School of Public Health

Torin Schaafsma

Estimating the impact of identifying and treating HIV-infected male circumcision clients in Uganda, Zambia, and Swaziland: a mathematical modeling analysis


Men represent an underserved population in sub-Saharan Africa with respect to HIV testing and treatment. Identification and treatment of HIV-infected men are vital both to improve their health and prevent onward transmission to their partners, especially for men in sero-discordant partnerships. One means of identifying HIV-infected men is via HIV testing and counseling in conjunction with voluntary medical male circumcision (VMMC) programs. Men in discordant couples could be identified by offering at-home testing to their cohabiting partners.


This analysis utilizes a modified version of the UNAIDS Modes of Transmission (MoT) mathematical model to estimate the number of primary HIV transmissions that could be averted over one year in 2014 and on average from 2015-2025, under two scenarios: 1) identifying and treating HIV-infected VMMC clients in discordant couples as per World Health Organization recommendations; and 2) identifying and treating all HIV-infected VMMC clients. The analysis was applied to three countries with differing magnitudes of HIV prevalence: Uganda, Zambia, and Swaziland. Univariate sensitivity analyses were performed on key model parameters.


To meet the WHO/UNAIDS target of 80% coverage of adult male circumcision by 2025, we estimate that each year from 2015 to 2025, an average of 390,945 circumcisions will need to be performed in Uganda, 189,387 in Zambia, and 20,486 in Swaziland. On average over 2015-2025 we estimate that about 2% of men presenting for male circumcision will be HIV-positive and in a discordant couple and that about 40% of those men could be identified and successful treated with ART for at least one year. Using the MoT model we predict that in one year this intervention could successfully identify and treat 2,517 men, preventing 298 new HIV cases in Uganda; 2,232 men, preventing 210 cases in Zambia, and 195 men, preventing 19 cases in Swaziland. Under a scenario in which all HIV-infected men, regardless of relationship status or CD4 count, were offered ART treatment we estimate that about 46% would start and stay on treatment for one year. In one year during 2015-2025, we predict this intervention could successfully identify and treat 9,089 men, preventing 537 new HIV cases in Uganda; 8,893 men, preventing 410 cases in Zambia, and 2,063 men, preventing 68 cases in Swaziland.

Discussion and Conclusions

VMMC programs are an important, relatively short-term opportunity to identify HIV-infected men that may not otherwise volunteer for HIV testing. A substantial proportion of the male national population in the 14 VMMC target countries will visit clinics between 2015 and 2025 and large numbers of HIV-infected men will be identified there. We predict that while circumcision programs to date have recruited mostly adolescents, as countries move toward the 80% coverage target, clients will be tend to be older and be more likely to be infected with HIV. If VMMC programs integrate HIV treatment as prevention, even if targeted only at men in discordant couples, thousands of transmissions to partners could be prevented each year.