Abstract
Aims: Drugs intelligence systems in Merseyside have identified a growing number of treatment-seeking clients who express a preference for injecting crack cocaine. We investigated the history of crack-cocaine injecting behaviours in individuals attending Merseyside drug-treatment agencies.
Methods: Data were drawn from the regional National Drug Treatment Monitoring System (NDTMS) database for all those individuals reported to have presented for structured treatment from 1 April 1999 to 31 March 2002 in Merseyside. This resulted in a sample size of 4055. Subjects were grouped according to whether they reported injection of crack cocaine. Details including demographics, current drug-use profile and behaviours surrounding drug use were derived from the NDTMS dataset. Prevalence ratios were calculated and significant variables were subsequently entered into a backwards stepwise logistic regression model to assess the relative strength of association between subject characteristics and incidences of crack-cocaine injection.
Findings: Clients that had injected crack cocaine and those that had not were well matched on demographic details and drug use other than heroin. Controlling for area of residence, crack-cocaine injectors were more likely to report use of injected heroin (p<0.001), use of non-injected cocaine (p<0.01), and less likely to report use of non-injected heroin (p<0.001). Logistic regression showed that crack injection was significantly predicted by heroin injection (p<0.001) and being a Liverpool resident (p<0.05).
Conclusions: This preliminary investigation suggested that heroin injection is also a strong indicator of crack-cocaine injection. Crack-cocaine injectors may represent a subset of heroin users rather than a distinct population. Patterns of crack-cocaine administration were locally determined and require a dynamic response from service providers at the local level. It is vital that intervention strategies adapt to varieties and changes in drug administration behaviours in order to remain effective.