Artikel

Abstract

Between simplification and complexity – structured interviews and treatment results in substance abuse treatment AIM The article draws on an empirical case study to analyse and discuss the application of structured interviews in evaluating outcomes of substance abuse treatment. METHODS The empirical example, a follow-up study, is based on the structured interview DOK (a Swedish interview instrument similar to the ASI) with 39 clients at an inpatient substance abuse treatment unit, at intake and six months after completed treatment. Changes in self-reported data were analysed between intake and follow-up. The analyses differentiate between subjective client estimations and verifiable data. This is followed by an analysis of individual variations in changes of alcohol and drug use. RESULTS The study shows how it is possible within a sample from a local treatment unit to illustrate relatively big changes on a group level in some of the objective, verifiable variables. Further, depending on the time period (last 30 days or last 6 months), the clients reported significantly differently on their use of the primary drug of abuse. At group level, there was a reduction of self-reported problems within all life areas. There were indications of some relation between verifiable variables and estimation variables. All clients reported, however, that their estimated problems in all life areas persisted to a relatively high degree at follow-up. In order to further point at individual variations in the patterns of change, the results of three different verifiable variables within the area of Alcohol and drugs are reported. The share of positive treatment results vary between 13–64 %, depending on definition and operationalisation. CONCLUSIONS The study is an example of an analytical model that can be applied in a local treatment unit with data from structured interviews. The results illustrate the problems with using questions about alcohol and drug use only for the last 30 days in follow-up studies. The results also indicate that subjective evaluations have no self-evident relation to reported changes within one life area. It is thus problematic to use a self-reported problem as an outcome measure or mix this measure with other outcome data.