Improving Outcomes of Depression in Primary Care

OBJECTIVE(S): The primary objective of the study was to determine the impact of a low-intensity, care management intervention on depression treatment outcomes of patients in a VA primary care setting. Secondary objectives were to determine the extent to which primary care clinicians adhere to major depression treatment guidelines, the extent to which the intervention impacts healthcare utilization, and to examine the impact of provider and patient covariates (including pain and adherence) on depression treatment response PLAN: Randomized controlled trial. METHODS: All clinicians first participated in the MacArthur Depression Education Program (DEP). After stratification by discipline and site, clinicians were randomized to receive the Depression Decision Support (DDS) intervention versus usual care (which included an on-site mental health team). Patients were nested within clinician intervention status. DDS intervention clinicians received serial reports of depression severity scores and other clinical data for their enrolled patients. The DDS care manager made one telephone contact with each intervention patient, and invited patients to attend a depression group education session. The DDS team monitored patient depression severity over time, and targeted additional time and effort towards patients not showing improvement. Additional DDS intervention included making further recommendations to primary care clinicians, and in some cases, a one-time consultation with the DDS psychiatrist.

Potential subjects were identified from lists of patients due for appointments with primary care clinicians, and were contacted for telephone screening. Patients were enrolled if they had Patient Health Questionnaire (PHQ-9) scores >10 or a Hopkins Symptom Checklist (SCL-20) score of > 1.0 at a subsequent in-person interview. There were no restrictions by age or sex. Patients were excluded if they had psychotic symptoms, dementia, serious suicidal ideation, very severe depression (PHQ-9 >25), or had been treated by mental health clinicians within the previous six months. The intervention lasted 12 months. Depression symptom severity (SCL-20) and health related quality of life (SF-36V) at 6 and 12-months were primary outcomes. Primary analyses employed mixed-model repeated measure analyses of covariance, which include a random effect for clinician and fixed effects for time and intervention status.

FINDINGS TO DATE: Five Portland VA clinic sites and 41 clinicians participated. 3,103 patients were screened, and 402 patients were enrolled. Three hundred seventy-five met final eligibility criteria. There were no adverse events. SCL-20 scores improved in both groups over 12 months (slope: -.382; 95% CI -.488, -.276), but there was no effect of the intervention on SCL-20 or Short Form-36V compared to usual care. There was greater initial improvement in PHQ-9 for intervention patients (p=.030); however, this effect diminished over time. At 12 months, intervention patients reported greater satisfaction (p=.002), and were more likely to have received antidepressants. To date, 10 manuscripts supported by the project have been published. One manuscript examining associations between stressful live events and depression outcomes has been submitted for publication.