Severity of drug dependence does not predict changes in drug users' behaviour over time

Michael Bloor, Joanne Neale, Christopher Weir, Michele Robertson, Neil McKeganey

Research output: Contribution to journalArticlepeer-review


This paper examines the power of a measure of drug dependence to predict future changes in drug consumption and related outcomes such as criminality and homelessness. Data from an ongoing, longitudinal, interview study (DORIS) of 1033 new attenders at a range of Scottish drug treatment facilities were analysed, 16 months on from sample recruitment, using forward stepwise logistic regression models. Dependence was measured at baseline and in subsequent interview sweeps using the five-item Severity of Dependence Scale (SDS). In addition to drug consumption measures, a range of non-drug consumption measures were used, including those measuring housing status, employment status and criminality. Respondents’ SDS scores fell significantly over time, and SDS score at 16 months was a significant independent predictor of outcome measures such as reports of acquisitive crimes. However, SDS score at baseline was not an independent predictor of either drug consumption measures at 16 months or non-drug outcomes such as criminality at 16 months. Treatment services are increasingly designed to view drug use as a chronic condition, with dependence as its core property. However, while SDS score functions as a useful outcome measure, puzzlingly, severity of dependence does not predict future drug using careers in these data, the largest longitudinal cohort of Scottish drug users ever studied. Accordingly, the implications of these findings for the conceptualisation of drug use and for services planning and delivery are both unsettling and unclear.
Original languageEnglish
Pages (from-to)381-389
Number of pages9
JournalCritical Public Health
Publication statusPublished - 29 Sep 2008


Dive into the research topics of 'Severity of drug dependence does not predict changes in drug users' behaviour over time'. Together they form a unique fingerprint.

Cite this