Improving treatment completion for young adults with substance use disorder: Machine learning-based prediction...

Area of Science:

• Addiction Medicine
• Public Health
• Social Psychology

Background:

• Substance use disorder (SUD) treatment completion is influenced by numerous factors, yet the interplay of psychosocial and system-level elements remains underexplored, particularly within publicly funded services.
• Understanding these intersections is crucial for improving treatment outcomes for vulnerable populations.

Purpose of the Study:

• To investigate the combined effects of psychosocial and system-related factors on SUD treatment completion among young adults in publicly funded outpatient settings.
• To identify specific interaction patterns that predict treatment success or failure.

Main Methods:

• Analysis of psycho-social assessment data from 2909 young adults in publicly funded outpatient SUD treatment in 2021.
• Utilized the Chi-square Automatic Interaction Detection (CHAID) method to explore complex interactions influencing treatment completion.

Main Results:

• Treatment completion rates varied significantly based on improvement in total actionable items (TAI).
• Young adults with high TAI improvement in rural areas showed lower completion rates than urban counterparts.
• Middle-level TAI improvement intersected with criminal justice involvement; those with justice involvement and opioid use disorders had lower completion rates than those with non-opioid SUD.

Conclusions:

• SUD treatment completion is multifactorial, requiring consideration of TAI improvement, family strengths, demographics, and social determinants.
• Targeted interventions addressing rural disparities and justice system involvement are needed for young adults with SUD.
• Monitoring individual progress via TAI improvement is vital for predicting and supporting treatment completion.