This study focused on the population in the U.S. from 2017 to 2020 who abused alcohol, marijuana, or heroin. We examined their socioeconomic factors to understand the relationship between these factors and treatment dropout. Our study presents several strengths, including a comprehensive analysis of demographic and socioeconomic association with treatment dropout among individuals with substance use disorders. By employing logistic regression analysis, we identified significant predictors of dropout probability both in the overall population and across different service types, enhancing our understanding of treatment retention dynamics. Additionally, the inclusion of a diverse sample strengthened the generalizability of our findings, as we examined patients of various age groups, racial backgrounds, geographical locations, and socioeconomic statuses. Further analysis was conducted on dropout probability in subpopulations, focusing on patients with varying histories of prior arrests or previous SUD treatments.

Substance abuse treatment has a dropout rate of approximately 24% among patients. The 24% dropout rate observed in our dataset aligns the findings of two U.S.-based studies, which reported rates of 21% and 25% [25,26]. However, this prevalence stands notably lower than the dropout rates reported in other studies, ranging from approximately 50% to 88% [27,28,29]. These discrepancies may be attributed to variations in how dropout is defined as well as differences in sample sizes—our study included 912,832 admission records, whereas prior studies involved only a few hundred.

To investigate the determinants of dropout, we evaluated discharged patients from treatment facilities to understand the causes of discontinuation and to identify connections between various socioeconomic factors and treatment attrition. Our analysis reveals a strong correlation between geographical factors, such as state and CBSA (Core-Based Statistical Area) regions, and patient dropout. This finding is consistent with previous studies [30,31]. Urban areas and rural areas may have different access to treatment. Compared to urban areas, people living in rural areas generally have significantly less access to healthcare and treatments due to a shortage of treatment facilities, longer travel distances to reach medical facilities, and limited availability, often leading to delayed or forgone medical treatment; this can result in poorer health outcomes for rural populations.

Our study shows that in both the overall population and subpopulations considering prior arrests and prior treatments, individuals over 65 consistently exhibit a lower probability of dropout compared to other age groups. This result is consistent with the findings of José J. López-Goñi et al. [15] and McKellar et al. [32], which indicate that younger individuals have a higher risk of dropping out. This suggests that older individuals may demonstrate greater adherence to treatment regimens, potentially due to factors such as increased health awareness or a stronger support network.

In the various populations we investigated, we observed that Black or African American individuals consistently have a higher probability of dropout compared to other racial groups. A report by Saloner and Cook found a similar trend, indicating that Black, Hispanic, and Native American individuals are less likely than white individuals to complete treatment for substance use disorders. The authors attributed these disparities to factors such as unemployment, housing instability, and other underlying issues [33]. The study by Mennis and Stahler also showed that African Americans are less likely to complete treatment compared to whites in urban outpatient substance use disorder programs, primarily due to social isolation [34].

SUD treatment dropout is closely related to patients’ financial situation. It has been observed that patients whose primary source of income is public assistance are more likely to discontinue treatment. Additionally, those with Medicaid insurance or no insurance at all have a higher likelihood of dropping out. Our findings align with a previous study highlighting the significant impact of patients’ financial problems on substance abuse treatment dropout [35]. Cook et al. found that patients with Medicaid insurance are more likely to enroll in SUD treatments; however, their study did not investigate dropout events [36].

Regarding clients in different treatment services, we found that Black or African American patients receiving public assistance exhibit elevated dropout rates in ambulatory services. In contrast, patients of other racial backgrounds without private insurance show higher dropout rates in detox services. Individuals aged 18 to 49 who are not in the labor force display increased dropout rates in rehab services. Interestingly, patients who identify as Native Hawaiian, American Indian, Pacific Islander, or another single race tend to have a lower treatment dropout rate as number of arrests increases. In addition, prior arrests and previous SUD treatments affect dropout rates differently. Patients in dependent living situations who pay for services through private insurance or receive them at no charge have lower dropout rates as the number of arrests increases. Conversely, independently living patients with prior SUD treatments exhibit higher dropout rates compared to those undergoing treatment for the first time.

Our study may be limited by its reliance on secondary data collected from treatment facilities, which could introduce data limitations and selection bias, particularly concerning the accuracy and completeness of patient records. While we identified significant predictors of treatment dropout, the cross-sectional nature of the data restricts our ability to infer causality. Future research utilizing longitudinal study designs could provide deeper insights into the temporal relationships between demographic factors, service utilization, and treatment outcomes.

Despite controlling for various demographic and socioeconomic variables, our analysis may not account for all potential confounders influencing treatment dropout. Factors such as social support networks, treatment engagement, and comorbid psychiatric conditions warrant further investigation in future studies. The treatment environment, for example, whether adjustments can be made to accommodate specific patient needs, is important to the patient’s decision to remain in a treatment [11]. Poorer psychiatric functioning [22], anxiety sensitivity [26], and having personality disorders [18] are related to dropout risks, too. Additionally, it is likely that some factors we studied are interdependent; for example, a patient’s primary source of income may influence the type of health insurance they use and their payment methods.

In conclusion, this study underscores the strong association between patients’ socioeconomic factors, demographic characteristics, and treatment dropout among individuals with substance use disorders (SUDs). Specifically, we found that geographic factors including state and core-based statistical area (CBSA) are highly correlated with treatment dropout. With the entire population studied, patients aged equal to or greater than 65 have a much lower chance of dropping out the treatment compared with other age groups. Additionally, Black or African American patients who rely on public assistance as their primary source of income, as well as those with Medicaid insurance or lacking coverage, are at a heightened risk of discontinuing treatment. Native American/American Indian/Pacific Islander groups tend to have a lower dropout rate when they had arrests prior to admission or patients have received prior SUD treatments. These findings highlight an urgent need for targeted interventions that address financial barriers to treatment access and retention, ultimately aiming to improve outcomes for individuals with SUDs.