Treatment Completion for Substance Use Disorders Among Older Adults: Differences by Race/Ethnicity and Gender

by Ivette Rodriguez Borja

Mentor: Andrea Acevedo, Community Health; funding source: NSF TRIPODS

rodriguezborjaivette_29681_2264499_SS_PosterPresentation_IvetteRb

The population of older adults in the United States is increasing at an alarming rate as baby boomers continue to age and people are living longer. Current trends in substance use have demonstrated an increase in the number of older adults who consume and depend on substances, a behavior that places this vulnerable population at greater risk for serious health conditions. In recognition of growing substance use issues, it is important to understand substance use treatment utilization and completion among older adults, more specifically, how treatment completion varies by race/ethnicity and gender. Research on race/ethnicity and gender in the realm of substance use remains scarce, so this project intends to address gaps in our current understanding of the disparities in treatment completion. The first objective in our research is to examine whether disparities in substance use disorder treatment completion exist by race/ethnicity within gender. The second objective is to explore whether racial/ethnic disparities in treatment completion vary because of the effect of predictor variables.

For this study, data was drawn from the Treatment Episodes Dataset on Discharges (TEDS-D) for the years 2014-2017. These are publicly available de-identified datasets sponsored by the Substance Abuse and Mental Health Data Archive (SAMHDA). The primary population of this study is older adults which was defined as any record of someone age 55 or older. In order to keep the definition of “treatment” consistent, focus was placed on ambulatory non-intensive outpatient treatment. Also, individuals who had one or more prior treatment episodes were excluded from the study. All the data in this study is being analyzed using STATA, a statistical analysis software program. The current sample accounts for 49,064 clients (44% NH White Males, 20% NH White Females, 18% NH Black Males, 5% NH Black Females, 9% Hispanic Males, 2% Hispanic Females, 1% NH American Indian/Alaskan Native Males, >1% NH American Indian/Alaskan Native Females).

Our primary outcome variable is treatment completion and our predictor variables examine demographic and substance use characteristics. In general, reports of educational attainment of at least 12 years or more were higher for most racial/ethnic groups, with the exception of Hispanic clients, than reporting less than 12 years. The highest percentages of homelessness status were found in NH Black Males (9.8%) and NH Am-Ind/Al-Nat Females (9.6%). Also, the primary substance with the highest rates of reported use at admission was alcohol across all racial/ethnic gender groups (>45%).

Research for this study is still in progress, but there are some principal findings on what we have learned so far. Overall, only 53.3% of the total sample completed substance use treatment. Men displayed higher rates of treatment completion than women and this remained true across all race/ethnicity groups. Also, NH White individuals had higher rates of treatment completion than other racial/ethnic groups. Logistic regression models predicting treatment completion demonstrated that there are many similarities in the effects of predictor variables, but some differences exist as well. For example, one similarity was that within most racial/ethnic gender groups, not being employed was associated with a lower likelihood of completing treatment compared to being employed.

There were a few limitations in this study, one was that we were not able to include Asian or Pacific Islander clients in our sample because of their small sample sizes. Another was that we have no information on treatment facilities, this would help us better understand if there is something else responsible for influencing treatment completion that we may not be accounting for since not all treatment facilities operate in the same way.

There are also some implications for this study going forward. First, given the low treatment outcome shown, state and national policy should increase funding for treatment programs. By increasing funding, more support can be given to individuals in treatment with hopes that completion rates will increase.  Another is that community-level interventions and programs should work closely to better understand and accommodate for the needs of older age groups. There are also some questions we’re left with to consider outside of this research project like: How can we equitably address the needs of different groups most equitably to increase treatment completion? Also, why do some predictor variables positively affect some racial/ethnic groups while negatively affecting others?

This is an ongoing study and we are working towards submitting a paper for publishing. I’d like to thank the Tufts Summer Scholars program for supporting this research project as well as the T-Tripods Institute (NSF HDR grant 1934553) for partial funding.