The Management of Depression and Anxiety in Primary Care: Examining Predictors of Adherence to a Psychopharmacological Collaborative Care Management Program for Veterans.

Abstract

Collaborative care management (CoCM) is an evidenced based approach to psychiatric treatment in primary care, yet literature examining factors associated with program adherence is lacking. This study analyzed predictors of adherence to a CoCM model of psychopharmacological treatment of depression and anxiety in primary care by conducting a retrospective cohort analysis on Veterans referred to a large VA Medical Center's CoCM program over an 18-month period. Baseline characteristics, symptomatic assessments, and covariates of interest were collected. For the primary outcome, the association between covariates and programmatic completion were analyzed. Secondary analyses assessed improvements in psychiatric symptoms. A total of 757 Veterans with depressive or anxiety disorders were included, and 256 completed the CoCM program. Baseline covariates associated with differences in completion rates included the following: age, contact with psychology prior to referral, baseline PHQ-9, baseline GAD-7, and a number of comorbid psychiatric/substance abuse covariates. After controlling for baseline differences, age remained a significant positive predictor of completion (OR 1.019, 95% CI 1.008‒1.030) and cannabis use a significant negative predictor (OR 0.507, 95% CI 0.275‒0.934). Both early improvement in PHQ-9 (OR 1.864, 95% CI 1.210‒2.872) and GAD-7 (OR 1.762, 95% CI 1.154‒2.691) scores were positive predictors. Secondary analyses showed that programmatic completion was associated with greater reductions in psychiatric symptoms. Results identified a number of modifiable parameters associated with differences in completion rates and greater symptomatic reduction for those who complete the program. Additional studies should be conducted examining interventions to optimize CoCM programs by supporting positive predictors while minimizing negative predictors.