Predictors of substance use disorder treatment completion: a multilevel analysis of longitudinal administrative data in Western Australia

Background Treatment non-completion remains a major barrier to positive substance use disorder (SUD) outcomes, yet jurisdiction-specific analyses of completion predictors using longitudinal, person-linked data remain scarce. This study examined sociodemographic, treatment engagement, and service-related predictors of treatment completion using 15 years of linked administrative data from a major SUD treatment provider in Western Australia.Methods Two-level logistic regression models were applied to data from 8,979 clients across 14,938 treatment episodes, decomposing within- and between-person variation in treatment completion (finishing the prescribed treatment regimen, versus ceasing to participate before completion). Missing data were addressed using multiple imputation, with estimates pooled across 50 imputed datasets. Models were adjusted for client age and calendar year. Sociodemographic and treatment engagement predictors were tested as pre-registered confirmatory hypotheses; substance category, episode program, and treatment type were examined in exploratory analyses.Results Sociodemographic disadvantage, including not being in paid employment, lower education, and being unpartnered, was associated with reduced odds of completion, while older age was associated with increased odds. Session attendance and episode duration were the strongest predictors at both the episode and client levels. Completion increased over calendar time, and the associations above were adjusted for this secular trend. In exploratory analyses, several differences in completion between treatment programs and types were attenuated or no longer evident after adjustment for client characteristics, suggesting that unadjusted modality comparisons may overstate the differences between them. Exploratory substance analyses indicated lower odds of completion for depressants, opiates or opioids, and cannabinoids relative to (meth-)amphetamines, and higher odds for other pharmaceuticals.Conclusion These findings underscore the need for retention strategies that direct greater support to clients facing socioeconomic disadvantage, for substance-specific engagement approaches, and for comprehensive covariate adjustment when interpreting treatment completion data from observational studies.

Authors

Institutions

Publication Details

Journal
Drugs Education Prevention and Policy
Published
2026-11-18
DOI
https://doi.org/10.1080/09687637.2026.2762884
Primary Topic
Substance Abuse Treatment and Outcomes
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Predictors of substance use disorder treatment completion: a multilevel analysis of longitudinal administrative data in Western Australia

Carol Daws, Hugh Riddell, Sascha B. Thal, BRONWYN J. MYERS et al.
2 links
Drugs Education Prevention and Policy
Substance Abuse Treatment and Outcomes
article

Predictors of substance use disorder treatment completion: a multilevel analysis of longitudinal administrative data in Western Australia

Carol Daws, Hugh Riddell, Sascha B. Thal, BRONWYN J. MYERS, James Hunter, Manisha Luhar
article en

Abstract

Background Treatment non-completion remains a major barrier to positive substance use disorder (SUD) outcomes, yet jurisdiction-specific analyses of completion predictors using longitudinal, person-linked data remain scarce. This study examined sociodemographic, treatment engagement, and service-related predictors of treatment completion using 15 years of linked administrative data from a major SUD treatment provider in Western Australia.Methods Two-level logistic regression models were applied to data from 8,979 clients across 14,938 treatment episodes, decomposing within- and between-person variation in treatment completion (finishing the prescribed treatment regimen, versus ceasing to participate before completion). Missing data were addressed using multiple imputation, with estimates pooled across 50 imputed datasets. Models were adjusted for client age and calendar year. Sociodemographic and treatment engagement predictors were tested as pre-registered confirmatory hypotheses; substance category, episode program, and treatment type were examined in exploratory analyses.Results Sociodemographic disadvantage, including not being in paid employment, lower education, and being unpartnered, was associated with reduced odds of completion, while older age was associated with increased odds. Session attendance and episode duration were the strongest predictors at both the episode and client levels. Completion increased over calendar time, and the associations above were adjusted for this secular trend. In exploratory analyses, several differences in completion between treatment programs and types were attenuated or no longer evident after adjustment for client characteristics, suggesting that unadjusted modality comparisons may overstate the differences between them. Exploratory substance analyses indicated lower odds of completion for depressants, opiates or opioids, and cannabinoids relative to (meth-)amphetamines, and higher odds for other pharmaceuticals.Conclusion These findings underscore the need for retention strategies that direct greater support to clients facing socioeconomic disadvantage, for substance-specific engagement approaches, and for comprehensive covariate adjustment when interpreting treatment completion data from observational studies.

Drugs Education Prevention and Policy
South African Medical Research Council (ZA), University of Cape Town (ZA), Curtin University (AU)
Good health and well-being, Reduced inequalities
Substance Abuse Treatment and Outcomes
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon