NDARC Technical Report No. 173 (2004)


In early 2001 Australia experienced a dramatic disruption to heroin supply. In order to investigate the impact of this phenomenon, heroin users were interviewed retrospectively to determine changes in drug use behaviour over a two year period.

Two methods were used to obtain such data:

  1. A 24 month retrospective recall of drug use and associated behaviour;
  2. Retrospective interviews with maintenance pharmacotherapy clients entering treatment prior to and during the shortage, using qualitative and quantitative methods.

Time-line follow-back
Test-retest interviews were conducted with heroin users to examine the reliability of the timeline follow-back (TLFB) technique in determining heroin users’ patterns of drug use, drug treatment seeking, criminal behaviour and adverse events associated with their drug use over a two year period.

Current heroin users were recruited through a number of services and interviewed retrospectively about their drug use behaviour using the calendar method of the TLFB technique. Test-retest reliability was measured over seven days. Validity was assessed against data collected from a sub-sample of participants involved in a longitudinal cohort study.

Recall of drug use was generally high, even after 24 months. Recall was poorest during January to April 2001, the peak period of the heroin shortage. Recall of criminal activity and weekly expenditure on drugs was variable, though generally poor. Recall of treatment entry and drug related health problems such as overdose was also variable.

The two year TLFB did not obtain information reliable enough to examine sporadic drug use or to assess overall changes the patterns of drug use associated with the heroin shortage at the level necessary for this research.

Pharmacotherapy client interviews
Interviews with pharmacotherapy clients who entered treatment before and during the heroin shortage were conducted in Sydney, Melbourne and Adelaide. Participants were similar across a number of demographic variables. State data were combined and analysed as two groups: those who entered pharmacotherapy maintenance treatment before the shortage (pre-shortage group) and those entered during the shortage (shortage group).

There were few demographic differences between the two groups, therefore representing two relatively similar groups of people receiving maintenance pharmacotherapy. The two groups differed slightly on reason for entering treatment: those who entered during the shortage were more likely to report entering treatment due to police and/ or legal reasons or increased time take to procure heroin, whereas those who entered prior to the shortage were more likely to report that they did so because they ‘wanted to stop using heroin’. This, in conjunction with qualitative reports, suggests the heroin shortage acted as catalyst for treatment entry and retention.

There were few differences between the two groups in terms of level of social functioning, patterns of drug use and criminal activity. Given the time lag between the onset of the heroin shortage and interviews this was not surprising.

This research identified numerous problems with interviewing heroin users retrospectively about changes in drug supply. The techniques trialled were either not reliable enough to examine changes in the patterns of drug use at the level necessary or, the time lag between the shortage and interview was too great to determine changes. Although qualitative data provided an insight to some of the changes consequential to the shortage, it is subject to recall bias and was not accurate enough to detect definitive population change. The development of reliable, alternative methods of data collection, such as prospective cohort studies is warranted.

Citation: Day, C., Gibson, A., Collins, L., Degenhardt, L. and Dietze, P. (2004) Research methodologies used to obtain retrospective self-reports of the impact of the changes in heroin supply on regular users, Sydney: National Drug and Alcohol Research Centre.



C. Day, A. Gibson, L. Collins, L. Degenhardt, P. Dietze
Date Commenced
31 Mar 2004
Resource Type
Technical Reports