Implementing a Virtual Collaborative Behavioral Health Care Program in an Obstetrics and Gynecology Practice [ID: 1376661]

Abstract

INTRODUCTION: More women require behavioral health services than exist psychiatrists to treat them. In order to meet this need, the collaborative care model (CoCM) was proposed in the 1990s to extend the reach of behavioral health services. Relatively little is known about the applicability of the CoCM to obstetrics. We instituted a virtual collaborative behavioral health care program in an obstetrics and gynecology practice in our health system. METHODS: This was an IRB-approved retrospective chart review of a collaborative care program using the model initially proposed in the IMPACT study. Patients presenting to an obstetrics and gynecology office were assessed for depression using validated tools (PHQ-9). Women who screened positive were referred to a behavioral health care manager (BHCM) embedded in the practice. The BHCM conducted a behavioral health assessment, provided short-term psychotherapy, coordinated referral to a higher level of care if necessary, maintained a registry of patients, and worked with a supervising psychiatrist who provided medication recommendations. Our primary outcomes were the incidence of screenings completed, the number of contacts with the BHCM, and the number of cases that underwent review by a psychiatrist. RESULTS: Between February 2020 and February 2022, 166 women were enrolled in the CoCM out of 213 referred by the obstetrician–gynecologist (78% enrollment rate); participants were enrolled for an average of 177 days. Ninety participants (n=166, 54%) had at least one PHQ-9 completed and 81 participants (n=166, 49%) had at least one GAD-7 completed. One hundred six (n=166, 64%) patients had therapy for an average of five visits. One hundred four (n=166, 63%) participants participated in care coordination. Thirty-four (n=166, 20%) cases were reviewed by a psychiatrist. CONCLUSION: Our results indicate the high utilization of collaborative care services in this context, and moderate feasibility with only half of enrolled participants receiving validated assessments. Future studies should examine effectiveness and the components of the model that are most beneficial.