Abstract 13686: An Implementation Science Informed, Patient Activation Intervention for Depression Treatment in Coronary Heart Disease Patients: Impact on Clinician Referral and Prescribing Behaviors

Abstract

Intro: Implementation strategies are integral to addressing gaps in AHA depression treatment guidelines in coronary heart disease (CHD) patients. Methods: From 2019 to 2023, we randomized 12 Columbia University affiliated primary care and cardiology clinics to timing of receiving a multi-level intervention. We enrolled separate pre and post EHR eligible patients/site (≥21 years old, English or Spanish speaking, CHD ICD10 and upcoming cardiologist/PCP index visit) with Patient Health Questionnaire(PHQ)9≥10 and not seeing a psychiatrist. Pre-implementation patients and their providers/sites received usual care. Post-implementation, a theory-informed intervention included problem-solving meetings for mental health social workers (MHSW), cardiologist/PCP [“provider”] education and a patient psychoeducation, activation, and treatment selection tool (with patient treatment preferences sent to MHSW/providers via EPIC). Blinded physicians extracted index visit pre vs. post provider behavior (i.e., documented CHD risk factor or mental health counseling, therapy/cardiac rehab referral, and/or antidepressant prescribing vs. none). We conducted a pre-post analysis using descriptive statistics and thematically analyzed any MHSW/provider responses to EPIC reports. Results: Of 10,625 EHR eligible CHD patients; 1759 patients were screened, 304 (16.5%) eligible and 253 (82.9%) enrolled; 52.0% were female, 56.1% Hispanic, 14.6% Black; mean age was 67.8 years old, PHQ9 13.1. Overall, all (n=71) intervention providers received EPIC messages with 67.8% of patients selecting ≥1 treatment; 29.7% of reports prompted a MHSW/provider response (themes: thinking, talking, action) ; more cardiologists acknowledged receipt/relayed plans than PCPs as did providers caring for commercially insured (vs. Medicaid) patients; 25% of providers optimized treatment pre- vs. 40% post-implementation. Conclusion: A multi-level strategy centered around patient activation may marginally improve provider guideline adherence. Key limitations are pre-post design and descriptive analyses without adjustment for temporal trends. Our ongoing study will adjust for temporality and examine impact on depressive symptoms and patient behavior.