CLINICIAN’S CAPSULE
What is known about the topic?
Early warning scores reduce morbidity and mortality by identifying patients at risk for deterioration on the medical wards.
What did this study ask?
This study sought opinions from emergency department (ED) staff on the use of early warning scores in the ED.
What did this study find?
Emergency doctors and nurses did not value early warning scores despite quantitative data showing efficacy in the ED.
Why does this study matter to clinicians?
We gleaned insight into how to implement a computer-based early warning score into the ED to reduce morbidity and mortality from septic shock.
INTRODUCTION
Early warning scores (EWS) identify patients at risk of critical deterioration or deathReference Smith, Prytherch, Meredith, Schmidt and Featherstone 1 , Reference Tam, Xu and Kwong 2 and are derived by scoring derangements in commonly measured physiological parameters. Many patients with elevated EWS have sepsis, a common and time-sensitive emergency department (ED) presentation and a major cause of morbidity and mortality.Reference Martin, Priestap and Fisher 3 , Reference Gaieski, Mikkelsen and Band 4 Sepsis is sometimes not identified during ED triage, leading to preventable death.Reference Keep, Messmer and Sladden 5 Use in both medical and surgical wards is recommended, and evidence suggests that EWS are useful in the ED.Reference Keep, Messmer and Sladden 5 , Reference Churpek, Snyder and Han 6 The Hamilton Early Warning Score (HEWS; Appendix 1) is a predictor of critical in-patient eventsReference Tam, Xu and Kwong 2 ; an elevated HEWS at the time of ED triage predicts sepsis.Reference Skitch, Tam and Xu 7 In 2015, we integrated an automatic calculation of HEWS into a digital triage and ED charting system, training nurses and physicians through meetings and in-services. This study assessed how ED staff perceived HEWS post-implementation.
METHODS
Participants
We intentionally sampled ED registered nurses (both triage and bedside), resident physicians, and emergency medicine attending physicians. Participation was voluntary, verbal consent was obtained, and no incentives were offered. Institutional ethics approval was obtained.
Data collection
Our interview guide was informed by both published literature and local experience. The guide (Appendix 2) was pilot tested with non-participating physicians and nurses and refined based on feedback. A single, trained interviewer recorded semi-structured interviews, which were subsequently transcribed by a trained medical transcriptionist and checked for errors by the original interviewer.
Data analysis
We analyzed data using a constructivist grounded theory,Reference Charmaz 8 reviewing previous literature on implementation of clinical decision tools and risk stratification scores as sensitizing concepts. We used a constant comparative technique extracting themes as we went, constantly seeking out new or divergent themes as we collected data.Reference Kolb 9
Transcripts were reviewed by two qualitative researchers (BB and TMC) and coded into thematic coding trees, developing a codebook with definitions and relevant exemplary quotes. These investigators frequently met to refine the coding structure until consensus was reached. To ensure rigour, a third investigator (SS) listened to the interviews and conducted an audit of the trial; participants also reviewed the final themes and codes in a member check.
RESULTS
Twelve participants (five nurses, three residents, and four attending physicians), ranging in experience from 1 to 30 years, were recruited and interviewed, yielding 241 minutes of interview tape (median 27 minutes, range 5-62 minutes) and 98 transcript pages. We identified five themes: knowledge of sepsis and HEWS, utility of HEWS in emergency triage, utility of HEWS at the bedside, utility in communicating acuity and deterioration, and feasibility and accuracy of data collection. We then identified nine barriers and seven enablers for the use of EWS in the ED (Table 1). Our data analysis reached a point of sufficiency after 12 interviews.
Table 1. Barriers and enablers
ED=emergency department; HEWS=Hamilton Early Warning Score; RP=resident physician; SP=staff physician; RN=registered nurse.
Feasibility and accuracy of HEWS data gathering
Vital sign accuracy was generally thought to be high, although many acknowledged that certain vital signs (particularly temperature and respiratory rate) are often estimated if no specific clinical concern would prompt accurate measurement.
Knowledge of sepsis and HEWS
Participants were aware of vital sign changes in sepsis and the purpose of the HEWS score and were comfortable interpreting it. Vital signs were valued in the detection of sepsis, with temperature, heart rate, and respiratory rate being the most highly valued. The details of the HEWS scoring were not known by the participants, but “cut scores” and trend patterns were well known by most.
Utility in the triage area
Participants felt that the HEWS score was not helpful for the assignment of a triage level because of strong clinical assessment skills and gestalt: “Sick is relatively common for us,” said one physician (SP01). A registered nurse (RN04) stated, “We are able to identify what sick looks like and what abnormal vitals are.”
Triage nurses universally felt that their clinical judgment already accounted for abnormal vital signs and that they had the gestalt abilities to contextualize abnormalities: “My critical thinking should come into play when I look at vital signs, so I really don’t need an additional tool to tell me that vitals are off” (RN04). RNs reported that their decision to upgrade or downgrade a patient’s triage level was based on vital signs, rather than the HEWS. One participant (RN03) stated, “I think when somebody brings you a patient and they are like, ‘oh, they’ve got a HEWS score of six,’ then there is an eye roll.”
Participants perceived that the sensitivity and specificity of the HEWS fared poorly. “The patient having a [heart attack] or a stroke may have pristine vital signs,” so HEWS is not “designed as a sorting tool, it is not designed to differentiate between which of these 30 patients you have to see next” (SP01). They said that a high HEWS was often a false alarm; they believe many abnormal vital signs could be explained with a clinical rationale such as a low oxygen saturation in a patient with chronic lung disease or a high heart rate in a patient with atrial fibrillation. Triage nurses also expressed that a low HEWS score may be falsely reassuring, such as in supraventricular tachycardia in which only the heart rate is abnormal, but the patient requires prompt care.
Utility of HEWS for bedside nurses
Nurses and physicians believed that HEWS was of limited value in the acute section of the ED because of their experience and skill in recognizing and caring for very sick patients: “We have highly experienced and highly trained nurses to know what is wrong with the patient” (SP04). Some nurses and physicians believed that the EWS might be of greatest benefit in the low-acuity (fast-track) area of the ED, where subtle deteriorations may not be recognized quickly: “I think the place where it is likely of the most use is in [rapid assessment zone]” (SP01).
Utility in communicating acuity and deterioration
Participants universally rejected HEWS as part of the ED lexicon for communicating: “I don’t think I have ever heard anybody have a conversation and say a HEWS score” (RN02); “HEWS is never at the forefront of my conversation with a physician” (RN04). Physicians also expressed a lack of utility knowing just the score: “I want to know why the score is that way . . . the score itself isn’t helpful to me” (SP03).
RNs generally found HEWS helpful when calling senior medical residents (SMRs) or the most responsible patient for admitted patients boarding in the ED, though many surgical specialties were thought not to appreciate or understand the score: “the SMR will show up right away [and say] ‘What did I miss?’” (RN01).
DISCUSSION
Our results indicate unsuccessful ED implementation of an early warning score currently well accepted in other areas of our hospital. Despite implementing training and technology to integrate the score into triage and bedside functions, there appears to be a “culture clash.”
Three beliefs formed the basis of this rejection: 1) ED practitioners are expert at detecting patients with abnormal vital signs who are at risk of deterioration; 2) the standardized language of HEWS was unnecessary because of an existing lexicon to communicate acuity between doctors and nurses; and 3) they did not understand the science behind the score, nor the quality concerns of department leadership.
Clinicians in our study believed they are experts in determining if a patient is “sick or not sick.” There is some validity to this construct. There is evidence to support that experienced emergency physicians and nurses can rapidly predict the disposition of patients with a single look.Reference Sibbald, Sherbino and Preyra 10 , Reference Rohacek, Nickel, Dietrich and Bingisser 11 ED physicians can predict acuity based on a triage note 75% of the time.Reference Cabrera, Thomas and Wiswell 12 Despite strong intuitive abilities, ED patients can be at risk for unrecognized deterioration and death.Reference Goulet, Guerand and Bloom 13 , Reference Bilben, Grandal and Søvik 14 A failure to recognize deterioration is the second-highest cost for hospital insurers in Canada. 15
While an EWS can provide a succinct method of communicating a patient’s condition on a ward, this advantage was not felt to be too germane for the ED. Callen found that hospital subcultures can view new information technology with hostility,Reference Callen, Braithwaite and Westbrook 16 resulting in an apparent “culture clash.”
LIMITATIONS
Our study was conducted around a specific EWS implementation that could limit the transferability of our findings. The authorship team consists of physicians, and the single interviewer was a resident physician from the same institution that might have influenced our interpretation of the data. There might also have been a tendency for certain types of volunteers to engage in this study, who might not have been representative of all clinicians.
CONCLUSION
ED clinicians believe that their expertise, an existing lexicon related to critically ill patients, and the current triage score lessens the utility of HEWS that is sufficient for patient safety. Nurses find EWS useful to communicate with admitting services. Effective integration of an EWS into the ED would require broader engagement of front-line clinicians.
Acknowledgements
We thank the study participants for their time and participation. All authors contributed to the study design, analysis, and writing of this manuscript.
Competing interests
The authors have no financial conflicts of interest to declare.
SUPPLEMENTARY MATERIAL
To view supplementary material for this article, please visit https://doi.org/10.1017/cem.2018.392
