Redesigning an emergency department for interprofessional teamwork : a longitudinal evaluation of the impact on patient flow and team behaviour

Abstract: Objective: Crowding in emergency departments (EDs) has negative effects on patients and staff, whereas effective teamwork in healthcare has positive effects. Triage interventions and streaming of specific patient categories are common strategies used by EDs to reduce the negative effects of crowding, but few studies have evaluated interprofessional teamwork as a strategy to reduce ED crowding. This thesis evaluated the redesign of an ED for interprofessional teamwork with the aim of studying the impact on patient flow and team behaviour. Methods: The project was conducted at an adult ED, where teamwork modules replaced the triage, fast track, and main ED sections in November 2014. Study I, II, and III were quantitative before- and-after studies using patient data, whereas study IV collected qualitative and quantitative data from health professionals for four years. Study I and II used registry data from May 2012 to November 2015. Study I included all patient arrivals on weekdays from 8 am to 9 pm for three one-year periods and compared the first year of teamwork to two previous triage processes, nurse-led triage and physician-led triage. Study II included patients who presented limb injuries or back pain to the ED, where the first year of teamwork was compared to a previous period, when ambulant low acuity patients were streamed to the fast track. An equal number of non-orthopaedic presentations were also included to control for the impact on other patient categories. The waiting time to physician assessment and ED length of stay were outcome measures in both studies. We used multiple linear regression analysis to adjust the ED length of stay for differences in patient and background characteristics between the periods. Study III included all patients 80 years or older arriving on weekdays during a pilot period in 2016, when a teamwork module was dedicated to older patients, and a corresponding period in 2015 with only regular teamwork modules for mixed age groups. The outcome measures were the ED length of stay and the total hospital admission rate within seven days. Study IV included triangulated data from three staff sources; structured observations of team behaviours in June 2016 and June 2018, semi-structured interviews in June 2018, and a questionnaire of the perceived workload, collaboration, and patient satisfaction repeated from October 2014 through June 2018. Results: Study I included 185 806 patient arrivals. The crude median ED length of stay was shortest for the teamwork period, 228 minutes (95% CI: 226.4 to 230.5) compared to 232 minutes (95% CI: 230.8 to 233.9) for the nurse-led triage period, and longest for the physician-led triage period, 250 minutes (95% CI: 248.5 to 252.6). The adjusted ED length of stay for the teamwork period was 16 minutes shorter than for the nurse-led triage period (p<0.001), and 23 min shorter than for the physician-led triage period (p<0.001). The median waiting time to physician assessment was 74 minutes (95% CI: 73 to 75) for the teamwork period, 116 minutes (95% CI: 114 to 118) for the nurse-led triage period, and 56 minutes (95% CI: 55 to 57) for the physician- led triage period. Study II included 22 551 orthopaedic patient presentations. In the fast track period, 70% were low acuity patients and 70% of these were dispositioned from the fast track. The crude median ED length of stay was shorter for the teamwork period compared to the fast track period, -13 minutes (95% CI: -18 to -8). The difference of the adjusted ED length of stay was -23 minutes (95% CI: -27 to -19). The mean waiting time to physician assessment was also shorter in the teamwork period, -57 minutes (95% CI: -60 to -54) compared to the fast track period. For the additional 21 780 non-orthopaedic presentations, the adjusted ED length of stay was also shorter in the teamwork period, -20 minutes (95% CI: -25 to -16), as was the mean waiting time to physician assessment, -30 minutes (95% CI: -33 to 26). Study III included 4 584 presentations by patients 80 years or older and there was no difference in patient characteristics between the periods. In the intervention period, 27% (n=634) of the patients received care in the geriatric module, and the remaining patients in standard teamwork modules. The total hospital admission rate within seven days was lower in the intervention period, compared to the control period. However, the ED length of stay was longer in the intervention period. Study IV included 50.5 hours of structured observations in 2016, when fidelity was observed for four of five key team behaviours. In 2018, fidelity remained only for one team behaviour and observation saturation was reached after 37.5 hours. Qualitative content analysis of 18 interviews in 2018 exposed several issues of the staff and context fidelity, for instance, team training and feedback were discontinued. In the questionnaire, positive ratings approximately doubled for items relating to the work experience when teamwork was introduced. However, in 2018 the ratings had deteriorated to pre-implementation levels. Conclusions: Effective interprofessional teamwork reduced the waiting time to physician assessment and the ED length of stay for the patients. However, the fidelity to the teamwork process decayed over time and the positive outcomes were not sustained.