Impact of the emergency department streaming decision on patients' outcomes

BACKGROUND: Streaming occurs in emergency department (ED) to reduce crowding, but misallocation of patients may impact patients' outcome. AIM: The study aims to determine the outcomes of patients misallocated by the ED process of streaming into likely admission or discharge. METHODS: This is a retrospective cohort study, at an Australian, urban, tertiary referral hospital's ED between January 2010 and March 2012, using propensity score matching for comparison. Total and partitioned ED lengths of stay, inpatient length of stay, in-hospital mortality and 7- and 28-day unplanned readmission rate were compared between patients who were streamed to be admitted against those streamed to be discharged. RESULTS: Total ED length of stay did not differ significantly for admitted patients if allocated to the wrong stream (median 7.6 h, interquartile range 5.7-10.6, cf. 7.5 h, 5.3-11.2; P = 0.34). The median inpatient length of stay was shorter for those initially misallocated to the discharge stream (1.8 days, 1.1-3.0, cf. 2.4 days, 1.4-3.9; P < 0.001). In-hospital mortality and 7- and 28-day readmission rates were not adversely affected by misallocation. When considering patients eventually discharged from the ED, those allocated to the wrong stream stayed in the ED longer than those appropriately allocated (5.2 h, 3.7-7.3, cf. 4.6 h, 3.3-6.4; P < 0.001). CONCLUSION: There were no significant adverse consequences for an admitted patient initially misallocated by an ED admission/discharge streaming process. Patients' discharge from the ED was slower if they had been allocated to the admission stream. Streaming carries few risks for patients misallocated by such a process.