical response. 2.1. Outcomes Two outcomes were scoring error and adequacy of the clinical response. For a clinical response to be adequate we required the prescribed actions to be carried out according to level of score as set out by the Royal College of Physicians report ‘Standardising the assessment of acute-illness severity in the NHS 2012’.9 This involves the correct frequency of observations and the correct action (Fig. 2). The correct action requires both the appropriate urgency and competency of the clinical responder and the appropriate clinical environment. Furthermore, we collected data on
Fig. 2. Clinical response to NEWS triggers. Figure published with permission from the Royal College of Physicians.
I. Kolic et al. / Resuscitation 90 (2015) 85–90
87
Table 1 Univariate analysis of factors associated with a NEWS scoring error. Factor
Correct score
Age
Incorrect score
77 (18–102)
Univariate analysis Crude OR (95% CI)
p-Value
77 (23–97)
1.01 (0.99–1.02)
0.160
Time of day
Day (9 am–9 pm) Night (9 pm–9 am)
149 (79.7) 151 (82.5)
38 (20.3) 32 (17.5)
0.83 (0.49–1.40)
0.487
Day of week
Weekday Weekend
237 (80.3) 63 (84.0)
58 (19.7) 12 (16.0)
0.78 (0.39–1.54)
0.471
NEW score
NEWS 0 NEWS 1–4 NEWS 5–6 NEWS 7
131 (87.9) 156 (78.0) 9 (69.2) 4 (50.0)
18 (12.1) 44 (22.0) 4 (30.8) 4 (50.0)
1.0 2.05 (1.13–3.72) 3.24 (0.90–11.60) 7.28 (1.67–31.68)
0.018 0.072 0.008
Data is presented as median (range) or number (percent). OR, Odds Ratio; CI, Confidence Interval; NEW score, National Early Warning Score. Binary logistic regression analysis with score error as the indicator dependent variable.
patient mortality, in order to assess whether inadequate NEWS responses were associated with worse outcomes. This outcome data was collected for the first cohort of patients. 2.2. Definitions The NEWS includes the following prescribed recommendations; a score of 0 (Category 1) requires minimum 12 hourly observation recording; a score of 1–4 (Category 2) requires minimum 4–6 hourly observation recording and the registered nurse to be informed; a score of 5 or 6, or 3 in one parameter, (Category 3) requires minimum 1 hourly observation recording, the registered nurse to immediately inform the medical team, urgent assessment by a competent clinical responder and clinical care in an environment with monitoring facilities; a score of 7 or more (Category 4) requires continuous monitoring of observations, a registered nurse to immediately inform the Specialist Registrar or more senior physician, emergency assessment by a clinical team competent in critical care with advanced airway skills and the consideration of transfer to a level 2 or 3 clinical environment.9 We defined there to be a scoring error if the score we calculated from the documented observations on the observation charts was different to the documented score on the NEWS chart. Patients were considered adults if their age was >16 years. We defined daytime to include time from 09:00 to 21:00.
Numerical data are reported as median with interquartile range (IQR) and categorical data as frequency (n) and percent (%). The Mann–Whitney U test was used to compare numerical data and Fisher’s Exact test was used to compare categorical data. For comparisons of multiple groups a Kruskal–Wallis test was used to compare numerical data and a chi-squared test was used to compare categorical data. A multivariable logistic regression model was developed to compare the effects of our variables on the adequacy of clinical responses. Patient characteristics significantly associated with inadequate clinical outcomes (p < 0.1) were included in the model. Statistical significance was set as a two tailed p-value of <0.05.
3. Results A total of 370 patients were included during the study period (Table 1). Median age was 77 years (range: 18–102). One hundred eighty seven patients (50.5%) were admitted during the day and 183 patients (49.5%) were admitted during the night. Two hundred ninety five patients (79.7%) were admitted on weekdays and 75 (20.3%) on the weekend.
3.1. Scoring error 2.3. Statistical analysis Statistical analysis was performed using SPSS version 20 (SPSS, Chicago, IL). Normal-quartile plots were used to test for normality.
In 70 patients (18.9%) the NEW score was calculated incorrectly. There was a significant increase in scoring errors as the NEW score increase (Table 1).
Table 2 Univariate and multivariate analysis of factors affecting the adequacy of the clinical response to NEWS score. Factor
Adequate response
Age
Inadequate response
Univariate analysis
Multivariate analysis
Crude OR (95% CI)
p-Value
Adjusted OR (95% CI)
p-Value
78 (18–102)
75 (21–97)
1.01 (0.99–1.02)
0.358
Time of day
Day (9am–9pm) Night (9pm–9am)
142 (75.9) 132 (72.1)
45 (24.1) 51 (27.9)
1.22 (0.77–1.94)
0.404
Day of week
Weekday Weekend
234 (79.3) 40 (53.3)
61 (20.7) 35 (46.7)
3.36 (1.97–5.73)
<0.0001
4.15 (2.24–7.69)
<0.0001
NEW score
NEWS 0 NEWS 1–4 NEWS 5–6 NEWS 7
137 (92.0) 135 (67.5) 0a 2 (25.0)
12 (8.0) 65 (32.5) 13 (100.0) 6 (75.0)
1.0 5.50 (2.84–10.64) 297 (16.64–5302) 34.25 (6.22–188.6)
<0.0001 <0.0001 <0.0001
1.0 6.13 (3.08–12.16) 177 (20.72–1510) 40.64 (7.04–234.7)
<0.0001 <0.0001 <0.0001
Data is presented as median (range) or number (percent). OR, Odds Ratio; CI, Confidence Interval; NEW score, National Early Warning Score. Binary logistic regression analysis with adequate response as the indicator dependent variable. a A contingency factor of 0.5 was used to enable logistic regression calculations.
88
I. Kolic et al. / Resuscitation 90 (2015) 85–90
Fig. 3. Graph to show the variation of adequate clinical responses with NEWS category.
3.2. Clinical response An appropriate clinical response to NEWS was observed in 274 patients (74.1%) with the remaining 96 patients (25.9%) receiving an inadequate response. There was a significant worsening of the clinical response with increasing NEW score (Table 2, Fig. 3). All 13 patients with a NEWS of 5 or 6 received an inadequate clinical response. The action taken was inappropriate in 12 of these patients and the frequency of repeat observations was inappropriate in 12 patients. Six patients (75%) with a NEWS ≥ 7 received an inadequate clinical response. Both the action taken and the frequency of repeat observations were incorrect for all six patients. The day of admission was also significantly associated with an inadequate clinical response (Table 2). Patients admitted on the weekend were more likely to receive an inadequate response than patients admitted during the week (p < 0.0001). After adjusting for confounders, day of admission (OR 4.15 (95% CI: 2.24–7.69)) and increasing NEWS remained significantly associated with an inadequate clinical response. Furthermore, our results demonstrate a small increase in inadequate NEWS responses at night, however this was not clinically or statistically significant ((night 53.1% vs. day 46.9%, p = 0.404)). 3.3. Outcome Complete outcome data was available for the first cohort of 199 patients. Eight patients who had an adequate response to the score died (6%) compared to six patients (8.5%) who had an inadequate clinical response (p = 0.573) (relative risk 1.35 (95% CI 0.49–3.74)). 4. Discussion This study shows that clinical response to NEWS is significantly worse on weekends. Our key finding is that there is a different standard of care at these times, which has implications for patient safety out of hours. Furthermore we showed scoring errors occurred more frequently with higher NEWS. EWS have been shown to predict mortality and length of hospital stay.18,19 To the best of our knowledge, there have been few studies looking at scoring errors and compliance with the NEWS. Even with the use of EWS, there are problems with sub-optimal recording of observations and clinical response to EWS.20 In some
cases there is little documentation of responses and intervention7 and in others the frequency of observations is incomplete.1 Gordon & Beckett have shown an accuracy in EWS documentation in 21% of patients,21 compared to our scoring accuracy of 81.1%. Shearer et al. showed an inadequacy in the clinical response to early warning scores in 42% of patients (n = 570),22 which is almost double that of the rate of inadequate responses in our study (25.9%). The MERIT study analysed 246 cardiac arrests and identified that in 44% of cases staff did not activate an emergency response despite having fulfilled the EWS criteria.20 Deficiencies in the overnight use of EWS have been shown by Gordon & Beckett, however no comparison was made with daytime, and limited only to patients already causing clinical concern (scores >4).21 Similar findings have been shown that weekends are associated with a different level of care. Peberdy et al. demonstrated that when critical events such as cardiac arrests occur out of hours patient outcomes are worse.23 Studies have demonstrated that mortality can be 10% higher in emergency admissions on weekends compared to weekdays.24 Specifically, mortality is higher for patients suffering from specific conditions, namely myocardial infarctions25 and upper gastrointestinal haemorrhage.26 In one study, 23 of the 100 leading causes of death had significantly increased mortality on weekends.17 In contrast to our study, compliance with EWS has been shown to vary with time of day. Peak compliance has been shown to occur during routine nursing observations.13 Nighttime work has been shown to be associated with a lower frequency of vital sign monitoring.27 Beckett et al.28 demonstrated that it was most common for delays in investigations and initiation of treatment to occur at night. However, a study on an intensive care unit has shown that the nighttime presence of an Intensivist on site did not improve patient outcomes.29 The primary aim of our study was to investigate if incorrect responses to NEWS occurred more frequently out of hours. There is evidence to show that the correct use of early warning scores improve medical care and patient outcomes.3 Our results show a trend towards increased mortality in people who had an incorrect response; eight patients who has an adequate response to the score died (6%) compared to six patients (8.5%) who had an inadequate clinical response. Complete outcome data was available for the first cohort of 199 patients. Clinically there was a significant difference but the results did not reach statistical significance (p = 0.573), demonstrating a type 2 error. The most likely explanation for this is the small sample size. Whether this difference would be significant with a larger sample size is unclear. Our above findings have several possible explanations. Numerous studies speculate whether the differences in care are as a result of lower staffing out of hours.17,24 Staffing levels have been shown to be lower on weekends,30 which results in a greater workload per healthcare provider.31 During times of high workload in the intensive care setting it was shown that patients had higher mortality rates.30 Often a large proportion of staff working on weekends are more junior and less experienced than those who work on weekdays and are often temporary staff and not familiar with the patients or workplace.17 Primary care services are less available out of hours24 and patients often experience difficulty in accessing these services at night or over the weekend.15 Although this has been largely thought not to be the case, Meynaar et al. have shown that patients tend to be more sick out of hours which may result in greater workload for staff.32 Staff have been interviewed as to the potential reasons for failing to activate rapid response systems and found key reasons to include that they felt the clinical situation was under control on the ward, that they had instigated other investigations or actions and that they feared negative reactions from colleagues about activating the system.22 The Foster Report
I. Kolic et al. / Resuscitation 90 (2015) 85–90
highlighted that if patients are admitted at the weekend they are less likely to receive prompt treatment and more likely to die and a possible explanation is the reduced number of senior doctors.14 In view of these concerns, we agree with the proposed seven day working week as set out by the Academy of Medical Royal Colleges report ‘Seven Day Consultant Present Care’,33 with particular focus on early and increased consultant level decision making. Increasing the availability of diagnostic services out of hours may contribute to improving care out of hours. Furthermore, we could recommend the use of automated observation calculation as is being used across trusts in the United Kingdom, with the aim of decreasing scoring. Our results showed no difference in frequency of vital sign recording for higher NEW scores. Contrary to this, Hands et al.27 demonstrated more frequent observation recording with sicker patients, based on their early warning scores. These findings raise some concern. This may be explained by a lack of staff available to provide the necessary increase in recording observations. Another possible explanation is that the frequency prescribed by the EWS imposes an unachievable routine, owing to staffing levels, the availability of medical diagnostic equipment and the business of the clinical shifts.27 The clinical workload may be as an indirect effect of reduced staffing levels, or due to heavy workload and sick patients. The key strength of our study is that we are directly assessing clinical practice. Furthermore, strength is the prospective nature of the study and the completeness of data collection from ward observation charts. This allows us to highlight specific areas of care that need attention. 4.1. Limitations Firstly, this is an observational study and does not demonstrate the effect of NEWS utilisation on patient outcome. We only looked at factors affecting scoring error and compliance with NEWS. The small sample size is a limitation. Our study looks at a snapshot of compliance with the NEWS for each patient, and only looks at patients in acute medical wards, which impacts on the generalisability of our results. However the factors found were not specific to acute medical wards and would be important to other areas of hospital. None of the above limitations detract from the fact that the illustrated factors affect clinical responses to NEWS. Patients should receive the same standard of clinical care at any time of day or day of the week that they are in hospital. Attempts to establish a 24/7 inpatient service has been restricted due to limited resources and restrictions on working hours for doctors in training.15 Aylin et al. suggest that hospitals need to re-evaluate the level of services they provide out of hours.24 Our study findings may be a favourable factor towards the seven day working week as set out by the Academy of Medical Royal Colleges report ‘Seven Day Consultant Present Care’.33 This may contribute towards achieving parity for inpatient care throughout the week.34 4.2. Conclusion We have shown that clinical responses to EWS were significantly worse on weekend days. This may reflect differences in quality of care out of hours and highlights an important patient safety concern. This in turn may contribute to the increased mortality seen at these times. Authors’ contributions IK and AT involved in study conception and design; IK, SM and SC involved in data collection; ZP and IK involved in data analysis and interpretation; IK, SC and ZP involved in drafting of manuscript; IK, SM, SC, ZP and AT involved in critical revision.
89
Conflicts of interest statement None declared. 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. 2015.02.009. References 1. Smith GB, Prytherch DR, Schmidt P, et al. Hospital-wide physiological surveillance – a new approach to the early identification and management of the sick patient. Resuscitation 2006;71:19–28. 2. Kause J, Smith G, Prytherch D, et al. A comparison of antecedents to cardiac arrests, deaths and emergency intensive care admissions in Australia and New Zealand, and the United Kingdom – the ACADEMIA study. Resuscitation 2004;62:275–82. 3. Alam N, Hobbelink EL, van Tienhoven AJ, et al. The impact of the use of the Early Warning Score (EWS) on patient outcomes: a systematic review. Resuscitation 2014;85:587–94. 4. Jansen JO, Cuthbertson BH. Detecting critical illness outside the ICU: the role of track and trigger systems. Curr Opin Crit Care 2010;16:184–90. 5. NIfHaCE. NICE clinical guideline 50 Acutely ill patients in hospital: recognition of and response to acute illness in adults in hospital. London: National Institute for Health and Clinical Excellence; 2007. 6. Smith GB, Prytherch DR, Meredith P, 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. 7. Hodgetts TJ, Kenward G, Vlachonikolis IG, et al. The identification of risk factors for cardiac arrest and formulation of activation criteria to alert a medical emergency team. Resuscitation 2002;54:125–31. 8. Jones DA, DeVita MA, Bellomo R. Rapid-response teams. N Engl J Med 2011;365:139–46. 9. Staff RCoPoL. National Early Warning Score (NEWS): standardising the assessment of acute-illness severity in the NHS. Royal College of Physicians of London; 2012. 10. Ludikhuize J, Borgert M, Binnekade J, et al. Standardized measurement of the Modified Early Warning Score results in enhanced implementation of a Rapid Response System: a quasi-experimental study. Resuscitation 2014;85:676–82. 11. Martin I, Mason M, Mason D, et al. Emergency admissions: a journey in the right direction? NCEPOD; 2007. 12. Force RCoPoLAMT, Royal College of Physicians, Williams B. Acute medical care: the right person, in the right setting – first time. Royal College of Physicians; 2007. 13. Jones D, Bates S, Warrillow S, et al. Circadian pattern of activation of the medical emergency team in a teaching hospital. Crit Care 2005;9:R303–6. 14. Intelligence DF. Inside your hospital: Dr Foster Hospital Guide 2001–2011. London: Dr Foster Limited; 2011. 15. Maggs F, Mallet M. Mortality in out-of-hours emergency medical admissions – more than just a weekend effect. J R Coll Physicians Edinb 2010;40:115–8. 16. Laupland KB, Shahpori R, Kirkpatrick AW, et al. Hospital mortality among adults admitted to and discharged from intensive care on weekends and evenings. J Crit Care 2008;23:317–24. 17. Bell CM, Redelmeier DA. Mortality among patients admitted to hospitals on weekends as compared with weekdays. N Engl J Med 2001;345:663–8. 18. Gao H, McDonnell A, Harrison DA, et al. Systematic review and evaluation of physiological track and trigger warning systems for identifying at-risk patients on the ward. Intensive Care Med 2007;33:667–79. 19. Groarke JD, Gallagher J, Stack J, 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. 20. Hillman K, Chen J, Cretikos M, et al. Introduction of the medical emergency team (MET) system: a cluster-randomised controlled trial. Lancet 2005;365:2091–7. 21. Gordon CF, Beckett DJ. Significant deficiencies in the overnight use of a Standardised Early Warning Scoring system in a teaching hospital. Scott Med J 2011;56:15–8. 22. Shearer B, Marshall S, Buist MD, et al. What stops hospital clinical staff from following protocols? An analysis of the incidence and factors behind the failure of bedside clinical staff to activate the rapid response system in a multi-campus Australian metropolitan healthcare service. BMJ Qual Saf 2012;21:569–75. 23. Peberdy MA, Ornato JP, Larkin GL, et al. Survival from in-hospital cardiac arrest during nights and weekends. JAMA 2008;299:785–92. 24. Aylin P, Yunus A, Bottle A, et al. Weekend mortality for emergency admissions. A large, multicentre study. Qual Saf Health Care 2010;19:213–7. 25. Kostis WJ, Demissie K, Marcella SW, et al. Weekend versus weekday admission and mortality from myocardial infarction. N Engl J Med 2007;356:1099–109. 26. Ananthakrishnan AN, McGinley EL, Saeian K. Outcomes of weekend admissions for upper gastrointestinal hemorrhage: a nationwide analysis. Clin Gastroenterol Hepatol 2009;7:296–3020.
90
I. Kolic et al. / Resuscitation 90 (2015) 85–90
27. Hands C, Reid E, Meredith P, 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. 28. Beckett DJ, Gordon CF, Paterson R, et al. Improvement in out-of-hours outcomes following the implementation of Hospital at Night. QJM 2009;102:539–46. 29. Kerlin MP, Small DS, Cooney E, et al. A randomized trial of nighttime physician staffing in an intensive care unit. N Engl J Med 2013;368:2201–9. 30. Tarnow L, Rossing P, Nielsen FS, et al. Cardiovascular morbidity and early mortality cluster in parents of type 1 diabetic patients with diabetic nephropathy. Diabetes Care 2000;23:30–3.
31. Goddard L. Vaccinations (or lack of them) for student doctors. N Z Med J 2009;122:128. 32. Meynaar IA, van der Spoel JI, Rommes JH, et al. Off hour admission to an intensivist-led ICU is not associated with increased mortality. Crit Care 2009;13:R84. 33. Academy of Medical Royal Colleges. Seven day consultant present care: implementation considerations. Academy of Medical Royal Colleges; 2013. 34. Temple J. Time for training. A review of the impact of the European working time directive on the quality of training. Med Educ Engl 2010:9–10 (Nil volume).