The effect of recovery capital on substance use cessation

Addiction recovery housing provides structured support to individuals and fosters positive habits essential for recovery. However, individuals enter recovery housing with diverse backgrounds, circumstances, and varying levels of recovery capital, which affects their substance use cessation. Facets of the relationship between recovery capital and substance use cessation remain unclear. Existing studies examine various recovery outcomes based on the development of recovery capital, but baseline levels of recovery capital have not been systematically explored. Further, the effect of different recovery capital domains on substance use cessation over time have not been clarified. This thesis addresses these research gaps by: • Identifying current methodologies and outcomes in machine learning applications in addiction by conducting a systematic review of machine learning applications in addiction. • Identifying "profiles" of individuals entering recovery housing using hierarchical clustering (n=622). • Modeling substance use cessation over time based on recovery capital profile. • Assessing the impact of different recovery capital domains on substance use cessation for each profile. • Comparing the effectiveness of a novel intensive recovery intervention to standard treatment. The thesis identifies two distinct types of people having different recovery capital profiles and drug use histories. For the most disadvantaged group, it was found that they were significantly slower to cease drug consumption. For individuals with lower recovery capital, Social Recovery Capital and participation in recovery groups were identified as the most crucial factors distinguishing between those who ceased drug consumption and those who continued. This suggests individuals entering recovery housing with low levels of recovery capital may require more intensive interventions. The novel intervention assessed in this thesis showed improved retention and growth in recovery capital among patients. These findings offer insights into the benefits of supplemental recovery support and has implications for allocating resources and improving outcomes for individuals with different recovery capital profiles.