(136) Opioid Treatment Engagement and Retention: A Risk and Protective Factor Examination

Background & Introduction: Opioid use disorder (OUD) remains a significant public health concern. Nationally, there were an estimated 80,391 overdose deaths in 2024. The persistence of OUD and impact of overdose deaths highlight the need for effective treatment. Medications for opioid use disorder (MOUD) are an evidence-based treatment1, but treatment gaps persist. In 2024, 4.8 million people in the United States had OUD but only 17% accessed MOUD2. Once treatment is accessed, retention in care is an ongoing challenge, despite its link to sustained recovery.1 Retention rates in the literature vary from a low of 19% at 3 months and 3% at 6 months.1 These challenges highlight the need for innovations that expand access to evidence-based treatment and sustain engagement. In 2016, Pennsylvania launched the Centers of Excellence for Opioid Use Disorder (COE) program to close access gaps and improve treatment engagement for individuals with OUD through the provision of community-based care management services. The reach of this innovative program has grown with 417,215 interactions reported statewide. Given the COE’s guiding principles are rooted in promoting sustained engagement for individuals with high disease acuity and low recovery capital who struggle to remain engaged, program evaluation focuses on understanding specific factors that influence retention.

Methods: Care management interactions were analyzed to investigate risk and protective factors influencing client retention. The analysis included a cohort of 35,260 clients enrolled across 56 organizations across 148 sites with at least 10 clients between 2022 and 2024. Each patient’s care management data was examined within a 12-month window, following each patient for a year from the date of their initial visit. Using a hierarchical ordinal regression model, retention was predicted by variables measuring client substance use severity (high = current IV use and any history of overdose; low = all other patterns of IV use and overdose history)1; client receipt of mental health, housing, transportation, or food services (yes/no for each service); client receipt of virtual or offsite services (yes/no); client criminal justice involvement (none, past, or current); highest client level of care assessment (LOCA) score (high = intensive outpatient treatment or higher; low = standard outpatient treatment or lower); and average total score for client Brief Assessment of Recovery Capital (BARC-10). Retention in care management was defined as length of first episode of care in days based on client interactions with no gaps greater than 61 days at which point care management was discontinued.

Results: The hierarchical ordinal regression model explained 40% of the variance in client retention, with 15% of this variance attributed to COE differences, indicating substantial organizational variation in retention outcomes beyond measured client characteristics. Mean retention in care management was 116 days (SD = 128) and median was 63 days (IQR = 197). Because the outcome variable distribution was bimodal, retention was divided into an ordinal variable with the following groups: immediate dropout (0 to 3 days; n = 8,495), early disengagement (4 to 30 days; n = 4,921), partial engagement (31 to 90 days; n = 6,998), substantial engagement (91 to 180 days; n = 5,235), and long term engagement (>180 days; n = 9,611).
Several factors demonstrated strong associations with longer client retention. Receipt of mental health services showed the strongest association (.21% long-term retention change/73 more long-term clients) with extended retention across COEs, though this relationship had considerable organizational variation, potentially reflecting differences between behavioral and physical health providers. Other supportive services, including housing (.11% long-term retention change/40 more long-term clients), transportation (.10% long-term retention change/34 more long-term clients), and food assistance (.08% long-term retention change/29 more long-term clients), also predicted longer retention periods. Non-traditional service delivery also emerged as a potentially effective retention strategy. Clients receiving virtual or off-site services demonstrated significantly longer engagement periods (0.17% long-term retention change/40 more long-term clients), suggesting that flexible service delivery options may reduce barriers to sustained care management.
Several client characteristics were associated with shorter retention periods. The variables with the strongest relationships with shorter retention were missing BARC-10 (-.20% long-term retention change) and LOCA scores (0.06% long-term retention change), which may reflect inconsistent data collection practices across COEs rather than representing a true client risk factor. Current or past criminal justice involvement emerged as the second most significant predictor of earlier disengagement from services (-0.04% long-term retention change/15 fewer long-term clients). Clients with high-risk substance use patterns (-0.03% long-term retention change/10 fewer long-term clients) and those requiring intensive levels of care as defined by high LOCA scores (-0.03% long-term client change/11 fewer long-term clients) also demonstrated shorter retention periods, though these effects were more modest.

Conclusion & Discussion: COEs have offered enhanced care management services to a growing number of individuals with OUD. Many of the identified services are widely recognized as factors influencing retention. Studies have found transportation to be a barrier in MOUD adherence3 and housing instability has been identified as a barrier to treatment.4 In addition, mental health services are a documented factor linked to retention in care as co-occurring mental health conditions are common among individuals with OUD.5 Results also suggest that in addition to linking clients to medical and non-medial support services, it may be helpful to increase recovery capital of those with severe SUD as higher recovery capital was shown to be a protective factor. As COEs place greater emphasis on serving priority populations, addressing the care management needs of higher-risk individuals may also increase the importance of protective factors in keeping clients engaged in care. Using this analysis as a basis for ongoing implementation of the COE program through a lens of quality improvement, future directions may include examining in more detail how the risk and protective factors differ based on COE programming model or population served especially related to telehealth and mobile options.

References: 1. Timko, C., Schultz, N. R., Cucciare, M. A., Vittorio, L., & Garrison-Diehn, C. (2016). Retention in medication-assisted treatment for opiate dependence: A systematic review. Journal of addictive diseases, 35(1), 22–35. https://doi.org/10.1080/10550887.2016.1100960
2. Substance Abuse and Mental Health Services Administration. (2025). Key substance use and mental health indicators in the United States: Results from the 2024 National Survey on Drug Use and Health (HHS Publication No. PEP25-07-007, NSDUH Series H-60). Center for Behavioral Health Statistics and Quality, Substance Abuse and Mental Health Services Administration. https://www.samhsa.gov/data/data-we-collect/nsduh-national-survey-drug-use-and-health/national-releases
3. Hutchison, M., Russell, B. S., Leander, A., Rickles, N., Aguiar, D., Cong, X. S., Harel, O., & Hernandez, A. V. (2023). Trends and Barriers of Medication Treatment for Opioid Use Disorders: A Systematic Review and Meta-Analysis. Journal of Drug Issues. https://doi.org/10.1177/00220426231204841.
4. Nguyen, CM, Kubiak, G, Dixit, N, Young, SA, Hayes JR. (2024). Evaluating Barriers to Opioid Use Disorder Treatment from Patients’ Perspectives. PRiMER. 2024;8:11. https://doi.org/10.22454/PRiMER.2024.458349.
5. Vekaria, V., Bose, B., Murphy, S. M., Avery, J., Alexopoulos, G., & Pathak, J. (2021). Association of co-occurring opioid or other substance use disorders with increased healthcare utilization in patients with depression. Translational Psychiatry, https://doi.org/10.1038/s41398-021-01372-0