Subtypes and Service Utilization Patterns Among Patients with Opioid Use Disorder at a Community Health Center in a Medically Underserved Urban Area

Background and Purpose: Substance use disorders (SUDs) and mental health disorders (MHDs) are prominent public health concerns relevant to the social work discipline. These groupings of conditions are associated with increased morbidity and mortality, frequently co-occur, and are often diagnosed and treated at community health centers. Opioid use disorder (OUD), one of the more common SUDs, is associated with multimorbidity and is a leading cause of overdose deaths. To better inform integrated treatment for OUD, this study examined multi-year electronic medical records from a community health center in a diverse, high-need community. These data were used to identify client subtypes based on demographic characteristics, healthcare plans, and SUD/MHD diagnoses, and how subtypes differed in their service utilization.

Methods: The dataset included adults 18-65 years old (N=6,780) who were screened for substance use while receiving treatment (2015-2021) at a community health center in a Northeast state. For this analysis, we limited the sample to individuals with an OUD diagnosis (n=705). Using univariate statistics, we examined demographic characteristics including age, race/ethnicity, sex, healthcare plans, psychiatric diagnoses, and service utilization sessions (individual psychotherapy, group therapy, and psychiatric). Using Bayesian Information Criteria, we performed a cluster analysis to examine potential groupings. A one-way analysis of variance (ANOVA) with Bonferroni and chi-square tests with adjusted standardized residuals were conducted to identify characteristic differences between clusters, and service utilization differences by cluster.

Results: The sample primarily identified as Black or African American (n=282, 40.0%), male (n=498; 70.6%), and used Medicare/Medicaid for clinical services (n=527; 74.8%). Over a quarter of the sample (n=196; 27.8%) had a MHD and a second SUD diagnosis. The most prevalent co-occurring diagnoses for MHDs were depressive (n=137, 19.4%) and bipolar (n=90; 12.8%) disorders; and for SUDs were cannabis use (n=224, 31.8%) and cocaine use (n=223; 31.6%) disorders. The cluster analysis had an average silhouette of 0.5, indicative of good clustering. Six clusters were revealed and labeled: Cluster 1. “Medicare/Medicaid healthcare plan with SUD needs”, Cluster 2. “Private pay and charity care healthcare plan and younger with cocaine use disorder needs”, Cluster 3. “Medicare/Medicaid and other publicly-funded healthcare plans with mood disorder needs”, Cluster 4. “Private healthcare plan with low co-occurring needs”, Cluster 5. “Other publicly-funded healthcare plan and male with cannabis use disorder needs”, and Cluster 6. “Medicare/Medicaid healthcare plan and older with MHD needs”. All ANOVAs for service utilization differences by cluster were significant (p<.001). In post-hoc tests, Cluster 3 had the highest number of all session types, whereas Clusters 6, 4, and 5, had the lowest number of individual psychotherapy, group therapy, and psychiatric sessions, respectively.

Conclusions and Implications: This study identified patient subtypes and their service utilization patterns at a community health center in a diverse, high-need community. The co-diagnosed conditions highlight the importance of comprehensive and integrated treatment. Importantly, patients’ healthcare plans, a socioeconomic factor that impacts access to care, plays a critical role in distinguishing treatment needs and utilization. These findings can guide assessment and treatment protocols tailored to common patient presentations and improve quality of care.