Follow-Up Patterns at a Low-Threshold Mobile Medical Unit Providing Opioid Use Disorder Care in an Urban Setting: A Group-Based Trajectory Modeling Approach
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background The University of Illinois Chicago (UIC) Community Outreach Intervention Project (COIP) established a low-threshold mobile medical unit that dispensed medications to treat opioid use disorder (MOUD) to individuals with high need. Knowledge of mobile unit utilization is limited to binary retention metrics. This study identified and characterized patterns of patient follow-up at a mobile unit dispensing MOUD. Methods Group-based trajectory modeling (GBTM) identified patterns of follow-up observed at the mobile unit. The data source was an internal database of every UIC COIP mobile unit patient encounter, maintained by clinicians. Individuals visiting the mobile unit for MOUD services while the UIC mobile unit carried buprenorphine onsite were included. In the GBTM, follow-up data for each patient over 12 months after their first visit were aggregated to yield groups of individuals with similar follow-up patterns. After GBTM, a multinomial logistic regression was used to compare demographic, substance use, and treatment characteristics between identified follow-up pattern groups. Results Of the 964 eligible individuals, 429 had more than one mobile unit visit and were included in the GBTM. The best fitting GBTM model had four-groups, and the resulting follow-up patterns were labeled based on trajectory shape. Adding a group of individuals who only visited the mobile unit once in the one-year observation period yielded five total groups of mobile unit utilizers: slow-decline (n, %: 53, 6%), fast-decline (159, 16%), infrequent (175, 18%), continued use (42, 4%), and single-visit (535, 56%). Characteristics within the five groups were generally similar, with the continued use group having the highest average number of visits in the one-year period. The slow-decline group had the highest rates of insurance and of receiving buprenorphine during their visit. Discussion/Conclusion This study identified five follow-up patterns at a mobile unit providing MOUD to neighborhoods with unmet needs: slow-decline, fast-decline, infrequent, continued use, and single-time visitors. These findings contextualize previously reported dichotomous retention rates and provide insights into how patients may utilize the mobile unit to accomplish individualized treatment goals. Future research should incorporate behavioral characteristics and treatment received elsewhere to contextualize how follow-up patterns at a mobile unit relate to MOUD adherence.