Voluntary medical male circumcision (VMMC) has been demonstrated to reduce the risk of female-to-male transmission of HIV by an estimated 60%. Quality and safety is essential to the successful implementation of VMMC programs at scale and their social acceptability. In 2010, the Uganda Ministry of Health (MOH) developed a policy to guide implementation of VMMC (termed safe male circumcision [SMC]) as part of the comprehensive package of HIV prevention strategies and set a goal of circumcising 1,250,000 males between the ages of 15 and 49 by 2013. In 2013, the MOH established and disseminated quality standards for VMMC along the following seven domains: management systems; supplies, equipment, and environment; registration, group education, and information, education, and communication (IEC); counseling and testing; surgical procedures; monitoring and evaluation; and infection prevention. An external quality assessment by PEPFAR in 2013 identified quality gaps in Uganda’s VMMC program. The USAID Applying Science to Strengthen and Improve Systems (USAID ASSIST) Project provided technical support to 30 sites to address these gaps. Using a continuous quality improvement (QI) approach, facilities tested changes to address gaps in VMMC service quality including improving rate of patients who return for follow-up within 48 hours, 7 days, and 6 weeks post-circumcision and meeting minimum quality standards.
n Evaluate quality improvement approaches in improving facility adherence to the Uganda Ministry of Health Quality Standards for SMC and rates of patient follow-up at 48 hours, 7 days, and 6 weeks.
n Explore decision-making around seeking circumcision services and returning for follow-up care, and the role of female partners in the decision-making process among males in Uganda.
Pre-post design was used to evaluate facility adherence (longitudinal data) and patient follow-up rates (two independent cross-sections). Qualitative interviews with patients who had been circumcised at participating facilities and either had or had not returned for follow-up care.
The work described to improve VMMC services in Uganda was supported by the American people through the United States Agency for International Development (USAID) Applying Science to Strengthen and Improve Systems (ASSIST) Project with funding from the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR). The project is managed by University Research Co., LLC (URC) under the terms of Cooperative Agreement Number AID-OAA-A-12-00101.
The 30 USAID ASSIST-supported sites were spread across Uganda. Of these, 9 facilities were purposively selected for participation in the study to reflect variability in facility type and size. The sample included five hospitals, three HC IVs, and one HC III.
Each facility was assessed at baseline and endline using the Ugandan MOH-approved QI for SMC for HIV Prevention Assessment Tool. Data were collected through discussion with key personnel, observation of environment, and review of materials and patient charts. For standards involving direct patient services, three observations were required. Points were tallied and each facility was given a score out of a possible 100. Data were imported into SPSS 21 for analysis after an independent audit of the dataset verified accuracy of the data entry. Dependent two-tailed t-tests were conducted to compare average number of criteria fulfilled by domain at baseline versus endline.
Patient follow-up data
Data on patient follow-up rates were extracted from randomly selected patient charts. A target sample of 230 for each baseline and endline was calculated assuming a 50% patient follow-up rate, difference of 20%, power of 0.8, alpha of 0.05, and design effect of 2. Due to limited charts at baseline, only 203 were reviewed; 231 were reviewed at endline. Pearson’s Chi-square tests were performed to test the null hypothesis of no association between measurement period (baseline v. endline) and follow-up category (missing v. follow-up v. no follow-up). Post-hoc analyses were then conducted using the Bonferroni correction for three tests.
Qualitative interviews on follow-up and sample characteristics
Males aged 13+ who had received circumcision services were included in the sampling frame. 21 interviews with males who did not return for follow-up (7 age 13-17; 14 age 18+) and 25 interviews with males who did return for follow-up (5 age 13-17; 20 age 18+) at least once in the 6 weeks following the procedure were conducted. All interviews were audio-recorded, transcribed, and translated into English for analysis. A codebook was developed along key topics included in the interview guide. An iterative approach was taken to refine and apply the codes.
All facilities consented to participate in the study. Signed consent was obtained from all respondents over the age of 18; for those respondents under the age of 18, assent was obtained along with consent from their guardian. Ethical approval was obtained from the University Research Co., LLC Institutional Review Board, the Mildmay Uganda Research and Ethics Committee, and the Uganda National Council for Science and Technology.
Facility adherence to standards
Overall, adherence to standards improved between baseline and endline. In table 2, findings for each standard domain are presented for 2 analysis scenarios. In the first, all undocumented or missing values were coded as 0 or non-adherent; in the second, missing values were excluded. In both scenarios, statistically significant improvement was observed in the domains of management systems; supplies, equipment, and environment; and monitoring and evaluation. Coding the missing values as non-adherent resulted in significant improvements along the remaining four domains of registration, group education and IEC; individual counseling and HIV testing; surgical procedures; and infection prevention, but, when coded as missing, improvements were observed but were not significant.
Reasons for getting circumcised
Respondents reported first learning of VMMC from friends, family, or the radio. Reasons for seeking circumcision included:
n Improved personal hygiene and reduced disease transmission, specifically HIV and cancer; helped eliminate “dirt” that collected in the skin of the penis, allowing them to go longer without bathing and not “smell, unlike the uncircumcised person”
n Improved sexual performance
n Social pressure among peers
n Socio-cultural or religion as barrier or facilitator of uptake of VMMC services; 3 respondents initially believed that “circumcision was only for Muslims”; 1 respondent reported cultural practices of his wife led to his decision to be circumcised
Female partner involvement
Among the 28 respondents who reported being in a relationship or married at the time of their circumcision, 19 (68%) stated their female partner was involved in or informed of their choice to seek circumcision. For 8, their female partners told them to get circumcised to prevent disease transmission and promote hygiene. Respondents reported being concerned their relationship would suffer if their female partners were not included in the decision. As one respondent explained:
I told her about my decision to get circumcised because I wanted her to know about it and be okay with it, such that she may not dump me after I get circumcised if she did not want a guy who is circumcised.
Respondents who reported being in a relationship but not involving their partner in the decision indicated they did not inform their partners because it was a “private matter” or because they did not want their partners to stop them or convince them to change their minds.
I never told my wife before circumcision, she only realized it after the procedure…Sometimes women can be suspicious; some complain that men become promiscuous after circumcision so she will not allow you to get circumcised.
Baseline quantitative data on follow-up rates extracted from patient charts had notable gaps. An association between measurement period (baseline v. endline) and follow-up category (missing v. follow-up v. no follow-up) was found for each of the 3 follow up periods (p < 0.001 for each period, 48 hours, 7 days, and 6 weeks), indicating the proportional distribution between the 3 follow-up categories was significantly different between baseline and endline.
Post-hoc analyses reveal significant decreases in missing information on follow-up at all 3 time points (48 hours, 7 days, and 6 weeks) from baseline to endline, indicating improvement in facility data management. In spite of data limitations, we did observe significant improvement in follow-up rates at 48 hours and 7 days. This was also matched by a significant increase in no follow-up at 7 days which may be accounted for by improved record keeping rather than actual change.
Interview respondents’ reasons for not returning for follow-up:
n Perceived absence of a complication: 14 respondents who sought no follow-up reported no problems; 10 respondents who returned for follow-up at least once indicated they did not return for additional follow-up because they had no problems.
n Confusing information from facility staff about follow-up care: One respondent reported being told he was not to ride a bicycle while recovering, yet riding a bicycle was necessary for him to return to the facility for follow-up care.
Six of the 46 respondents reported a mild adverse event. Only 1 returned to the facility where he had the procedure to receive follow-up care and 2 reported seeking care elsewhere. However, had those respondents not suffered pain, they likely would not have returned for follow-up services.
QI can yield greater adherence to VMMC standards. However, improving post-circumcision follow-up rates requires not just facility-based or care-oriented changes. Efforts must be made in the community and through social networks, including but not limited to female partners, to encourage men to return for post-operative care. Recognizing that men may seek follow-up care from facilities closer to their home, health information systems should be strengthened to capture and accurately reflect follow-up rates.