Unemployment and Low Education Raise Dropout Risk in Kenya’s Methadone Clinics

Methadone maintenance treatment has become one of the most important tools in the global response to opioid dependence, quietly transforming lives in clinics from Nairobi to Zurich. A new study from Mombasa, Kenya, however, reveals a counterintuitive pattern that could reshape how such programs are run: the clients most likely to abandon treatment are not the injectors with the deepest drug histories, but unemployed, non-injecting heroin users with little formal education. The findings, drawn from more than 440 clinical records at the Kisauni Medically Assisted Therapy clinic, expose the socioeconomic fault lines running beneath one of East Africa’s flagship harm reduction programs.

The research team, led by Nassoro Mwanyalu of the Mombasa County Department of Health Service together with collaborators from Kenya’s Field Epidemiology and Laboratory Training Program, conducted a retrospective review of records for opioid-dependent persons enrolled in methadone treatment at the Kisauni clinic between 2017 and 2019. Of 457 records examined, 443 met the inclusion criteria, excluding clients who had transferred into or out of the clinic or had been in treatment for less than three months. The team defined a poor clinical or health-related treatment outcome as any client who was lost to follow-up, acquired HIV, hepatitis B, or hepatitis C while on methadone, missed two or more monthly antiretroviral therapy pickups, or died during treatment, regardless of cause.

The demographic picture that emerged was strikingly male-dominated. Men comprised 90.7 percent of the eligible clients, with an overall mean age of 37 years. Female clients were markedly younger, with 79.1 percent aged 35 or below, compared with a male distribution concentrated in the 36 to 45 age bracket. Women were also more educated than their male counterparts, with 27.9 percent of female clients reaching post-primary education versus 15 percent of males. Yet women carried a disproportionate burden of infectious disease: 39.5 percent of female clients were HIV positive, compared with 8.75 percent of males, and 18.6 percent of women were hepatitis C positive versus 7.75 percent of men. Overall HIV prevalence across the cohort stood at 13.1 percent, hepatitis B at 7.2 percent, and hepatitis C at 8.8 percent.

Drug use histories reflected the coastal heroin trade that funnels through Mombasa’s position along Indian Ocean trafficking routes. At enrollment, 98.1 percent of clients were mixing heroin with other drugs, and the median duration of opioid use before joining methadone treatment was six years. Roughly 28.4 percent of clients were categorized as injecting drug users, while the remaining 71.6 percent were non-injectors, primarily smoking or inhaling heroin. During treatment, 15.8 percent of clients continued to use illicit substances, with cannabis accounting for the majority of concurrent use at 65.6 percent, followed by benzodiazepines and tricyclic antidepressants.

Treatment outcomes told a story of both success and attrition. Three-quarters of clients remained active in the program, a retention rate the authors attribute in part to a defaulter-tracking collaboration between clinic staff and two civil society organizations, the Muslim Education and Welfare Association and Reach Out Centre Trust, whose outreach workers identify and re-induct clients who drop out. Nevertheless, 22.8 percent of clients were lost to follow-up, defined under Kenya’s national protocol as missing 30 or more consecutive daily methadone doses. Five clients died during treatment, four acquired HIV, five became hepatitis B positive, and five turned hepatitis C positive while on methadone. In total, 27.5 percent of enrolled clients experienced at least one poor clinical or health-related outcome during the study period.

The statistical analysis, performed with Epi Info 7 using prevalence odds ratios and 95 percent confidence intervals, identified three factors significantly associated with being lost to follow-up. Lack of formal education carried the strongest association, with a prevalence odds ratio of 2.7, meaning clients without schooling had nearly triple the odds of dropping out. Unemployment followed with an odds ratio of 2.4, and being a non-injecting drug user at enrollment carried an odds ratio of 1.7. Notably, the period of opioid use before treatment, polysubstance use during methadone therapy, the specific type of illicit drug used while on treatment, and living with HIV were all assessed and found to have no significant effect on attrition.

The education and unemployment findings carry particular technical weight because they point to structural barriers rather than pharmacological ones. The authors suggest that clients lacking primary school education may lack employability skills, hampering reemployment during recovery and leaving them unable to finance daily travel costs to the clinic. The restricted morning service delivery model for daily observed therapy may compound the problem: employed clients cannot work regular hours before picking up their dose, and job-seeking clients may find the clinic schedule incompatible with interviews and work commitments. The authors argue this dynamic may create the perception that methadone clients are still active drug users, rendering employment difficult and increasing time spent in drug consumption spaces, which in turn raises dropout risk.

The non-injector finding stands in contrast to much of the international literature. Studies by Ledgerwood and colleagues and by Nordt and colleagues found no significant effect of heroin administration route on treatment outcomes. The Mombasa team proposes a plausible biological and behavioral explanation: injecting drug users are more likely to be HIV positive, which may increase their timely health-seeking behavior and adherence to scheduled clinical reviews and psychological support. This echoes findings by Morgan and colleagues in South Africa, who reported that non-injectors had poorer treatment outcomes than injectors in inpatient rehabilitation settings, adding African evidence to a pattern that challenges assumptions about which opioid users need the closest monitoring.

HIV treatment integration emerged as a technical bright spot. All 58 HIV-positive clients, including the four new infections acquired during treatment, were linked to a Comprehensive Care Clinic either at the Kisauni facility or nearby, and all were receiving antiretroviral therapy. Among the eight clients taking their antiretroviral medication directly at the Kisauni clinic, 75 percent had undetectable viral loads, implying that supervised antiretroviral therapy combined with daily methadone intake enhances treatment adherence. Only 3.6 percent of HIV-infected clients linked to care defaulted from antiretroviral therapy. These results align with prior research in Nairobi and Dar es Salaam reporting improved antiretroviral adherence among opioid-dependent clients on methadone, though the authors note that the proportion of HIV-positive clients taking their medication at the Kisauni clinic itself remained low.

The authors conclude that the Kisauni model confirms the effectiveness of a low-threshold, one-stop-shop methadone program integrating HIV, viral hepatitis, and sexually transmitted infection prevention and treatment. But they caution that without a more patient-centred approach, methadone risks being perceived as simply replacing one drug with another. Their recommendations include refresher training for staff, alternative dispensing models, short treatment breaks and take-home doses as incentives for stable and drug-free clients, and full integration of comprehensive HIV care into methadone services. They also call for closer monitoring of clients who are unemployed, non-injecting, or lack formal education, and for further research addressing the vulnerability of female and young opioid users, whose disproportionate infectious disease burden in Mombasa demands targeted intervention.

Subject of Research: A retrospective records review of methadone maintenance treatment clients in Mombasa, Kenya, identifying socioeconomic and behavioral risk factors for poor treatment outcomes and dropout.

Article Title: Risk factors for poor treatment outcomes among opioid-dependent clients taking methadone in Mombasa, Kenya

Article References: Mwanyalu, N., Nunga, M., Mwanyamawi, R., Abdallah, S., & Owiny, M. (2024). Risk factors for poor treatment outcomes among opioid-dependent clients taking methadone in Mombasa, Kenya. Advances in Drug and Alcohol Research, 4, Article 11791. https://doi.org/10.3389/adar.2024.11791

Image Credits: AI Generated

DOI: 10.3389/adar.2024.11791

Keywords: methadone maintenance treatment, opioid dependence, Kenya, Mombasa, lost to follow-up, HIV, hepatitis C, treatment retention, harm reduction, unemployment, non-injectors, antiretroviral therapy