Clinical competence and scientific audit

Abstract

<h3>Objectives</h3> Heroin users are at high risk of premature mortality. Despite the evidence supporting methadone maintenance programmes (MMT), methadone itself has been associated with drug-related deaths. This study aims to determine whether people prescribed methadone have an elevated risk of overdose mortality during periods of treatment transition, particularly during treatment initiation. <h3>Method</h3> Retrospective cohort study of 3162 Scottish people prescribed and dispensed liquid methadone between January 1993 and February 2004. Observation time was defined as a period during methadone treatment or a period of maximum 6 months after leaving treatment. Individual observation time was censored after 6 months off-treatment. A person9s observation time started again if they re-entered treatment after an off-treatment period. The main outcome measure was drug-related mortality by means of Cox-proportional hazards models during the 12 years of follow-up. Drug-related deaths occurring during treatment or within 3 days after last methadone prescription were considered as cases “on treatment”. Fatalities occurring 4 days or more after leaving treatment were considered to be drug-related deaths “off treatment”. <h3>Results</h3> Overall 130 people died, with 51 deaths identified as drug-related deaths (20 off treatment and 31 in treatment). Risk of drug-related mortality was higher during treatment than off treatment (adjusted hazard ratio 11.17, 95% CI 4.51 to 27.64). Inspection of timing of death showed that the risk of drug-related mortality was higher during the initial two weeks of treatment (adjusted hazard ratio 16.93, 95% CI 5.17 to 55.46) compared to the risk of mortality off treatment. Similarly, retention in treatment for more than 3 weeks was associated with increased mortality relative to being off treatment (adjusted hazard ratio 9.97, 95% CI 4.08 to 24.39). In relation to risk of mortality during treatment, being in treatment for 3–10 weeks (adjusted hazard ratio 0.36, 95% CI 0.15 to 0.85) or greater than 10 weeks (adjusted hazard ratio 0.13, 95% CI 0.04 to 0.39) was associated with a reduced risk of mortality compared to the initial two weeks on treatment. These effects were observed after adjusting for all or some of the following covariates; co-prescribing of benzodiazepines, psychiatric admission, number of methadone treatments, overuse of methadone and urine testing, where appropriate. <h3>Conclusion</h3> Excess mortality risk in the initial two weeks of methadone treatment indicates the need for more care in prescribing and monitoring of methadone when starting or restarting a patient on methadone maintenance therapy.