Tuberculosis case detection by trained inmate peer educators in a resource-limited prison setting in Ethiopia: a cluster-randomised trial

BACKGROUND: To improve tuberculosis case detection, interventions that are feasible with available resources are needed. We investigated whether involving trained prison inmates in a tuberculosis control programme improved tuberculosis case detection, shortened pre-treatment symptom duration, and increased treatment success in a resource-limited prison setting in Ethiopia.

METHODS: In this cluster-randomised trial we randomly assigned prisons in the regions Amhara and Tigray of Ethiopia to an intervention group or a control group, after matching them into pairs based on their geographical proximity and size. Larger prisons were considered eligible whereas smaller prisons were excluded. We selected three to six prison inmates in each intervention prison. The recruited prison inmates who received a 3-day training course and were capable of identifying presumptive tuberculosis cases then provided health education to all other prison inmates about tuberculosis prevention and control every 2 weeks for 1 year. They also actively searched for symptomatic prison inmates and undertook routine symptom-based tuberculosis screening. The control prisons followed the existing passive case finding system. The primary outcome was the mean case detection rate at the end of the year, measured at cluster (prison) level. This trial is registered at ClinicalTrials.gov, number NCT02744521.

FINDINGS: We randomly assigned 16 prisons with a total population of 18 032 inmates to either the intervention group (n=8) or the control group (n=8) from April 1, 2016, to March 31, 2017. During the 1-year study period, 75 new tuberculosis cases (1% of 8874 total inmates) were detected in the intervention prisons and 25 new cases (<1% of 9158 total inmates) were detected in the control prisons. The mean case detection rate was significantly higher in the intervention group than in the control group (mean difference 52·9 percentage points, 95% CI 17·5-88·3, p=0·010).

INTERPRETATION: Involving trained inmate peer educators in the tuberculosis control programme in Ethiopian prisons significantly improved the tuberculosis case detection rate. The findings have important implications for clinical and public health policy, particularly in prisons of low-income countries where tuberculosis burden is high and the recommended tuberculosis diagnostic and treatment algorithms have generally not been implemented.

FUNDING: Nuffic, Mekelle University.