Treatment completion among Australians attending publicly-funded specialist alcohol and other drug treatment services.

Introduction: Completion of alcohol and other drug (AOD) treatment is associated with improved health and social outcomes. Previous research has largely focused on individual (e.g. demographic) or service-level (e.g. treatment modality/setting) factors related to treatment completion in isolation. This study investigates the relationship between treatment completion and service-level and substance related factors, after controlling for demographics.

Study design: Retrospective cohort of 53,430 people engaging in government funded specialist AOD treatment across New South Wales, Australia, between 1 January 2015-31 December 2018.

Methods: Generalised linear mixed models were used to analyse factors associated with treatment completion, accounting for repeated measures across treatment episodes.

Results: Approximately two-thirds (69.8 %, n = 37,330) of the cohort completed treatment at least once during the study period, 42.3 % (22,605) on a single, and 27.5 % (14,725) on multiple occasions. After controlling for demographics, treatment episodes for amphetamines were least likely to be recorded as complete (52.5 %), while those for MDMA were 1.9 (95 % CI: 1.49, 2.45) times more likely (67 %) than amphetamine episodes to be completed. Treatment episodes mandated through criminal justice/child protection agencies were 1.25 (95 % CI: 1.20, 1.30) times more likely to be completed compared to those originating from self-referral. There were no differences in treatment completion between self or health professional referrals (aOR: 0.98, 95 % CI: 0.95, 1.02). Episodes involving involuntary AOD treatment modalities and residential withdrawal were 6.67 times (95% CI: 4.53, 9.81) and 5.02 times (95 % CI: 4.46, 5.64) more likely respectively to be completed compared to those for community rehabilitation. Case management episodes were also more likely (aOR: 2.43, 95 % CI: 2.16, 2.73) to be completed. Episodes of longer treatment duration (≥90 days) were 1.89 times (95 % CI: 1.82, 1.97) more likely to be completed compared to shorter (≤30 days) treatment episodes.

Conclusions: Treatment completion was moderated by a range of factors including drug type, treatment modality and duration, and referral source. Low rates of amphetamine treatment completion across treatment modalities confirms the urgent need for further research investigating more effective treatment options for amphetamine use disorders.