Embedding Primary Care Clinic Within Community Mental Health Services Reduces All-Cause Hospitalizations and Emergency Department (ED) Visits for Patients with Severe Mental Illness (SMI) And Cardiovascular Disease (CVD)

Friday, November 11, 2022

2:15 PM – 3:45 PM

Location: Room 206-207

Presenting Author(s)

Joy J. Choi, MD (she/her/hers)

Assistant Professor
University of Rochester Medical Center
Rochester, New York

PDF

Co-Author(s)

Daniel Maeng, PhD

Assistant Professor
University of Rochester Medical Center
Rochester, New York
Hochang Lee, MD, FACLP (he/him/his)

John Romano Professor and Chair
University of Rochester Medical Center
Rochester, New York
Telva Olivares, MD

Associate Chair of Diversity, Inclusion, Culture and Equity
University of Rochester Medical Center
Rochester, New York
MW

Marsha Wittink, MD, MBE

Medical Director Medicine in Psychiatry Unit
University of Rochester Medical Center
Rochester, New York

Abstract: BACKGROUND
Despite persistent mortality gap attributed to CVD in patients with SMI, access to primary care remains challenging (Lawrence, 2010). Different models of enhanced primary care have been studied to improve access for patients with SMI with varying degrees of success. One potential solution – “reverse integration model” – is to co-locate a primary care clinic within community mental health service, but little data is available on the effectiveness of this model for SMI (Ward 2017). Medicine in Psychiatry (MIPS) is a SMI-focused primary care clinic co-located alongside community mental health services at University of Rochester Medical Center. We examined all-cause acute care utilization pattern in MIPS patients with SMI and CVD, compared to non-MIPS primary care.

METHODS
A retrospective cohort analysis using multi-payer claims data (10/1/2015 through 6/30/2017) identified 303,824 patients who had diagnosis of SMI (ICD-10: F20-25, F30-31) and CVD (I00-99) in 12 Western New York counties. Inclusion criteria included a 6-month washout period of no primary care, to account for possible treatment effect of non-MIPS primary care received prior to treatment at MIPS. Final analysis included 3,646 subjects with 372 treated at MIPS. At baseline, the MIPS group was associated with younger age (51.9 vs. 52.8, p</em>=0.06). MIPS and non-MIPS groups did not significantly differ in baseline rates of inpatient hospitalization, ED visits, Charlson-Quan Comorbidity Index, or substance use disorder. A difference-in-difference approach was used to identify the MIPS treatment effect on acute care utilization rate.

RESULTS
In the first 6 months after the washout period, patients treated at MIPS had lower per-member-per-month (PMPM) rates of all-cause ED visit (0.62 vs 0.74, p=0.06) and inpatient admission (0.14 vs 0.17, p=0.1) that did not rise to the level of significance. After 6 months, the difference persisted and became highly significant (ED: 0.51 vs. 0.64, p=0.041; inpatient: 0.09 vs 0.13, p</em>=0.068).

DISCUSSION
Treatment at MIPS is associated with significant reduction in all-cause ED visits and inpatient hospitalizations in patients with SMI and CVD. MIPS is a form of “reverse integration” that allows for warm-handoffs and clinical collaboration between primary care and mental health providers to improve access to comprehensive care. Strengths of this analysis include large control group size, a wide geographical cohort, and use of all-claims data that does not restrict analysis to a specific healthcare organization. Generalizability is limited due to scarcity of reverse co-located primary care clinics specific for SMI. Well-powered, additional studies are needed to confirm the findings of this study and their implications on clinical outcome.

CONCLUSION
Our findings support the growing trend in enhancing primary care for SMI by co-locating primary care clinic in mental health settings to improve access and management of chronic CVD among SMI population.

REFERENCES
Lawrence 2010:doi:10.1177/1359786810382058
Ward 2017:doi:10.1176/appi.focus.20170011