Integrating Census Data with Electronic Health Records to Assess the Impact of Social Determinants of Health on Opioid Use Disorder

ObjectiveThis study investigated associations between social determinants of health (SDoH) and time to MOUD initiation across care settings, providing insights for targeted interventions to promote equitable care for OUD patients. Material and MethodsWe linked patient-level electronic health records (EHRs) from a regional integrated health system with census-tract-level SDoH from the Population Level Analysis and Community Estimates (PLACES) database. The study cohort included patients newly diagnosed with OUD (including overdose) between 2000 and 2024. We assessed temporal trends in newly diagnosed OUD cases and MOUD prescriptions, mapped the spatial correlation between OUD cases and the Area Deprivation Index, and used multivariable regression models to quantify associations between SDoH and MOUD initiation, adjusting for demographics, insurance type, and comorbidities. Analyses were stratified by care setting (emergency department, inpatient, and outpatient) to examine setting-specific associations. ResultsDuring 2000-2024, 51,521 patients with OUD or opioid overdose, among whom 14,858 (28.8%) received MOUD. OUD diagnoses peaked at 3,787 cases in 2017, then declined by 42.1% to 2,191 cases in 2024. MOUD initiation, especially buprenorphine, steadily increased throughout the study period. Geospatial analyses revealed more OUD cases in high-ADI neighborhoods. In multivariate analyses, older age and Black or African American race were associated with slower MOUD initiation. In the stratified analyses by care setting, significant associations between SDoH and MOUD initiation were primarily observed in the outpatient setting, though effect sizes were modest. ConclusionsIntegrating neighborhood-level SDoH with EHRs can uncover care-setting-specific disparities in treatment initiation and identify neighborhoods with unmet treatment needs.