Baseline Characteristics Associated With Referrals From a Primary Care Mental Health Collaborative Care Model to Specialty Mental Health Services.

Objective

Collaborative care models (CoCMs) are modalities for treating mental health conditions in primary care. One such iteration of CoCM, antidepressant monitoring (ADM), is a pharmacologic treatment modality for the management of depression and anxiety. ADM programs have established efficacy, yet little is known about ideal patient selection or approaches related to program retention. The objective of this study was to address this need by examining predictors of referral from a CoCM ADM program to higher levels of psychiatric services.

Methods

A retrospective cohort analysis was conducted on individuals enrolled in the Tampa Veterans' Affairs ADM program over 18 months (from June 4, 2018, through December 4, 2019). Data collected included information related to referral to a higher level of service, as well as baseline information and covariates of interest. Primary analysis was conducted utilizing a multivariable logistic regression model to evaluate whether baseline characteristics were associated with differences in referral rates to higher-level services.

Results

A total of 757 veterans were included in the analyses, with 131 (17.31%) referred on to a higher level of service for specialty psychiatric care. Multivariable modeling showed the following covariates to be associated with higher rates of referral to specialty psychiatric services: baseline 9-item Patient Health Questionnaire scores, sleep issues at the time of enrollment, alcohol use disorder, and cannabis use disorder.

Conclusions

Results show low rates of referral overall but identify a number of baseline characteristics associated with higher referral rates to specialty psychiatric services. Further research is needed, including prospective work and studies examining proactive interventions to limit required referrals to specialty mental health services. Prim Care Companion CNS Disord 2026;28(2):25m04149. Author affiliations are listed at the end of this article.