Mapping male circumcision for HIV prevention efforts in sub-Saharan Africa

Abstract

Background: HIV remains the largest cause of disease burden among men and women of reproductive age in sub-Saharan Africa. Voluntary medical male circumcision (VMMC) reduces the risk of female-to-male transmission of HIV by 50-60%. The World Health Organization (WHO) and Joint United Nations Programme on HIV/AIDS (UNAIDS) identified 14 priority countries for VMMC campaigns and set a coverage goal of 80% for men ages 15-49. From 2008 to 2017, over 18 million VMMCs were reported in priority countries. Nonetheless, relatively little is known about local variation in male circumcision (MC) prevalence.

Methods: We analyzed geo-located MC prevalence data from 109 household surveys using a Bayesian geostatistical modeling framework to estimate adult MC prevalence and the number of circumcised and uncircumcised men aged 15-49 in 38 countries in sub-Saharan Africa at a 5 × 5-km resolution and among first administrative level (typically provinces or states) and second administrative level (typically districts or counties) units.

Results: We found striking within-country and between-country variation in MC prevalence; most (12 of 14) priority countries had more than a twofold difference between their first administrative level units with the highest and lowest estimated prevalence in 2017. Although estimated national MC prevalence increased in all priority countries with the onset of VMMC campaigns, seven priority countries contained both subnational areas where estimated MC prevalence increased and areas where estimated MC prevalence decreased after the initiation of VMMC campaigns. In 2017, only three priority countries (Ethiopia, Kenya, and Tanzania) were likely to have reached the MC coverage target of 80% at the national level, and no priority country was likely to have reached this goal in all subnational areas.

Conclusions: Despite MC prevalence increases in all priority countries since the onset of VMMC campaigns in 2008, MC prevalence remains below the 80% coverage target in most subnational areas and is highly variable. These mapped results provide an actionable tool for understanding local needs and informing VMMC interventions for maximum impact in the continued effort towards ending the HIV epidemic in sub-Saharan Africa.

Keywords: Africa; Geospatial; Geostatistics; HIV; HIV prevention; Intervention; Male circumcision; Mapping; Medical male circumcision; Spatial statistics; Voluntary medical male circumcision.

Fig. 1

Estimated male circumcision prevalence, adult men ages 15–49 years, 2017. Male circumcision prevalence among adult men ages 15–49 years in 2017 at the country level (a), first administrative unit level (b), second administrative unit level (c), and 5 × 5-km grid-cell level (d). Maps reflect administrative boundaries, land cover, lakes, and population. Grid cells with fewer than ten people per 1 × 1-km [39] and classified as “barren or sparsely vegetated” [47] are colored in dark gray. Countries in light gray were not included in the analysis. Outlined by a thick black border are priority countries for VMMC campaigns in southern and eastern Africa, as identified by the WHO and UNAIDS

Fig. 2

Mean, lower, and upper bounds of estimated male circumcision, ages 15–49, 2017. MC prevalence among males ages 15–49 in 2017 at the first administrative unit level (a–c), second administrative unit level (d–f), and 5 × 5-km grid-cell level (g–i). Mean estimates and lower and upper bounds of the 95% uncertainty intervals are shown in the left, middle, and right columns, respectively. Maps reflect administrative boundaries, land cover, lakes, and population. Grid cells with fewer than ten people per 1 × 1-km [39] and classified as “barren or sparsely vegetated” [47] are colored in dark gray. Countries in light gray were not included in the analysis

Fig. 3

Range of estimated male circumcision prevalence at the second administrative level, 2017. Estimated range of male circumcision prevalence for each second administrative level unit in 2017 by country. Each point represents the estimated prevalence of a single second administrative level unit within a country in 2017, and the vertical bars indicate the prevalence of the highest and lowest second administrative level units in 2017 by country. Color indicates whether or not the country was a priority country for VMMC campaigns, as identified by the WHO and UNAIDS

Fig. 4

Estimated number of uncircumcised adult men at the second administrative level, 2017. Number of uncircumcised men ages 15–49 years in 2017 at the second administrative unit level. Maps reflect administrative boundaries, land cover, lakes, and population. Grid cells with fewer than ten people per 1 × 1-km [39] and classified as “barren or sparsely vegetated” [47] are colored in dark gray. Countries in light gray were not included in the analysis. Outlined by a thick black border are priority countries for VMMC campaigns in southern and eastern Africa, as identified by the WHO and UNAIDS

Fig. 5

Estimated change in male circumcision prevalence, adult men age 15–49, in 14 priority countries. Absolute change at the country level in male circumcision prevalence among adults age 15–49 between 2000 and 2008 (a) and between 2008 and 2017 (b). Absolute change at the second administrative unit level in male circumcision prevalence between 2000 and 2008 (c) and between 2008 and 2017 (d). Maps reflect administrative boundaries, land cover, lakes, and population. Grid cells with fewer than ten people per 1 × 1-km [39] and classified as “barren or sparsely vegetated” [47] are colored in dark gray. Countries in light gray were not included in the analysis. Outlined by a thick black border are priority countries for VMMC campaigns in southern and eastern Africa, as identified by the WHO and UNAIDS

Fig. 6

Posterior probability of exceeding 80% male circumcision coverage, adult men ages 15–49, 2017. Posterior probability of exceeding the 80% male circumcision prevalence target among adult men ages 15–49 in 2017 at the country level (a), first administrative unit level (b), second administrative unit level (c), and 5 × 5-km grid-cell level (d). Maps reflect administrative boundaries, land cover, lakes, and population. Grid cells with fewer than ten people per 1 × 1-km [39] and classified as “barren or sparsely vegetated” [47] are colored in dark gray. Countries in light gray were not included in the analysis. Outlined by a thick black border are priority countries for VMMC campaigns in southern and eastern Africa, as identified by the WHO and UNAIDS