Research That Matters (January 17 - 20, 2008)

Background: Non-elderly disabled Medicaid beneficiaries who are dually-eligible for Medicare face a variety of costly health conditions which may impact their ability to participate in community life. While people with disabilities are known to be at high risk for substance abuse, access to treatment is limited for this population, often as a result of stigma and cross-system collaboration challenges. Substance abuse may exacerbate health conditions among people with disabilities. Medicare is the primary source of coverage for people with disabilities who are dually eligible for both programs, and while Medicaid pays for up to 20% of specialty substance abuse treatment nationwide, Medicare has limited substance abuse benefits. Despite widespread concern about high health care costs incurred by this population, little is known about access to potentially cost-saving substance abuse treatment. This study assesses rates and predictors of access to treatment as well as treatment dropout. Methods: Using a cross-sectional design, substance abuse treatment episodes were examined via analysis of national Medicaid and Medicare administrative claims data from calendar year 1999. A weighting system was developed to address data missing as a result of capitated payment approaches. A sample of dually-eligible beneficiaries aged 18-64 (N=130,109) and a comparison group of non-dually eligible Medicaid beneficiaries (N=746,081) both with substance abuse diagnoses within the year were identified. Drawing on performance measures used by the National Committee on Quality Assurance to assess treatment access, data about treatment initiation and engagement were gathered. Results: Findings suggest that the sample was less likely to both initiate (age/gender-adjusted odds ratio (OR)= 0.9***, p<.001) and engage (age/gender-adjusted OR= 0.9***) in treatment. Guided by Anderson's sociobehavioral model of health care utilization, predictors of both initiation and engagement derived from multivariate logistic regression analyses suggest that women (OR=0.7***), those in the older portion of the sample (OR=0.9***) and Whites (OR=0.9***) were less likely to access treatment. Predictors of access included serious and persistent mental illness (SPMI), depression, substance abuse-related medical conditions (i.e. alcoholic gastritis), fee-for-service coverage or residence in an area with a high percentage of people living below the Federal poverty level. Despite identification of apparent pathways to accessing treatment, the sample as a whole was more likely to drop out of treatment within a month (age/gender-adjusted OR=1.2***) while sample members with depression (OR=1.2***) were more likely to drop out within two weeks. Conclusions: Implications relate to how existing service systems support the provision of substance abuse-related care to this population and the need for social work case managers to facilitate cross-system collaboration in support of improving access and reducing treatment dropout among vulnerable and often stigmatized populations of people who are dually eligible. Given the importance of treatment retention in fostering better recovery outcomes, population-specific improvements in service provision and dropout prevention are necessary, especially for women, older people and Whites. Indications that the presence of either a SPMI or depression may facilitate access to treatment, despite higher rates of later dropout among the latter suggest an important boundary-spanning role for clinical social workers in mental health settings.