RESUMO
Background: Methadone maintenance treatment (MMT) is the most effective and cost-effective strategy to control HIV in Central Asian countries, where the epidemic is concentrated among people who inject drugs (PWID) who use opioids. Methods: Using data from a prospective observational database of all people initiated on MMT in Kyrgyzstan since 2008, we analyzed a more contemporary subset of data for all persons receiving MMT from January 2017 through June 2021 after the national treatment guidelines were changed. Retention on MMT was assessed at 1, 6, 12, and 24 months and predictive variables included were dosage levels, HIV status, and type of clinical setting using survival analysis. Predictors of treatment dropout were estimated using Cox multivariate regression models. Results: Among the 940 MMT patients, the proportion receiving low (<40mg), medium (40-85mg), and high (>85mg) dosage levels was 37.9%, 42.2%, and 19.9%, respectively. Increasing MMT dosage level was significantly (p<0.0001) correlated with retention at 1 (90%, 98%, 100%), 6 (42%, 63%, 95%), 12 (33%, 55%, 89%), and 24 (16%, 45%, 80%) months, respectively, with no differences between community and correctional settings. Significant predictors of dropout at 12 months included low (aHR=8.0; 95%CL=5.8-11.0) and medium (aHR=3.5; 95%CL=2.5-4.9) methadone dosage level relative to high dose, receiving MMT in three administrative regions relative to the capital Bishkek, and lower in the tuberculosis-specialized clinic in Bishkek. Clients with HIV receive higher average MMT doses (79.5mg vs 63.1mg; p<0.0001), but MMT retention did not differ after controlling for dosage in this group. Conclusions: The proportion of patients receiving optimal dosage was low (19.9%). An implementation strategy that focused on getting a larger proportion of MMT on the optimal dosage to promote retention could potentially improve the quality of existing treatment and promote further scale-up of MMT in Kyrgyzstan.