NDARC Technical Report No. 190 (2004)


Introduction: Cocaine use among heroin users in NSW rose dramatically during the ‘heroin drought’ of 2001. A recent study (Williamson et al., 2003) demonstrated that heroin users who also used cocaine were a more ‘at risk’ group. As such, they displayed higher current levels of social dysfunction, drug use, needle-risk taking behaviour and criminality. It was unknown what effect the different clinical profile of heroin users who also use cocaine would have on short-term treatment outcome for heroin dependence. Moreover, it was unclear whether heroin users who also use cocaine were an inherently more dysfunctional group, or whether instead, it was cocaine use per se exacerbating dysfunction.

The current study was conducted as part of the Australian Treatment Outcome Study (ATOS). ATOS is the first large scale longitudinal study of treatment outcome for heroin dependence to be conducted in Australia. The aims of the current study were: to determine the effect of cocaine use on the short-term outcomes of treatment for heroin dependence in NSW; to examine the relationship between cocaine use and dysfunction among heroin users.

At baseline, approximately 40% of the ATOS sample had used cocaine in the month prior to interview (CU), but only half as many (19%) reported cocaine use in the month prior to 3 month follow-up. The average frequency of last month cocaine use amongst current cocaine users did not change at three months. While 35% of CU continued to use cocaine at follow up, nearly 10% of those who did not report current cocaine use at baseline had commenced cocaine use during the follow up period. Having reported recent cocaine use at baseline and not enrolling in treatment at baseline were the strongest predictors of cocaine use at three months.

Cocaine use at baseline did not appear to be significantly related to short-term treatment retention in the ATOS sample. At three months, CU and non-cocaine users (NCU) were equally likely to have completed their index detoxification or residential rehabilitation program and to still be enrolled in their index maintenance pharmacotherapy. CU and NCU also received similar average methadone and buprenorphine doses during the first three months of their index treatment.

Social functoning
At both baseline and follow up, CU displayed a greater level of dysfunction than NCU, being more likely to report being homeless and gaining the majority of their income from criminal activity and less likely to report current employment. Comparisons within groups on the basis of current cocaine use revealed social functioning to decrease during periods of cocaine use and increase subsequent to cessation of such use.

Drug use
CU continued to display more extensive polydrug use and higher levels of heroin use and dependence at three months. As was the case with social functioning, drug use behaviour appeared to be strongly influenced by cocaine use, with current cocaine users using more drug types and meeting criteria for more heroin dependence symptoms than those not currently using cocaine, irrespective of baseline cocaine use status.

At three months CU remained more likely than NCU to report injecting a drug in the month prior to follow up and to report daily injection. Needle sharing at three months was also more common among CU than NCU. All needle risk-taking behaviours were seen to decrease substantially upon cessation of cocaine use and increase with the initiation of such use.

While reported levels of criminal activity in the sample decreased substantially at three months, the higher prevalence of criminal involvement in the CU group noted at baseline was maintained. Current cocaine use was shown to be a key predictor of criminal activity in the sample, with those who reported cocaine use at three months being more likely to report recent criminal activity than those who did not, irrespective of baseline cocaine use status.

Physical and mental health
As at baseline, self-reported general physical health did not vary as a function of cocaine use. As would be expected due to their greater frequency of injection however, CU were more likely to report an injection related health problem at three months. In keeping with this, current cocaine users from both groups were more likely to have recently suffered an injection related health problem than those who did not report cocaine use at follow up. General mental health and rates of current major depression did not appear to be effected by cocaine use at either baseline or three months.

Conclusion: Cocaine use had halved amongst the sample at follow up but frequency of use among those using cocaine at three months remained unchanged. Retention in treatment was not significantly affected by cocaine use at baseline however, short-term outcomes were poorer among CU. Comparisons within groups on the basis of three month cocaine use status revealed that decreased performance on outcome measures was associated with the commencement and/or continuation of cocaine use, while cessation of cocaine use resulted in significant improvements on these measures. Thus, in the short-term, cocaine use appears to exacerbate dysfunction, rather than serving as a marker for a more dysfunctional group of individuals. The relationship between cocaine use among heroin dependent individuals and outcome will be explored further at 12 and 24 months post-treatment entry.

Citation: Williamson, A., Darke, S., Ross, J. and Teesson, M. (2004) The effect of cocaine use on short term outcomes for heroin dependence, Sydney: National Drug and Alcohol Research Centre.



A. Williamson, S. Darke, J. Ross, M. Teesson
Date Commenced
23 Jul 2004
Resource Type
Technical Reports