Cox J et al. – Understanding predictors of specific loss to follow–up outcomes may help methadone maintenance treatment programs improve patient retention.

• Data were collected using the Addiction Severity Index Lite.
• The probability of loss to follow-up was evaluated using a time-to-event analysis.
• Involuntary and voluntary program discharges were treated separately as the outcomes of interest.
• Multivariate Cox proportional hazards models were used to explore the role of various patient-related attributes.

• The probabilities of involuntary and voluntary discharges at 1 year were 20% and 14%, respectively.
• In this exploratory analysis, determinants of loss to follow-up were characteristics related to drug use history (e.g., use of sedatives) and its consequences (e.g., number of lifetime arrests), and differed for each outcome.
• Some determinants of involuntary discharge were modified by sex.