Proactive Integration of Mental Health Care in Hospital Medicine: PRIME Medicine

Background: Proactive consultation-liaison (C-L) psychiatry has been shown to reduce hospital length of stay (LOS), increase psychiatric C-L consult rate, and improve hospital staff satisfaction. Nursing attrition has not been studied in relation to proactive C-L. Objective: Our primary aim in evaluating the proactive C-L service called Proactive Integration of Mental Health Care in Medicine (PRIME Medicine) is to analyze change in LOS over 10 months using historical and contemporary comparison cohorts. As secondary aims, we assess change in psychiatric consultation rate, time to consultation, and change in nurse attrition. Methods: PRIME Medicine was implemented on 3 hospital medicine units as a quality-improvement project. Team members systematically screened patients arriving to assigned units for psychiatric comorbidity. Identified patients were reviewed with hospitalist teams and nurses with the goal of early intervention. Results: Including historical and contemporary comparison cohorts, the mean sample age was 62.4 years (n = 8884). Absolute LOS was unchanged, but difference-in-difference analysis trended toward reduced LOS by 0.16 day (P = 0.08). Consultation rate increased from 1.6% (40 consults) to 7.4% (176 consults). Time to consultation was unchanged (4.0-3.8 d). Annual per-unit nursing turnover increased from 4.7 to 5.7 on PRIME units but from 8.5 to 12.0 on comparison units. Nurses citing "population" as the reason for leaving decreased from 2.7 to 1.7 on PRIME units but increased from 1.5 to 4.5 on comparison units. PRIME Medicine led to increased consultation rate, and our unit-wide outcomes provide a conservative estimate of effect. Factors that may have influenced effect size include our cohort's advanced age, considerable emergency department boarding times, increasing proportion of patients discharged to skilled nursing facilities, and concurrent LOS-reduction initiatives on all units. The favorable trends in nursing attrition on PRIME units may be explained in part by our prior finding that PRIME Medicine was associated with enhanced nursing satisfaction. Conclusions: While PRIME Medicine had no more than a modest effect on LOS, it was associated with a markedly increased psychiatric consult rate and favorable trends in nursing retention. This analysis highlights important factors that should be considered when implementing and determining value metrics for a proactive C-L service.