EXAMINING THE INTERACTION OF INTERNALIZING AND EXTERNALIZING DISORDERS WITH HIV STATUS IN RELATION TO RESIDENTIAL SUBSTANCE USE TREATMENT DROPOUT

Each year approximately 34 billion dollars are spent on treatment for alcohol and drug use problems in the United States, yet less than half of clients complete their programs. Treatment dropout is a significant predictor of negative outcomes including substance use relapse and incarceration. The consequences of dropping out of substance use treatment are more severe for HIV-positive individuals—active drug use is associated with poor HIV medication adherence and higher mortality. Previous literature on risk factors for dropout has not explored what predictors are most relevant to HIV-positive clients and most studies have focused primarily on outpatient treatment settings, which may not generalize to residential programs. For example, substance users who enter residential care frequently present with a history of co-occurring medical and psychiatric problems, such as HIV, internalizing disorders (INT), and externalizing disorders (EXT). Internalizing and externalizing psychopathology are known to relate to both substance treatment dropout and worse health outcomes among people living with HIV. Thus, this study aimed to explore the interactions of HIV status with INT and EXT as predictors of residential substance use treatment dropout. Utilizing intake data from 1613 clients entering residential substance use treatment, we used regression models to examine several hypotheses concerning dropout risk. We hypothesized that the two-way interactions between HIV status and 1) INT and 2) EXT would be associated with dropout, such that HIV-positive individuals with higher INT or EXT would be more likely to drop out. We also predicted that those interactions would relate to fewer days spent in treatment until dropout. For the three-way interaction between HIV, INT, and EXT, we hypothesized that greater INT and EXT among HIV-positive individuals would relate to a greater likelihood of dropout and fewer days in treatment. We did not find differences between HIV-positive and HIV-negative clients in dropout rates, and the interactions between HIV and INT/EXT were not significant predictors of dropout. Results suggest that other variables such as age, education level, marital status, and court-mandated treatment status may be stronger predictors of residential substance use treatment dropout. Future directions are discussed in light of these nonsignificant findings