Aims: To assess relationships between characteristics of methadone maintenance treatment and long-term cessation of injecting (> or = 1 year).
Design and participants: The incidence of cessation of injecting and relapse from non-injecting to injecting was estimated among 488 participants of the Amsterdam cohort study among drug users. We used a nested matched case-control design to identify methadone treatment characteristics significantly and independently related to cessation of injecting. To ensure detailed and valid assessment of methadone treatment, data of the Central Methadone Register were linked with cohort data. For 339 of 488 subjects of the initial study group methadone data were available.
Findings: The incidence of cessation of injecting increased from 2.2/100 person years in 1985-89 to 5.5/100 per year in 1995-97 (Ptrend = 0.005). Relapse to injecting was high: 17.2/100 person years (no trend). Methadone dosage and frequency of methadone programme attendance in themselves were not significantly related to cessation of injecting. However, an individual increase of 5 mg or more per year (OR 4.20, 95% CI 1.54-11.46) and receiving methadone mainly via the outpatient clinic for drug-abusing prostitutes and foreigners (OR 0.18, 95% CI 0.05-0.59) were independent predictors of cessation of injecting. After cessation of injecting, there were no HIV-seroconversions during the period of non-injecting (129 person years). After relapse to injecting there was one seroconverter; however, follow-up was small (23 person years). The HIV-incidence of those who continued injecting was 3.2/100 per year.
Conclusions: Steadily increasing the methadone dosage in a harm reduction setting may be useful in supporting injecting drug users in the process of cessation of injecting and reducing the spread of HIV-infection.