Background:
Voluntary medical male circumcision (VMMC) is a proven intervention for reducing HIV incidence in men. Identifying specific populations and geographic areas with low VMMC coverage for targeted service outreach may help prioritize limited resources and achieve HIV epidemic control.
Methods:
We used data from the Rakai Community Cohort Study, a population-based HIV surveillance cohort in Uganda, collected June 2018-November 2020, to assess individual and community-level factors associated with MC coverage among non-Muslim HIV-negative men across 39 communities. Muslims participants were excluded because circumcision is routine religious practice. Individual factors assessed included age, marital status, religion, education level, and past year number of sexual partners. Community-level factors included community type (agrarian, trading, and Lake Victoria fishing), percentage of population <30 years, percentage with primary education or less, percentage Muslim, average distance of households to the nearest health facility (level III/IV), average distance of households to a main road, and HIV seroprevalence. Hierarchical multivariable Poisson regression with community random effects was used to estimate individual- and community-level associations with MC coverage.
Results:
Of the 8,757 men in the cohort, 86.4% (7,570) identified as non-Muslim, and 85.2% of these (n=6,448) were HIV-seronegative and enrolled in the study; Muslims were excluded from this analysis due to the expected full MC coverage as a religious practice. Of those enrolled, 61.7% were circumcised. Being circumcised was strongly associated with age, with 77% of boys and young men 15-24 years circumcised, compared to 51% of men 40-49 years old (prevalence ratio (PR)=0.64, 95% CI: 0.59-0.70). Median community coverage of MC was 66% ranging from 32% to 88% across communities. MC coverage was lower in communities with a greater percentage of people living with HIV (adj. PR=0.87,95% CI: 0.77,0.97; Figure). MC coverage decreased with greater distances to the main road (adj. PR=0.81, 95% CI: 0.69,0.95) and non-significantly with health facilities. Communities with larger Muslim populations had significantly higher MC coverage among non-Muslim HIV-negative men (adj. PR=1.13, 95% CI: 1.04,1.23).
Conclusions:
We observed individual and community-level variation in VMMC coverage among HIV-negative non-Muslim men. Prioritizing VMMC services in communities with higher HIV burden and limited health facility access may improve VMMC coverage .
![]()