Predictors of treatment non-adherence in an inpatient substance abuse rehabilitation programme
The research study aimed to identify the factors contributing to premature termination of treatment for substance addiction. The investigation took the form of a differential research design based on archival data obtained from patient files at an inpatient drug rehabilitation centre in Gauteng. One independent variable (treatment adherence) and five dependent variables (past and present patterns of scheduled medication use, legal history and DSM-IV-TR Axis I and II co-morbidity) were chosen. Eighty-five patient files were drawn, constituting 41 treatment adherent and 44 treatment non-adherent addicts. Extraneous variables of age, age of onset, duration of addiction, previous treatment history, drug of choice, current physical health status, gender, race, and level of education were equivalent in both samples. Univariate analysis indicated that treatment adherent and drop-out groups differed significantly in terms of legal history (Fisher\'s exact test = 12.369; p = 0.002) and past patterns of use of scheduled medication (Fisher\'s exact test = 29.131; p = 0.000). A logistic regression indicated that a history of abusing a combination of scheduled psychiatric and other medication prior to treatment is the single most accurate predictor of treatment non-adherence (Wald statistic = 11.1035, p = 0.0009). Although certain combinations of past medication patterns and legal history increase predicted probabilities of treatment non-adherence, legal history on its own failed to explain any further variance that past medication could not explain on its own.
Keywords: treatment adherence; treatment non-adherence; drop-out; substance rehabilitation programmes; substance abuse
Health SA Gesondheid Vol. 12 (4) 2007: pp. 53-68
The author(s) retain copyright on work published by AOSIS unless specified otherwise.
Licensing and publication rights
Author(s) of work published by AOSIS are required to grant AOSIS the unlimited rights to publish the definitive work in any format, language and medium, for any lawful purpose. AOSIS requires journal authors to publish their work in open access under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence.
The authors retain the non-exclusive right to do anything they wish with the published article(s), provided attribution is given to the applicable journal with details of the original publication, as set out in the official citation of the article published in the journal. The retained right specifically includes the right to post the article on the authors’ or their institution’s websites or in institutional repositories.
Previously published work may have been published under a different licence. We advise the community that if they would like to reuse the work to consult the applicable licence at article level.
Note: If you need to comply with your funding body policy, you can apply for the CC BY license after your manuscript is accepted for publication.