Implementing and Evaluating a Standard of Care for Clinical Evaluations in Emergency Psychiatry

Abstract Background Psychiatric presentations are common in emergency departments (EDs), but the standard of care for treatment remains poorly defined. We introduced standards for emergency psychiatric evaluations that included obtaining collateral information, writing a safety plan for discharging patients, identifying the next best provider, and alerting that provider to the patient's visit. Objective We sought to demonstrate the feasibility and clinical impact of implementing standards for emergency psychiatric evaluations. Methods To evaluate feasibility, physicians attested to completion in the electronic health record. To evaluate the effect on clinical outcomes, we compared admission rates, 30-day return rates, and median length of stay from a 4-month pre-implementation period to a 4-month post-implementation period. Data were extracted from a quality-improvement database. Results There were 1896 patient encounters in the pre-implementation period and 1937 encounters post-implementation. Pre-and post-cohorts were similar demographically. Collateral was obtained for 1035 (86%) encounters, a written safety plan was completed for 793 (77%) eligible patients, the next-best provider was identified for 1094 (91%), and that provider was contacted for 837 (70%). There was no difference from pre to post periods in admission rates (17% vs. 18%; p  = 0.36), median length of stay (13.3 ± 0.6 vs. 12.5 ± 1.4; p  = 0.35), or 30-day return rates (15% vs. 16%; p  = 0.66). Conclusions This standard work for emergency psychiatric evaluations was feasible even in a highly acute patient population. However, the benefits of this intervention are less clear. We question the utility of prevailing metrics in emergency psychiatry.