Baseline screening tools as indicators for symptom outcomes and health services utilization in a collaborative care model for depression in primary care: a practice-based observational study

Abstract

OBJECTIVE: Within a practice-based collaborative care program for depression, we examined associations between positive baseline screens for comorbid mental and behavioral health problems, depression remission and utilization after 1 year. METHODS: This observational study of 1507 depressed adults examined baseline screens for hazardous drinking (Alcohol Use Disorders Identification Test score ≥ 8), severe anxiety (Generalized Anxiety Disorder 7-item score ≥ 15) and bipolar disorder [Mood Disorders Questionnaire (MDQ) positive screen]; 6-month depression remission; primary care, psychiatric, emergency department (ED) and inpatient visits 1 year postbaseline; and multiple covariates. Analyses included logistic and zero-inflated negative binomial regression. RESULTS: At unadjusted baseline, 60.7% had no positive screens beyond depression, 31.5% had one (mostly severe anxiety), 6.6% had two and 1.2% had all three. In multivariate models, positive screens reduced odds of remission versus no positive screens [e.g., one screen odds ratio (OR) = 0.608, p = .000; all three OR = 0.152, p = .018]. Screening positive for severe anxiety predicted more postbaseline visits of all types; severe anxiety plus hazardous drinking predicted greater primary care, ED and inpatient; severe anxiety plus MDQ and the combination of all three positive screens both predicted greater psychiatric visits (all p < .05). Regression-adjusted utilization patterns varied across combinations of positive screens. CONCLUSIONS: Positive screens predicted lower remission. Severe anxiety and its combinations with other positive screens were common and generally predicted greater utilization. Practices may benefit from assessing collaborative care patients presenting with these screening patterns to determine resource allocation.