Early exit: Estimating and explaining early exit from drug treatment

Abstract

Background: Early exit (drop-out) from drug treatment can mean that drug users do not derive the full benefits that treatment potentially offers. Additionally, it may mean that scarce treatment resources are used inefficiently. Understanding the factors that lead to early exit from treatment should enable services to operate more effectively and better reduce drug related harm. To date, few studies have focused on drop-out during the initial, engagement phase of treatment. This paper describes a mixed method study of early exit from English drug treatment services. Methods: Quantitative data (n = 2,624) was derived from three English drug action team areas; two metropolitan and one provincial. Hierarchical linear modelling (HLM) was used to investigate predictors of early-exit while controlling for differences between agencies. Qualitative interviews were conducted with 53 ex-clients and 16 members of staff from 10 agencies in these areas to explore their perspectives on early exit, its determinants and, how services could be improved. Results: Almost a quarter of the quantitative sample (24.5%) dropped out between assessment and 30 days in treatment. Predictors of early exit were: being younger; being homeless; and not being a current injector. Age and injection status were both consistently associated with exit between assessment and treatment entry. Those who were not in substitution treatment were significantly more likely to leave treatment at this stage. There were substantial variations between agencies, which point to the importance of system factors. Qualitative analysis identified several potential ways to improve services. Perceived problems included: opening hours; the service setting; under-utilisation of motivational enhancement techniques; lack of clarity about expectations; lengthy, repetitive assessment procedures; constrained treatment choices; low initial dosing of opioid substitution treatment; and the routine requirement of supervised consumption of methadone. Conclusion: Early exit diminishes the contribution that treatment may make to the reduction of drug related harm. This paper identifies characteristics of people most likely to drop out of treatment prematurely in English drug treatment services and highlights a range of possibilities for improving services. © 2008 Stevens et al; licensee BioMed Central Ltd.