Ireland United Kingdom Neth. Belarus Russia Germany Poland Bel. K a z a k h s t a n Lux. Czech Rep. Ukraine Slovakia Moldova Switz. Liech. Austria Hungary Slovenia Romania France Croatia Bosnia & Herz. Uzbekistan San Marino Serbia Andorra Mont. Georgia Corsica Kosovo Bulgaria Turkmenistan Portugal Italy Mac. Armenia Azores Spain Albania Azerbaijan
Sardinia Balearic Greece Islands Turkey Sicily
I r a n Madeira Malta Cyprus Syria Islands Crete Ireland Tunisia Lebanon Morocco United Kingdom Iraq Belarus Russia Israel Neth. Germany Poland Bel. Canary Islands Jordan K a z a k h s t a n Kuwait Lux. Czech Rep. Ukraine A l g e r i a Slovakia Moldova Switz. Liech. Austria Bahrain Western Sahara Hungary (Occupied by Morocco) L i b y a E g y p t Slovenia Romania France Croatia Qatar Bosnia & Herz. U. A. E. Uzbekistan San Marino Serbia Andorra Mont. Georgia Corsica Kosovo Bulgaria Turkmenistan Portugal Italy Mac. Armenia Azores Spain Albania Azerbaijan Mauritania Saudi Arabia Sardinia M a l i Balearic www.ploscollections.org/vmmc2011Greece Oman Islands Turkey Sicily
I r a n Senegal Madeira Malta Cyprus Syria VoluntaryIslands Medical Male Circumcision forYe HIV m e n Prevention:Crete The Cost, Gambia Morocco Tunisia Lebanon Burkina Faso Impact, and Challenges of Accelerated Scale-Up in Southern and Iraq Israel
Guinea-Bissau Canary Islands Jordan Benin Eastern Africa Kuwait Guinea Sierra Ghana Togo A l g e r i a Leone Cote d'Ivoire A UNAIDSWestern Sahara and PEPFAR Collection Bahrain (Occupied by Morocco) L i b y a E g y p t Qatar Liberia U. A. E.
Equatorial Guinea Male CircumcisionMauritania Saudi Arabia Sao Tome & Principe M a l i Uganda Prevalence Oman N i g e r Rwanda 0%–9.9% C h a d Senegal Eritrea Ye m e n Gambia 10%–19.9% Burkina Faso Tanzania Dijbouti Guinea-Bissau 20%–29.9% Benin Guinea 30%–87%Sierra Ghana Togo Leone Cote d'Ivoire
Liberia Malawi Zambia Equatorial Guinea
Sao Tome &Mozambique Principe HIV Prevalence Uganda Zimbabwe Namibia 3%–9.9% Reunion Rwanda Botswana 10%–14.9%
15%–19.9% Tanzania Swaziland 20%–26% Lesotho South Africa Malawi Zambia
Mozambique Zimbabwe Namibia Reunion Botswana
Swaziland
Lesotho South Africa
Produced with support from the Joint United Nations Programme on HIV/AIDS (UNAIDS) and the United States President’s Emergency Plan for AIDS Relief (PEPFAR).
The PLoS Medicine editors have sole editorial responsibility for the content of this collection. The views expressed in this collection are those of the authors and do not necessarily reflect the official policy or position of the U.S. Government and UNAIDS. Image credit: PEPFAR Male Circumcision Technical Working Group
Voluntary Medical Male Circumcision for HIV Prevention The Cost, Impact, and Challenges of Accelerated Scale-Up in Southern and Eastern Africa
November 2011
The views expressed in this publication do not necessarily reflect the views the U.S. Government.
Table of Contents Introduction to the UNAIDS and PEPFAR Collection……………………………………..…….… 1 Prepared Talking Points on the Collection..……………..…….….….….….….….….….….….….… 3 Voluntary Medical Male Circumcision: An Introduction to the Cost, Impact, and Challenges of Accelerated Scaling Up……………………………………………………………… 9 Catherine Hankins, Steven Forsythe, and Emmanuel Njeuhmeli Voluntary Medical Male Circumcision: Modeling the Impact and Cost of Expanding Male Circumcision for HIV Prevention in Eastern and Southern Africa…………………………… 15 Emmanuel Njeuhmeli, Steven Forsythe, Jason Reed, Marjorie Opuni, Lori Bollinger, Nathan Heard, Delivette Castor, John Stover, Timothy Farley, Veena Menon, and Catherine Hankins Voluntary Medical Male Circumcision: A Framework Analysis of Policy and Program Implementation in Eastern and Southern Africa…………………………………………. 31 Kim E. Dickson, Nhan T. Tran, Julia L. Samuelson, Emmanuel Njeuhmeli, Peter Cherutich, Bruce Dick, Tim Farley, Caroline Ryan, and Catherine A. Hankins Voluntary Medical Male Circumcision: A Cross-Sectional Study Comparing Circumcision Self-Report and Physical Examination Findings in Lesotho…………………………… 45 Anne Goldzier Thomas, Bonnie Robin Tran, Marcus Cranston, Malerato Cecilia Brown, Rajiv Kumar, and Matsotetsi Tlelai Voluntary Medical Male Circumcision: Logistics, Commodities, and Waste Management Requirements for Scale-Up of Services………………………………………………. 51 Dianna Edgil, Petra Stankard, Steven Forsythe, Dino Rech, Kristin Chrouser, Tigistu Adamu, Sameer Sakallah, Anne Goldzier Thomas, Jennifer Albertini, David Stanton, Kim Eva Dickson, and Emmanuel Njeuhmeli Voluntary Medical Male Circumcision: A Qualitative Study Exploring the Challenges of Costing Demand Creation in Eastern and Southern Africa…………………………………………. 61 Jane T. Bertrand, Emmanuel Njeuhmeli, Steven Forsythe, Sarah K. Mattison, Hally Mahler, and Catherine A. Hankins Voluntary Medical Male Circumcision: Strategies for Meeting the Human Resource Needs of Scale-Up in Southern and Eastern Africa………………………………………………… 69 Kelly Curran, Emmanuel Njeuhmeli, Andrew Mirelman, Kim Dickson, Tigistu Adamu, Peter Cherutich, Hally Mahler, Bennett Fimbo, Thembisile Khumalo Mavuso, Jennifer Albertini, Laura Fitzgerald, Naomi Bock, Jason Reed, Delivette Castor, and David Stanton Voluntary Medical Male Circumcision: Translating Research into the Rapid Expansion of Services in Kenya, 2008–2011…………………………………………………………………… 77 Zebedee Mwandi, Anne Murphy, Jason Reed, Kipruto Chesang, Emmanuel Njeuhmeli, Kawango Agot, Emma Llewellyn, Charles Kirui, Kennedy Serrem, Isaac Abuya, Mores Loolpapit, Regina Mbayaki, Ndungu Kiriro, Peter Cherutich, Nicholas Muraguri, John Motoku, Jack Kioko, Nancy Knight, and Naomi Bock Voluntary Medical Male Circumcision: Matching Demand and Supply with Quality and Efficiency in a High-Volume Campaign in Iringa Region, Tanzania…………………… 83 Hally R. Mahler, Baldwin Kileo, Kelly Curran, Marya Plotkin,Tigistu Adamu, Augustino Hellar, Sifuni Koshuma, Simeon Nyabenda, Michael Machaku, Mainza Lukobo- Durrell, Delivette Castor, Emmanuel Njeuhmeli, and Bennett Fimbo
i
Introduction to the UNAIDS and PEPFAR Collection This collection of nine new articles—published in PLoS Medicine and PLoS ONE, in conjunction with the Joint United Nations Programme on HIV/AIDS (UNAIDS) and the United States President’s Emergency Plan for AIDS Relief (PEPFAR)—highlights how scaling up voluntary medical male circumcision (VMMC) for HIV prevention in eastern and southern Africa can help prevent HIV not only at the individual level, but also at the community and population levels. Scaling up can also lead to substantial cost savings for countries due to averted treatment and care costs.
The first article by Catherine Hankins of UNAIDS, Steven Forsythe of The Futures Institute, and Emmanuel Njeuhmeli of PEPFAR/USAID, offers an introduction to the cost, impact, and challenges of accelerated scaling up and lays out the rationale for the series. This article, as well those that follow, signposts the way forward to accelerate the scaling up of VMMC service delivery safely and efficiently to reap individual-and population-level benefits.
The remaining eight papers also focus on the various factors that go into effective program expansion of VMMC, including data for decision making, policy and programmatic frameworks, logistics, demand creation, human resources, and translating research into services.
The cost savings are clear: an initial investment of US$1.5 billion between 2011 and 2015 to achieve 80% coverage of VMMC services in 14 priority countries in southern and eastern Africa and thereafter US$0.5 billion between 2016 to 2025 to maintain that coverage of 80% would result in net savings of US$16.5 billion between 2011 and 2025. However, as the articles in the collection show, strong political leadership, country ownership, and stakeholder engagement, along with effective demand creation, community mobilisation, and human resource deployment are essential to effectively expanding and maintaining VMMC programs.
The views expressed in this collection are those of the authors and do not necessarily reflect the official policy or position of the U.S. Government and UNAIDS. These articles were produced with support from UNAIDS and PEPFAR. The PLoS Medicine editors have sole editorial responsibility for the content of this collection.
The series was published on November 29, 2011 in PLoS Medicine and PLoS ONE. All articles can be found online via the Public Library of Science at: http://www.ploscollections.org/VMMC2011.
1 2 Prepared Talking Points on the Collection
Voluntary Medical Male Circumcision: estimated the impact and cost of scaling up An Introduction to the Cost, Impact, and adult VMMC using updated, country-specific data. Challenges of Accelerated Scaling Up • Given its demonstrated acceptability, low cost, • The Male Circumcision: Decision Makers’ and potential impact when provided by well- Program Planning Tool (DMPPT) was used by trained, well-equipped providers in hygienic the authors to model the impact and cost of settings, voluntary medical male circumcision scaling up adult VMMC in Botswana, Lesotho, (VMMC) should be scaled up rapidly in high Malawi, Mozambique, Namibia, Rwanda, South HIV prevalence settings to reap both individual- Africa, Swaziland, Tanzania, Uganda, Zambia, and population-level HIV prevention benefits. Zimbabwe, and Nyanza province in Kenya. Epidemiologic and demographic data from • The opportunity costs of not taking action recent household surveys for each country was now are too high to ignore. VMMC scale up is used. cost saving and creates fiscal space in the future that otherwise would have been encumbered • The cost of VMMC ranges from US$ 65.85- by antiretroviral treatment costs. Net savings US$ 95.15 per VMMC performed, based on of US$16.6 billion are expected from an a cost assessment of VMMC services aligned investment of US$1.5 billion between 2011 and with the World Health Organization’s (WHO’s) 2015 to achieve 80% coverage in 13 priority considerations of models for optimizing volume countries. and efficiencies.
• Strong political leadership, country ownership, • Results from the DMPPT models suggest that and stakeholder engagement, along with scaling up adult VMMC to reach 80% coverage effective demand creation, community in the 13 countries by 2015 would entail mobilization, and human resource deployment, performing 20.33 million circumcisions between are essential ingredients of successful scale-up 2011 and 2015 and an additional 8.42 million strategies. between 2016 and 2025 (to maintain the 80% coverage). • In this collection of articles on the cost and impact of VMMC for HIV prevention, we • The impact of such a scale-up would result highlight progress to date, identify challenges in averting 3.36 million new HIV infections to overcome, and signpost the way forward to through 2025. In addition, while the model scaling up VMMC service delivery safely and shows that this scale-up would cost a total of efficiently. US$2 billion between 2011 and 2025, it would result in net savings (due to averted treatment and care costs) amounting to US$16.51 billion. Voluntary Medical Male Circumcision: Modeling the Impact and Cost of • The study suggests that rapid scale-up of VMMC in eastern and southern Africa is Expanding Male Circumcision for HIV warranted based on the likely impact on the Prevention in Eastern and Southern Africa region’s HIV epidemics and net savings. Scaling • There is strong evidence showing that voluntary up of safe, VMMC in eastern and southern medical male circumcision (VMMC) reduces Africa will lead to a substantial reduction in HIV incidence in men. To inform the VMMC HIV infections in the countries and lower health policies and goals of 13 priority countries system costs through averted HIV-care costs. in eastern and southern Africa, the authors
3 Voluntary Medical Male Circumcision: ensure that factors critical to supporting and A Framework Analysis of Policy accelerating scale-up are incorporated into the programme. The most successful programme and Program Implementation in required country ownership and sustained Eastern and Southern Africa leadership to translate research into a national • Following confirmation of the effectiveness policy and programme. of medical male circumcision for HIV prevention (MMC), WHO/UNAIDS issued Voluntary Medical Male Circumcision: recommendations in 2007. Less than five years later, priority countries are at different stages of A Cross-Sectional Study Comparing programme scale-up. Self-Report and Physical Examination Findings in Lesotho • This paper analyses the progress towards the scale-up of MMC programmes. It analyses • Overwhelming evidence including three clinical the adoption of MMC as an additional HIV trials shows that medical male circumcision prevention strategy and explores the factors (MMC) reduces the risk of HIV infection may have expedited or hindered the adoption of among men. However, data from recent policies and initial programme implementation Lesotho Demographic and Health Surveys in priority countries to date. (LDHSs) do not demonstrate male circumcision (MC) to be protective against HIV. • MMCs performed in priority countries between 2008 and 2010 were recorded and used to • These contradictory findings could partially be classify countries into five adopter categories due to inaccurate self-reported MC status used according to the Diffusion of Innovations to estimate MC prevalence. This study describes framework. The main predictors of MMC MC characteristics among men applying for programme adoption were determined and Lesotho Defence Force (LDF) recruitment factors influencing subsequent scale-up and seeks to assess MC self-reported accuracy explored. through comparison with physical examination- based data. • By the end of 2010, over 550,000 MMC’s have been performed representing approximately 3% • During LDF applicant screening in 2009, of the target coverage level in priority countries. 241 (77%) of 312 men, aged 18 - 25 years, consented to a self-administered demographic • The ‘early adoptor’ countries developed national and MC characteristic survey and physician- MMC policies and initiated MMC programme performed genital examination. The extent of implementation soon after the release of the foreskin removal was graded on a scale of 1 (no WHO Recommendations. evidence of MC) to 4 (complete MC).
• However, only Kenya appears to be on track • MC was self-reported by 27% (n = 64/239) towards achievement of the DMPPT estimated of participants. Of the 64 men self-reporting 80% coverage goal by 2015, having already MC, physical exam showed that 23% had no achieved 61.5% of the DMMPT target. None evidence of circumcision, 27% had partial, and of the other countries appear to be on track 50% had complete circumcision. to achieve their targets. Potential predicators of early adoption of MC programmes include • Of the MCs reportedly performed by a medical having a MMC focal person, establishing provider, 3% were Grade 1 and 73% were national policy, having an operational strategy Grade 4. Of the MCs reportedly performed and the establishment of a pilot programme. by traditional circumcisers, 41% were Grade 1, while 28% were Grade 4. Among participants • Early adoption of MMC policies did not self-reporting being circumcised, the odds necessarily result in rapid programme scale- of MC status misclassification were 7 times up. A key lesson is the importance of not only higher among those reportedly circumcised by being ready to adopt a new intervention but to
4 initiation school personnel (odds ratio (OR) • Previous costing studies of VMMC programs = 7.22; 95% confidence interval (CI) = 2.29– did not capture supply chain costs, the full range 22.75). of commodities needed for VMMC program implementation or waste management. The • Approximately 27% of participants self- authors estimates indicate that depending upon reported being circumcised. However, only the volume of services provided supply chain 50% had complete MC as determined by a and waste management, including commodities physical exam. Given this low MC self-report and associated labor, contribute between accuracy, countries scaling up MMC should $58.92USD and $73.57USD to the cost of obtain physical exam-based MC data to guide performing one adult male circumcision in service delivery and cost estimates. HIV Swaziland. prevention messages promoting MMC should provide comprehensive education regarding the • Experience with the VMMC program in definition of MMC. Swaziland indicates that supply chain and waste management add approximately $60 per circumcision. Voluntary Medical Male Circumcision: Logistics, Commodities, and Waste • Therefore, program planners and policymakers Management Requirements for should consider the significant contribution of Scale-Up of Services supply chain and waste management to VMMC program costs as they determine future resource • The global HIV prevention community needs for VMMC programs. is implementing voluntary medical male circumcision (VMMC) programs across Eastern and Southern Africa with a goal of reaching Voluntary Medical Male Circumcision: 80% coverage in adult males by 2015. Successful A Qualitative Study Exploring the implementation will depend on the accessibility Challenges of Costing Demand Creation of commodities essential for VMMC in Eastern and Southern Africa programming and the appropriate allocation of resources to support the VMMC supply chain. • This paper proposes an approach to estimating the costs of demand creation for voluntary • The U.S. President’s Emergency Plan for medical male circumcision (VMMC) scale-up in AIDS Relief programs, in collaboration with 13 countries of Eastern and Southern Africa. the World Health Organization and the Joint It addresses two key questions: (1) what are the United Nations Programme on HIV/AIDS, has elements of a standardized package for demand developed a standard list of commodities for creation? and (2) What challenges exist and must VMMC programs. be taken into account in estimating the costs of demand creation? • This list was used to inform program planning for a one-year program to circumcise 152,000 • Study was conducted using key informant adult men in Swaziland. During this process, interviews with communications experts in additional key commodities were identified, priority countries. expanding the standard list to include commodities for waste management, HIV • The key finding was the wide range of views, counseling and testing, and the treatment of suggesting that a standard package of core sexually transmitted infections. demand creation elements would not be universally applicable. This underscored the • The approximate costs for the procurement of importance of tailoring demand creation commodities, management of a supply chain, strategies and estimates to specific country and waste disposal, were determined for the contexts before estimating costs. The key VMMC program in Swaziland using current informant interviews, supplemented by the market prices of goods and services. researchers’ field experience, identified these issues to be addressed in future costing
5 exercises: variations in the cost of VMMC • These approaches provide models for other demand creation activities by country and countries to replicate and adapt. program, decisions about the quality and comprehensiveness of programming, and lack • Importantly, the effect of VMMC scale up on of data on critical elements needed to “trigger other health services is not known and must be the decision” among eligible men. investigated in future studies.
• The case of demand creation in a campaign setting in Iringa, Tanzania is discussed - Voluntary Medical Male Circumcision: revealing the rationale behind selection of Translating Research into the Rapid communication elements in a field setting. Expansion of Services in Kenya, 2008–2011 • Based on this study’s findings, the authors • Kenya’s male circumcision for HIV prevention propose a seven-step methodological approach policy prioritizes Nyanza Province, the to estimate the cost of VMMC scale-up in a region with the highest HIV burden and low priority country, based on key assumptions. circumcision rates, for scale up of voluntary However, further work is needed to better medical male circumcision (VMMC) services. understand core components of a demand creation package and how to cost them. • Since the policy’s implementation in October Notwithstanding the methodological challenges, 2008, approximately 290, 000 adult males have estimating the cost of demand creation remains been circumcised in Kenya, most of them in an essential element in deriving estimates of the Nyanza. total costs for VMMC scale-up in Eastern and Southern Africa. • Government leadership and a documented implementation strategy have been key factors in Kenya’s rapid scale up of VMMC. Voluntary Medical Male Circumcision: • A third key factor has been program flexibility; Strategies for Meeting the Human the introduction of innovative approaches Resource Needs of Scale-Up in including task shifting, short intensive service Southern and Eastern Africa campaigns, and most recently diathermy for hemostasis, have all helped the program respond • Scaling up voluntary medical male circumcision to challenges. (VMMC) could avert millions of HIV infections in Southern and Eastern Africa but shortages of • Kenya’s successful approach to VMMC scale up health professionals are likely to limit progress. provides a model that other countries can adapt to their own circumstances. • Potential responses to this human resource challenge include task-shifting, task-sharing, temporary redeployment of public-sector Voluntary Medical Male Circumcision: staff during VMMC campaigns, expansion of Matching Demand and Supply with the health workforce through recruitment of unemployed, recently retired, newly graduated Quality and Efficiency in a High-Volume or on-leave healthcare workers, and the use of Campaign in Iringa, Tanzania foreign volunteer medical staff. • The Government of Tanzania has adopted • Approaches to solving the human resource voluntary medical male circumcision (VMMC) challenge associated with VMMC scale up that as an important component of its HIV have already been implemented include: moving prevention strategy and aims to reach 2.8 public sector clinicians to high volume VMMC million uncircumcised men within the next three sites during campaigns (Tanzania), empowering years. nurses to conduct VMMC surgery (Kenya); and identifying untapped reserves of qualified nurses (Swaziland).
6 • In June and July 2010 a 6-week VMMC campaign in Tanzania’s Iringa region performed 10, 352 circumcisions.
• Strategies adopted by the campaign to generate demand included the widespread dissemination of messages focused on the provision of free VMMC by specially trained health-care providers and on the HIV-prevention benefits of VMMC.
• Clinical efficiency was improved through, for example, the use of multiple beds in an assembly line, and the efficient use of staff time through task shifting and task sharing.
• The experiences of this campaign suggest that high volume VMMC can be performed without compromising client safety and provide a model for matching supply and demand for VMMC services elsewhere.
7 8 Review Voluntary Medical Male Circumcision: An Introduction to the Cost, Impact, and Challenges of Accelerated Scaling Up
Catherine Hankins1*, Steven Forsythe2, Emmanuel Njeuhmeli3 1 Joint United Nations Programme on HIV/AIDS, Geneva, Switzerland, 2 Futures Institute, Glastonbury, Connecticut, United States of America, 3 United States Agency for International Development, Washington, District of Columbia, United States of America
southern Africa with settings of high HIV prevalence and low Abstract: Scaling up voluntary medical male circumci- levels of male circumcision [12,13]. sion (VMMC) for HIV prevention is cost saving and creates Male circumcision is the oldest and most common surgical fiscal space in the future that otherwise would have been procedure. With 30% of men globally and 67% of men in sub- encumbered by antiretroviral treatment costs. An invest- Saharan Africa circumcised [14], social and cultural factors are the ment of US$1,500,000,000 between 2011 and 2015 to main determinants of acceptability [6,15–18]. In sub-Saharan achieve 80% coverage in 13 priority countries in southern Africa, male circumcision was found to be acceptable to men and and eastern Africa will result in net savings of women in non-circumcising communities if readily accessible and US$16,500,000,000. Strong political leadership, country provided safely [18]. Mathematical modelling has shown that ownership, and stakeholder engagement, along with medical male circumcision is highly cost-effective, with costs to effective demand creation, community mobilisation, and human resource deployment, are essential. This collection avert one HIV infection ranging from US$150 to US$900 using a of articles on determining the cost and impact of VMMC ten-year time horizon, and one new HIV infection averted for for HIV prevention signposts the way forward to scaling every five to fifteen procedures performed [19]. up VMMC service delivery safely and efficiently to reap Given these levels of acceptability, cost, and potential impact, individual- and population-level benefits. VMMC provided by well-trained, well-equipped providers in hygienic settings should be scaled up rapidly in high HIV prevalence settings to reap both individual- and population-level Male Circumcision for HIV Prevention benefits. Policy makers and program planners interested in moving services to scale are faced, however, with decisions about which Despite a 22-fold increase since 2001 in the number of people populations to prioritise (newborns, adolescents, adults, men at receiving antiretroviral therapy, two people acquire HIV infection higher risk of HIV exposure, such as those in serodiscordant for every person starting treatment [1]. Scaling up evidence- couples), what service delivery models to use (fixed, outreach, informed HIV prevention programs is imperative. In the HIV mobile), which human resources to deploy (physicians, clinical prevention toolbox of behavioural, biomedical, and structural officers, nurses), how to create demand and match supply to it, and approaches to combine for maximum effect [2], VMMC is an what speed of scale-up is both desirable and feasible. essential tool in all high HIV prevalence, predominantly Policy makers in high HIV prevalence countries will find heterosexual epidemic settings. It provides lifelong partial valuable information to support rapid scale-up of services in this protection for men against HIV infection [3–6] and reduces their collection on the cost and impact of VMMC for HIV prevention. likelihood of genital ulcers [7,8], syphilis [9], and penile cancer The articles highlight progress to date, explore challenges to [10]. Observational data and ecological studies have suggested for decades that male circumcision provides a level of protection from Citation: Hankins C, Forsythe S, Njeuhmeli E (2011) Voluntary Medical Male HIV infection for men [11]. Three randomised controlled trials Circumcision: An Introduction to the Cost, Impact, and Challenges of Accelerated Scaling Up. PLoS Med 8(11): e1001127. doi:10.1371/journal.pmed.1001127 [3–5] conducted in the last decade found a 57% protective effect Academic Editor: Stephanie L. Sansom, Centers for Disease Control and against HIV for men who became circumcised [6]. All three trials Prevention, United States of America were stopped prematurely because it was deemed unethical to withhold VMMC from men in the control arm waiting to be Published November 29, 2011 circumcised. This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for Following trial results, the World Health Organization (WHO) any lawful purpose. The work is made available under the Creative Commons CC0 and the Joint United Nations Programme on HIV/AIDS public domain dedication. (UNAIDS) rapidly convened stakeholders in March 2007 to Funding: The funding source for this manuscript is PEPFAR through USAID evaluate the strength of the evidence and to consider the policy Health Policy Initiative and UNAIDS. Technical staff from USAID and UNAIDS were and programmatic implications. The resulting recommendations involved in the design, data collection and analysis, decision to publish, and addressed the essential components for program implementation preparation of this manuscript. in 13 priority countries (Botswana, Kenya, Lesotho, Malawi, Competing Interests: The authors have declared that no competing interests exist. Mozambique, Namibia, Rwanda, South Africa, Swaziland, Tanzania, Uganda, Zambia, and Zimbabwe) in eastern and Abbreviations: PEPFAR, United States President’s Emergency Plan for AIDS Relief; UNAIDS, Joint United Nations Programme on HIV/AIDS; VMMC, voluntary medical male circumcision; WHO, World Health Organization
Review articles synthesize in narrative form the best available evidence on a topic. * E-mail: [email protected] Provenance: Submitted as part of a sponsored Collection; internally reviewed.
9 overcome, and provide solutions to facilitate scale-up. By moving actively seek out VMMC services. Timely matching of supply to forward now, leaders will set their countries on course to achieve demand is critical to avoid men seeking unsafe procedures because the 50% reduction in sexual transmission of HIV by 2015 that waiting times are too long. Mahler et al. [32] describe how this they signed on to at the United Nations General Assembly in June challenge was met while ensuring service quality and efficiency in 2011 [20]. Iringa, Tanzania. As Bertrand et al. [33] underscore, demand creation strategies must be tailored to specific country contexts to Numbers of Male Circumcisions Needed determine the most effective mix of mass media, small media, outreach, and community mobilisation communication approach- How many medical male circumcisions will need to be es. Recognising that no standard package applies in all settings, performed in the 13 priority countries for maximum impact? they propose a seven-step methodology to estimate demand The first step towards answering this question is assessing baseline creation costs for inclusion in country VMMC costing estimates. prevalence of male circumcision in each country, recognising that The United States President’s Emergency Plan for AIDS Relief self-reported circumcision status, as Thomas et al. [21] demon- (PEPFAR) (through the US Agency for International Development strate for Lesotho, can be inaccurate. Njeuhmeli et al. [22] Health Policy Initiative) and the UNAIDS supported primary data estimate the number of VMMC procedures needed to reach 80% collection activities to estimate facility-based unit costs for prevalence using the Decision Makers’ Program Planning Tool VMMC, as currently delivered, in Uganda, Kenya, Zambia, [23] for male circumcision, an interactive tool that incorporates Zimbabwe, South Africa, and Namibia. These estimates were country-specific demographics, epidemic dynamics, and locally revised to include a more comprehensive assessment of cost derived cost estimates. An estimated 20.3 million circumcisions components, including waste management, supply chain manage- among men 15–49 years of age are needed to close the gap by ment, training, and overhead costs. Assumptions were made about 2015. Once eligible uncircumcised men have been reached future task shifting and task sharing to optimise the volume and through ‘‘catch-up’’ campaigns, annual costs fall because ongoing efficiency of services [31]. Unit cost estimates, excluding demand programs need only maintain coverage while shifting to long-term creation costs, were estimated at US$80.13 per adult VMMC. sustainable strategies. These could include systematically offering Country unit costs were then adjusted to account for differences in the procedure to one cohort per year, such as young men turning labour costs, and sensitivity analyses assumed a 20% higher and 15 years of age or all male newborns. 20% lower unit cost. As Mwandi et al. [24] detail, Kenya is on track, with over 66% Combining unit cost data with the numbers of VMMCs needed of its target for Nyanza Province met through sustained to achieve 80% coverage in all 13 countries results in an estimated government leadership, financial and technical support of US$1,500,000,000 required between 2011 and 2015. Maintaining international donors, and a partnership strategy that has engaged 80% coverage in all 13 countries between 2016 and 2025 would a broad array of stakeholders. Swaziland has reached 13%, require an additional US$500,000,000. Sensitivity analyses under- Zambia and Botswana have attained 4%, and Tanzania and score the importance of minimising costs while ensuring safe, high- South Africa have met 3% of their respective 2015 80% objectives, quality procedures accompanied by effective HIV prevention through campaigns that picked up speed in 2010 (see Figure 1). messages. If unit costs were 20% higher, an additional US$500,000,000 would be required over the full 2011–2025 period, Efficiently Mobilising Human Resources while a 20% decrease in unit costs would reduce overall resources Among the impediments to scaling up VMMC services are required by US$500,000,000. Given the estimated discounted concerns about human resources. Curran et al. [25] review the individual lifetime cost of antiretroviral therapy of US$7,400, concepts of task shifting and task sharing and describe approaches VMMC is not only cost-effective in these countries—it is cost to expanding the health sector workforce through redeployment of saving. Net savings from 2011 to 2015 due to averted treatment and existing personnel and use of expatriate volunteers, drawing on care costs amount to US$16,5 00,000,000. Therefore, an initial experiences in Tanzania, Kenya, and Swaziland. Scaling up investment now in VMMC, although substantial, will be returned VMMC requires a streamlined campaign footing and rapid many fold. It will create fiscal space in the future that otherwise mobilisation of human resources to eventually achieve a steady would have been encumbered by antiretroviral treatment costs. state that will make fewer demands on the health care system. Both surgical and non-surgical efficiency gains have significant Estimating the Impact impacts on per procedure costs and therefore on the numbers of medical male circumcisions that can be performed. Significant Just how many infections could be averted by scaling up VMMC efforts are underway to test a number of medical devices that, if to reach 80% male circumcision prevalence in all countries in eastern found to be safe, acceptable, and effective, would dramatically and southern Africa in five years and maintaining that coverage reduce procedure times [26–28] and speed the scale-up. through to 2025? Where would the largest impact be seen? Through 2025, 3.4 million new HIV infections would be averted, as Figure 2 Determining Costs of Scale-Up shows, with South Africa alone averting over 1 million new HIV infections between 2011 and 2025. More than 20% of projected new VMMC scale-up planning requires cost estimates for commod- HIV infections would be averted in Botswana, Lesotho, Malawi, ity procurement, supply chain management, and disposal of waste Namibia, Rwanda, Swaziland, Uganda, Zambia, and Zimbabwe. generated by male circumcision programs. Edgil et al. [29] The number of VMMCs needed to avert one HIV infection ranges describe a costing framework for these program components for from a low of four in Zimbabwe to a high of 44 in Rwanda. Swaziland, emphasising the advantages of a standard kit of the Among the infections averted are those among women who consumables and dedicated instruments necessary for one male benefit indirectly from VMMC scale-up. As more men become circumcision procedure [30,31]. circumcised, women are less likely to encounter sexual partners VMMC program planning necessitates the design and imple- who have HIV infection. Eventually, even uncircumcised men will mentation of effective demand creation strategies that encourage benefit indirectly from VMMC scale-up [19]. Early on, most HIV men to consider male circumcision for HIV prevention and infections averted occur among men, but the proportion among
10 Figure 1. Achievement towards target of 80% coverage. This figure illustrates that most countries have had only limited success in bridging the gap between historical male circumcision levels and the 80% target. The one exception is Kenya, which has achieved more than 66% of its objective, primarily in Nyanza Province. doi:10.1371/journal.pmed.1001127.g001
Figure 2. Cumulative number and percentage of HIV infections averted between 2011 and 2025 by scaling up adult VMMC to reach 80% coverage in five years. This figure illustrates the significant impact that achieving 80% VMMC coverage of 15- to 49-year-old men would have on the epidemics in 13 countries in eastern and southern Africa. South Africa can avert the largest number of HIV infections (over 1 million between 2011 and 2025); Zimbabwe can avert the highest percentage of new HIV infections (almost 42%). More than 20% of new HIV infections would be averted between 2011 and 2025 in nine countries: Botswana, Lesotho, Malawi, Namibia, Rwanda, Swaziland, Uganda, Zambia, and Zimbabwe. Nyanza refers to Nyanza Province in Kenya: the data presented are only for Nyanza Province in Kenya, as this is the only province in Kenya with prevalence of male circumcision lower than 80% and is the province with the highest HIV prevalence compared to the national average. doi:10.1371/journal.pmed.1001127.g002
11 women steadily increases over time until almost half of all HIV 2011 political declaration on HIV/AIDS [20]. It will take community infections averted in the year 2025 are those that would have conversations to create new social norms about male circumcision in occurred among women. previously non-circumcising communities. It will take women speaking The anticipated impact of scaling up VMMC is directly out on the HIV prevention benefits and the desirability of male proportional to the pace and scale of implementation. Decreasing circumcision for their sexual partners, brothers, and sons. It will take VMMC coverage targets from 80% to 50% results in a decline in tailored communication strategies to create demand for VMCC the total number of HIV infections averted from 3.4 million to 1.1 services, and it will take program planners who have anticipated million. Reducing the time to achieve 80% coverage from five increasing demand and are matching it with supply of safe, acceptable, years to one year has the opposite effect, increasing the number of and accessible services. It will take continued innovations in medical HIV infections averted from 3.4 million to 4.1 million. This device development and testing, along with other efficiency gains, to decreases the cost per HIV infection averted and increases total decrease procedure times and thereby increase access. cost savings due to HIV infections averted. Maximum epidemic Above all, it will take vision by government leaders who impact accrues the more quickly maximum coverage is achieved. understand that promoting effective VMMC programming now will create synergies to more rapidly halt and reverse their Determinants of Early Adoption and Sustained countries’ epidemics. Scaling up VMMC requires considerable Scale-Up short-term investment of financial and human resources to accelerate gains toward coverage objectives. Once achieved, Some countries need to offer medical male circumcision to countries will arrive at a sustained and sustainable ‘‘cruising’’ fewer than 500,000 men and could achieve 80% coverage in less level requiring far fewer resources to maintain 80% coverage. The than two years, while those that have to offer circumcision to more challenge before them is to climb over the ‘‘catch-up’’ hump to than 2–3 million men will take longer. However, it is not just the reap substantive HIV prevention benefits. The faster that size of the task that will determine whether objectives are met, impact is realised, and cost savings occur. As Dickson et al. [34] countries do this, the more rapidly will both direct benefits for demonstrate, key barriers and facilitators are influencing the speed men and indirect benefits for women accrue. Six million people of scale-up. Characteristics of the response to the compelling are on antiretroviral therapy, 9 million more are eligible today, scientific evidence of the HIV prevention benefits of VMMC in according to the current WHO recommendations of providing each of the 13 countries permit a classification of countries into antiretroviral treatment for patients with CD4 cell counts of ,350 four categories: innovators, early adopters, early and late majority, cells/ml [35], and pressure is mounting to offer even more people and laggards. Against a backdrop of varying sociopolitical and treatment for prevention [1,36]. The opportunity costs of not cultural contexts, the key drivers of early adoption and sustained taking action to scale up VMMC safely and rapidly now to prevent scale-up are country ownership, explicit political leadership, new HIV infections and create fiscal space are too high to ignore. engagement of stakeholders, and community mobilisation. This is on our watch—what role is each one of us playing to ensure that VMMC contributes fully to halting and reversing the HIV Conclusion epidemic in eastern and southern Africa? What will it take for the citizens of the 13 priority countries in eastern Author Contributions and southern Africa to reap the prevention benefits of VMMC? Domestic funding can be mobilised and international funding accessed Conceived and designed the experiments: CH SF EN. Performed the experiments: CH SF EN. Analyzed the data: CH SF EN. Contributed through PEPFAR, the Bill & Melinda Gates Foundation, and the reagents/materials/analysis tools: CH SF EN. Wrote the first draft of the Global Fund to Fight AIDS, Tuberculosis and Malaria. However, it manuscript: CH. Contributed to the writing of the manuscript: SF EN. will take leadership and visible champions at all levels to mobilise and ICMJE criteria for authorship read and met: CH SF EN. Agree with deploy this funding for maximum effect. This is precisely the ‘‘decisive, manuscript results and conclusions: CH SF EN. inclusive, and accountable leadership’’ called for in the United Nations’
References 1. Joint United Nations Programme on HIV/AIDS (2011) AIDS at 30: nations at 9. Tobian AAR, Serwadda D, Quinn TC, Kigozi G, Gravitt PE, et al. (2009) Male the crossroads. Geneva: Joint United Nations Programme on HIV/AIDS. circumcision for the prevention of HSV-2 and HPV infections and syphilis. 2. Hankins CA, de Zalduondo BO (2010) Combination prevention: a deeper N Engl J Med 360: 1298–1309. understanding of effective HIV prevention. AIDS 24: S70–S80. doi:10.1097/ 10. Morris BJ, Gray RH, Castellsague X, Bosch XF, Halperin DT, et al. (2011) The 1001.aids.0000390709.0000304255.fd. strong protective effect of circumcision against cancer of the penis. Adv Urol 3. Auvert B, Taljaard D, Lagarde E, Sobngwi-Tambekou J, Sitta R, et al. (2005) 2011: 812368. Randomized, controlled intervention trial of male circumcision for reduction of 11. Weiss HA, Quigley MA, Hayes RJ (2000) Male circumcision and risk of HIV HIV infection risk: The ANRS 1265 Trial. PLoS Med 2: e298. doi:10.1371/ infection in sub-Saharan Africa: a systematic review and meta-analysis. AIDS journal.pmed.0020298. 14: 2361–2370. 4. Bailey RC, Moses S, Parker CB, Agot K, Maclean I, et al. (2007) Male circumcision 12. World Health Organization, Joint United Nations Programme on HIV/AIDS for HIV prevention in young men in Kisumu, Kenya: a randomised controlled trial. (2007) New data on male circumcision and HIV prevention: policy and Lancet 369: 643–656. programmes implications—WHO-UNAIDS recommendations. Geneva: World 5. Gray R, Kigozi G, Serwadda D, Makumbi F, Watya S, et al. (2007) Male Health Organization, Joint United Nations Programme on HIV/AIDS. circumcision for HIV prevention in men inRakai,Uganda:arandomisedtrial. 13. Weiss HA, Halperin D, Bailey RC, Hayes RJ, Schmid G, et al. (2008) Male Lancet 369: 657–666. circumcision for HIV prevention: from evidence to action? AIDS 22: 567–574. 6. Weiss HA, Dickson KE, Agot K, Hankins CA (2010) Male circumcision for HIV doi:510.1097/QAD.1090b1013e3282f1093f1406. 14. Joint United Nations Programme on HIV/AIDS, World Health Organization prevention: current research and programmatic issues. AIDS 24: S61–S69. (2007) Male circumcision: global trends and determinants of prevalence, safety, doi:10.1097/1001.aids.0000390708.0000366136.f0000390704. and acceptability. Geneva: World Health Organization. 7. Weiss H, Thomas S, Munabi S, Hayes R (2006) Male circumcision and risk of 15. Halperin DT, Fritz K, McFarland W, Woelk G (2005) Acceptability of adult syphilis, chancroid, and genital herpes: a systematic review and meta-analysis. male circumcision for sexually transmitted disease and HIV prevention in Sex Transm Infect 82: 101–110. Zimbabwe. Sex Transm Dis 32: 238–239. 8. Nasio JM, Nagelkerke NJ, Mwatha A, Moses S, Ndinya-Achola JO, et al. (1996) 16. Kebaabetswe P, Lockman S, Mogwe S, Mandevu R, Thior I, et al. (2003) Male Genital ulcer disease among STD clinic attenders in Nairobi: association with circumcision: an acceptable strategy for HIV prevention in Botswana. Sex HIV-1 and circumcision status. Int J STD AIDS 7: 410–414. Transm Infect 79: 214–219.
12 17. Lagarde E, Dirk T, Puren A, Reathe RT, Bertran A (2003) Acceptability of male 27. Bitega JP, Ngeruka ML, Hategekimana T, Asiimwe A, Binagwaho A (2011) Safety circumcision as a tool for preventing HIV infection in a highly infected and efficacy of the PrePex(TM) device for rapid scale up of male circumcision for community in South Africa. AIDS 17: 89–95. HIV prevention in resource-limited settings. J Acquir Immune Defic Syndr;E-pub 18. Westercamp N, Bailey RC (2007) Acceptability of male circumcision for ahead of print. doi:10.1097/QAI.1090b1013e3182354e3182365. prevention of HIV/AIDS in sub-Saharan Africa: a review. AIDS Behav 11: 28. Musau P, Demirelli M, Muraguri N, Ndwiga F, Wainaina D, et al. (2011) The 341–355. safety profile and acceptability of a disposable male circumcision device in 19. UNAIDS/WHO/SACEMA Expert Group on Modelling the Impact and Cost Kenyan men undergoing voluntary medical male circumcision. J Urol 186: of Male Circumcision for HIV Prevention (2009) Male circumcision for HIV 1923–1927. prevention in high HIV prevalence settings: what can mathematical modelling 29. Edgil D, Stankard PS, Forsythe S, Reed J, Rech D, et al. (2011) Voluntary contribute to informed decision making? PLoS Med 6: e1000109. doi:10.1371/ medical male circumcision: logistics, commodities, and waste management journal.pmed.1000109. requirements for scale-up of services. PLoS Med 8: e1001128. doi:10.1371/ 20. United Nations General Assembly (2011 June 10) Political declaration on HIV/ journal.pmed.1001128. AIDS: intensifying our efforts to eliminate HIV/AIDS. Resolution adopted by 30. The United States President’s Emergency Plan for AIDS Relief Male the General Assembly on 10 June 2011. Resolution A/Res/65/277. Circumcision Technical Working Group (2009) PEPFAR Male Circumcision 21. Thomas AG, Tran BR, Cranston M, Brown MC, Kumar R, et al. (2011) Partners’ Meeting: commodities and improved coordination of male circumci- Voluntary medical male circumcision: a cross-sectional study comparing sion for HIV prevention. Washington (District of Columbia): The United States circumcision self-report and physical examination findings in Lesotho. PLoS President’s Emergency Plan for AIDS Relief. ONE 6: e0027561. doi:10.1371/journal.pone.002756. 31. World Health Organization (2010) Considerations for implementing models for 22. Njeuhmeli E, Forsythe S, Reed J, Opuni M, Bollinger L, et al. (2011) Voluntary optimizing the volume and efficiency of male circumcision services. Geneva: medical male circumcision: modeling the impact and cost of expanding male World Health Organization. 56 p. circumcision for HIV prevention in eastern and southern Africa. PLoS Med 8: 32. Mahler HR, Kileo B, Curran K, Plotkin M, Adamu T, et al. (2011) Voluntary e1001132. doi:10.1371/journal.pmed.1001132. medical male circumcision: matching demand and supply with quality and 23. United States Agency for International Development Health Policy Initiative efficiency in a high-volume campaign in Iringa Region, Tanzania. PLoS Med 8: (2009) Decision makers’ program planning tool (DMPPT): calculating the costs e1001131. doi:10.1371/journal.pmed.1001131. and impacts of a male circumcision program. Washington (District of 33. Bertrand JT, Njeuhmeli E, Forsythe S, Mattison SK, Chideya S, et al. (2011) Columbia): United States Agency for International Development Health Policy Voluntary medical male circumcision: a qualitative study exploring the Initiative. challenges of costing demand creation in eastern and southern Africa. PLoS 24. Mwandi Z, Murphy A, Reed J, Chesang K, Njeuhmeli E, et al. (2011) Voluntary ONE 6: e0027562. doi:10.1371/journal.pone.0027562. medical male circumcision: translating research into the rapid expansion of 34. Dickson KE, Tran NT, Samuelson JL, Njeuhmeli E, Reed J, et al. (2011) services in Kenya, 2008–2011. PLoS Med 8: e1001130. doi:10.1371/ Voluntary medical male circumcision: a framework analysis of policy and journal.pmed.1001130. program implementation in eastern and southern Africa. PLoS Med 8: 25. Curran K, Njeuhmeli E, Mirelman A, Dickson K, Adamu T, et al. (2011) e1001133. doi:10.1371/journal.pmed.1001133. Voluntary medical male circumcision: strategies for meeting the human resource 35. World Health Organization (2010) Antiretroviral therapy for HIV infection in needs of scale-up in southern and eastern Africa. PLoS Med 8: e1001129. adults and adolescents. Recommendations for a public health approach, 2010 doi:10.1371/journal.pmed.1001129. revision. Geneva: World Health Organization. 26. Barone MA, Ndede F, Li PS, Masson P, Awori Q, et al. (2011) The Shang Ring 36. Cohen MS, Chen YQ, McCauley M, Gamble T, Hosseinipour MC, et al. (2011) device for adult male circumcision: a proof of concept study in Kenya. J Acquir Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med Immune Defic Syndr 57: e7–e12. doi:10.1097/QAI.1090b1013e3182158967. 365: 493–505.
13 14 Voluntary Medical Male Circumcision: Modeling the Impact and Cost of Expanding Male Circumcision for HIV Prevention in Eastern and Southern Africa
Emmanuel Njeuhmeli1*., Steven Forsythe2., Jason Reed3, Marjorie Opuni4, Lori Bollinger2, Nathan Heard5, Delivette Castor1, John Stover2, Timothy Farley6, Veena Menon7, Catherine Hankins8 1 United States Agency for International Development, Washington, District of Columbia, United States of America, 2 Futures Institute, Glastonbury, Connecticut, United States of America, 3 Centers for Disease Control and Prevention, Atlanta, Georgia, United States of America, 4 UNAIDS, Geneva, Switzerland, 5 Office of the U.S. Global AIDS Coordinator, United States Department of State, Washington, District of Columbia, United States of America, 6 World Health Organization, Geneva, Switzerland, 7 Futures Group, Washington, District of Columbia, United States of America, 8 Joint United Nations Programme on HIV/AIDS, Geneva, Switzerland
Abstract
Background: There is strong evidence showing that voluntary medical male circumcision (VMMC) reduces HIV incidence in men. To inform the VMMC policies and goals of 13 priority countries in eastern and southern Africa, we estimate the impact and cost of scaling up adult VMMC using updated, country-specific data.
Methods and Findings: We use the Decision Makers’ Program Planning Tool (DMPPT) to model the impact and cost of scaling up adult VMMC in Botswana, Lesotho, Malawi, Mozambique, Namibia, Rwanda, South Africa, Swaziland, Tanzania, Uganda, Zambia, Zimbabwe, and Nyanza Province in Kenya. We use epidemiologic and demographic data from recent household surveys for each country. The cost of VMMC ranges from US$65.85 to US$95.15 per VMMC performed, based on a cost assessment of VMMC services aligned with the World Health Organization’s considerations of models for optimizing volume and efficiencies. Results from the DMPPT models suggest that scaling up adult VMMC to reach 80% coverage in the 13 countries by 2015 would entail performing 20.34 million circumcisions between 2011 and 2015 and an additional 8.42 million between 2016 and 2025 (to maintain the 80% coverage). Such a scale-up would result in averting 3.36 million new HIV infections through 2025. In addition, while the model shows that this scale-up would cost a total of US$2 billion between 2011 and 2025, it would result in net savings (due to averted treatment and care costs) amounting to US$16.51 billion.
Conclusions: This study suggests that rapid scale-up of VMMC in eastern and southern Africa is warranted based on the likely impact on the region’s HIV epidemics and net savings. Scaling up of safe VMMC in eastern and southern Africa will lead to a substantial reduction in HIV infections in the countries and lower health system costs through averted HIV care costs. Please see later in the article for the Editors’ Summary.
Citation: Njeuhmeli E, Forsythe S, Reed J, Opuni M, Bollinger L, et al. (2011) Voluntary Medical Male Circumcision: Modeling the Impact and Cost of Expanding Male Circumcision for HIV Prevention in Eastern and Southern Africa. PLoS Med 8(11): e1001132. doi:10.1371/journal.pmed.1001132 Academic Editor: Stephanie L. Sansom, Centers for Disease Control and Prevention, United States of America Received May 26, 2011; Accepted October 19, 2011; Published November 29, 2011 This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication. Funding: This work was funded by PEPFAR (through the USAID Health Policy Initiative) and UNAIDS (through the Technical Support Facility (TSF)). Technical staff from USAID (EN, DC), CDC (JR), WHO (TF), and UNAIDS (CH) had a role in study design, data collection and analysis, decision to publish, and preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. Abbreviations: ART, antiretroviral therapy; CDC, Centers for Disease Control and Prevention; DMPPT, Decision Makers’ Program Planning Tool; MOVE, Models for Optimizing Volume and Efficiencies; PEPFAR, United States President’s Emergency Plan for AIDS Relief; STI, sexually transmitted infection; UNAIDS, Joint United Nations Programme on HIV/AIDS; USAID, United States Agency for International Development; VMMC, voluntary medical male circumcision; WHO, World Health Organization * E-mail: [email protected] . These authors contributed equally to this work.
15 Impact and Cost of VMMC for HIV Prevention
Introduction modeling tool, which has been reviewed by an expert panel [10], allows analysts and decision makers to estimate the epidemiologic Three randomized controlled trials have shown that voluntary impact and cost of alternative programmatic options for scaling up medical male circumcision (VMMC) reduces heterosexual HIV VMMC. acquisition in men by up to 60% [1–3]. On the basis of these trial The objective of this study is to estimate the country-specific results, the World Health Organization (WHO) and the Joint epidemiologic impact and the cost and net savings associated with United Nations Programme on HIV/AIDS (UNAIDS) now scaling up VMMC services based on MOVE considerations in the recommend that VMMC be offered to heterosexual men in 13 priority countries in eastern and southern Africa. The combination with other effective HIV risk reduction interventions perspective that is taken is that of governments and their in settings with generalized HIV epidemics and where a international partners in their roles as health program funders. substantial proportion of men are not circumcised [4]. To do this, we run country-specific DMPPT models and explore The long-term population-level impact of implementing and how the results for each country vary by VMMC effectiveness, scaling up VMMC services is expected to be considerable in terms VMMC coverage level, time to scale-up, level of post-circumcision of HIV infections averted, as well as net savings associated with the behavior change, VMMC unit cost, and antiretroviral therapy reduced need for treatment, care, and support of infected (ART) cost. individuals. Previous model-based studies estimated that VMMC This study expands on previous related work [8,9,11] in several scale-up in countries with generalized HIV epidemics could result ways. All of our analyses are country-specific and based on the in substantial reductions in HIV transmission and prevalence over most recent data on HIV and male circumcision prevalence time among both men and women [3,5–9]. A recent review of available. Our calculations provide more comprehensive estimates these studies concluded that one HIV infection could be averted of VMMC-related costs than was the case in previous studies since for every five to 15 VMMCs performed [10]. Modeling studies they include costs for supply chain management and waste also have found that expansion of VMMC services produces net management, as per WHO’s recent recommendations for efficient savings when compared to lifetime HIV treatment costs [8,9,11]. VMMC scale-up [24]. In addition, we provide country-specific In addition, studies have shown VMMC to be protective against results on impact (VMMC per HIV infection averted), cost- some other sexually transmitted infections (STIs) in both men and effectiveness (cost per HIV infection averted) and cost savings (cost women. VMMC has been found to reduce the risk of herpes of VMMC relative to the averted cost of lifetime provision of simplex virus-2 in men [12] and human papillomavirus in men ART). [12–15] and their female partners [16], and is associated with a reduction in the risk of genital ulcer disease [3,17] and genital Methods cancers [18–20] in both men and women. In light of this evidence, in early 2007, WHO and UNAIDS The Model identified the following priority countries for VMMC scale-up: The DMPPT is a Microsoft-Excel-based modeling tool that Botswana, Lesotho, Malawi, Mozambique, Namibia, Rwanda, estimates the epidemiologic impact (HIV infections averted) and South Africa, Swaziland, Tanzania, Uganda, Zambia, Zimbabwe, the cost and net savings associated with different programming and Nyanza Province in Kenya (Figure 1). Four years later, most scenarios for VMMC scale-up. A list of all the equations used in of these 13 countries have developed plans for VMMC scale-up the model can be found in Table S1. Additional details, including and are at various stages of program implementation. However, a copy of the workbook and the model manual [25], can be found with a few exceptions (e.g., Nyanza Province in Kenya and the at http://www.malecircumcision.org/programs/DMPPT.html. Iringa Region in Tanzania), progress in expanding VMMC The DMPPT is a compartmental deterministic model in which services remains slow [21]. the population is disaggregated into four sex/age groups (females There is consensus that VMMC scale-up will require substantial ages 15–24 y, males ages 15–24 y, females ages 25–49 y, and funding and massive efforts to adequately train personnel, equip males ages 25–49 y). For each population group, a susceptible and facilities, and ensure the regular distribution of necessary infected population is described [26,27]. The susceptible popula- commodities—mostly in settings with weak health systems and tion is increased by people aging into it (15-y-olds entering the resource constraints. As background for the scale-up of safe younger age group and 25-y-olds leaving the younger age group VMMC, WHO, UNAIDS, and collaborating partners have and entering the older age group) and decreased by non-AIDS developed a number of guidelines and toolkits, including a deaths and new infections. The infected population is increased by recommended minimum package of VMMC services [22] and new infections and decreased by non-AIDS deaths, AIDS deaths, clinical guidelines for the provision of these services [23]. In and aging out. addition, to facilitate a more rapid scale-up of safe VMMC, WHO In the model, HIV incidence (the proportion of the susceptible has outlined considerations for models to optimize the volume and population becoming infected each year) is the product of HIV efficiency (MOVE) of VMMC services [24]. Central to these prevalence in the population and the force of infection. The force considerations is the efficient use of facility space through the of infection is determined by the base rate of infection (which is a dedication of multiple surgical beds to one surgical team and the fitting parameter specific to each population group), changes in coordination of client flow; the efficient use of staff time through behavior, and changes due to male circumcision. Changes in task shifting and task sharing, including deployment of non- behavior are assumed to occur as the epidemic progresses because physicians to complete all or specific steps in VMMC surgery; and of two key influences. Individuals with the riskiest behaviors are the bundling of commodities and supplies required to perform assumed to become infected first and die sooner than the rest of VMMC, including consumable materials and surgical instruments the population. Also, as AIDS deaths accumulate, a powerful effect [24]. on individual behavior is assumed, as those who know someone To support decision making and planning for VMMC scale-up, who has died from AIDS are motivated to adopt safer behaviors. the United States Agency for International Development (USAID) Thus, the force of infection can drop over time as the cumulative Health Policy Initiative collaborated with UNAIDS to develop the number of AIDS deaths increases. The amount of the effect is Decision Makers’ Program Planning Tool (DMPPT) [25]. This determined by a fitting parameter that produces the best fit to the
16 Impact and Cost of VMMC for HIV Prevention
Figure 1. Geographic distribution of HIV and male circumcision prevalence. For the 13 countries included in the study, the map on the left indicates the prevalence of HIV in adults ages 15 to 49 y. The map on the right indicates the percentage of men who are circumcised. doi:10.1371/journal.pmed.1001132.g001 historical HIV prevalence trend. The impact of other prevention change being modeled as having a 0% post-circumcision behavior interventions is incorporated through exogenous reductions in the change effect. force of infection. The effect of male circumcision on the force of The DMPPT model calculates the yearly cost of the additional infection is simply the increase in the prevalence of male circumcisions (the number of VMMCs required above and beyond circumcision multiplied by the reduction in susceptibility due to those occurring at pre-scale-up coverage levels) as the sum of the circumcision. additional circumcisions performed per annum times the VMMC For each scenario of interest, the DMPPT model generates two unit cost. The model also calculates the net savings associated with projections: a baseline projection, in which male circumcision HIV infections averted and the subsequent treatment costs coverage is held constant at the pre-scale-up level, and a scale-up averted. projection where VMMC coverage is scaled up to the desired coverage level. In the baseline projection, new infections increase Demographic, Epidemiologic, and Sexual Behavior Data over time in response to the initial force of infection. This causes The demographic and epidemiologic inputs required to HIV prevalence to rise rapidly in early years. The force of estimate the epidemiologic impact of VMMC scale-up include infection then declines somewhat as AIDS deaths accumulate and the following: prevalence stabilizes. Implementation of a male circumcision program can cause a further reduction in the force of infection N Demographic data—size of adult population by sex and age (depending on the increase in male circumcision coverage), leading group, adult population growth rate, crude birth rate, crude to a decline in new infections and HIV prevalence. There is some death rate, proportion of population surviving to age 15 y, pre- concern that men who undergo circumcision may develop a false scale-up male circumcision prevalence. sense of protection against HIV and thereafter decrease or even N Epidemiologic data—HIV prevalence among adults by sex stop previously implemented protective behaviors [28,29]. Post- and age group; effectiveness of male circumcision in circumcision behavior change is modeled in the DMPPT as a preventing HIV infection; underlying HIV transmission factors proportional change in the force of infection, with a VMMC scale- based on scientific evidence, including probability of transmis- up scenario in which circumcision does not lead to any behavior sion from mother to child; fertility reduction due to HIV.
17 Impact and Cost of VMMC for HIV Prevention
Table 1. Pre-scale-up population characteristics, by country.
Adult Population Male Circumcision Population Aged 15– Growth Rate Prevalence Adult HIV Prevalence Adult HIV Incidence Country 49 y (Millions) (Percentage) (Percentage) (Percentage) (Percentage)
Botswana 1.08 2.42 10.2 22.9 2.0 Lesotho 1.06 1.34 0.0 23.6 2.9 Malawi 6.84 3.02 20.7 12.5 1.0 Mozambique 10.32 2.45 59.5 12.9 1.4 Namibia 1.09 2.22 21.0 14.1 0.8 Nyanza Province, Kenya 1.51 2.80 44.8 13.1 1.6 Rwanda 4.49 3.29 10.0 3.1 0.3 South Africa 26.84 0.89 44.7 16.9 2.1 Swaziland 0.59 1.51 8.2 26.4 3.2 Tanzania 19.63 2.89 66.8 6.7 0.8 Uganda 14.16 3.87 24.8 6.5 0.7 Zambia 6.85 2.61 10.8 14.9 1.6 Zimbabwe 6.19 1.50 10.3 17.9 2.2
Table 1 provides the size of the 2008 population aged 15–49 y, adult population growth rate, and adult HIV incidence from Spectrum, which is populated with UN population data [27,43,44]. The years for the pre-scale-up male circumcision and adult HIV prevalence data are as follows: Botswana (2004) [30], Lesotho (2004) [9,39], Malawi (2004) [40], Mozambique (2003) [35], Namibia (2007) [38], Nyanza Province, Kenya (2008/2009) [36], Rwanda (2005) [9,34], South Africa (2008), Swaziland (2006/ 2007) [32], Tanzania (2007/2008) [42], Uganda (2004) [37], Zambia (2007) [31], and Zimbabwe (2005/2006) [33]. doi:10.1371/journal.pmed.1001132.t001
N Sexual behavior data—sexual mixing patterns by sex and age costs on a costing analysis conducted in Swaziland, a detailed group, post-circumcision behavior change (change in condom description of which is provided elsewhere [52]. The VMMC unit use and/or change in number of sexual partners following cost used for each country is shown in Table 3. VMMC resulting from lowered perceived risk). For lifetime HIV treatment costs, we use the default discounted value in the DMPPT model (US$7,400) [25]; we use the same We obtained the required demographic, epidemiologic, and value for all countries. This value includes costs of AIDS treatment sexual behavior data for each country from recent household and care, including ART, treatment of major opportunistic surveys [30–42] and Spectrum, a modeling software package, which infections, laboratory tests, and home-based care. The lifetime includes a demographic platform populated with UN population HIV treatment cost is based on a unit cost of US$155 for first-line data [27,43,44]. Country-specific pre-scale-up population charac- antiretroviral drugs and US$1,678 for second-line antiretroviral teristics, including size of the adult population, adult population drugs [53]. We also assume that need for treatment begins after growth rate, male circumcision and adult HIV prevalence, and HIV 8 y of infection and that the annual continuation rate on ART is incidence, are shown in Table 1. A complete listing of the data used 97.5% [27]. for each country is available from the authors upon request.
Cost Data A breakdown of the VMMC unit cost used in the model is Table 2. VMMC unit cost components. provided in Table 2. With the exception of the waste management and supply chain costs, the unit cost components are derived from a VMMC costing study conducted in Zimbabwe in 2010 as part of Category Cost Component Cost (US$) a multi-country study that also included Kenya, Namibia, South Direct costs Consumables 28.67 Africa, Uganda, and Zambia [45–50]. Zimbabwe was chosen Waste management 9.39 because it is one of the first countries to scale up VMMC services Other supply chain 10.91 following WHO’s MOVE considerations for more efficient use of Staff costs Variablea facility space and staff time and for the bundling of the commodities required to perform VMMC [24]. Information on Training costs 1.70 required inputs and costs was collected from sites providing Indirect costs Capital costs 0.45 VMMC in Zimbabwe. This includes data on both direct costs Maintenance and utility costs 3.24 (consumables, staff costs, and training costs) and indirect costs Support overhead 4.42 (capital costs, maintenance and utility costs, support overhead, and Management overhead 2.32 management overhead). Because significant variations in labor costs were noted across countries, we vary the staff costs by Table 2 presents the components of the VMMC unit cost used in this study. The country, adjusting them by after tax median monthly disposable VMMC unit cost was derived based on a VMMC costing study conducted in salary. We also adjust the costing information collected in the Zimbabwe in 2010 [50] and supplemented with analysis of data from Swaziland. aWe vary staff costs by country, adjusting them by after tax median monthly Zimbabwe study for underreporting on waste management and disposable salary. other supply chain costs [51]. We base the waste management doi:10.1371/journal.pmed.1001132.t002
18 Impact and Cost of VMMC for HIV Prevention
Table 3. VMMC unit cost by country.
Country VMMC Unit Cost—Base Case (US$) VMMC Unit Cost Range (US$)
Botswana 78.07 62.45–93.68 Lesotho 83.78 67.03–100.54 Malawi 83.78 67.03–100.54 Mozambique 86.29 69.03–103.55 Namibia 86.60 69.28–103.92 Nyanza Province, Kenya 74.89 59.92–89.87 Rwanda 80.13 64.10–96.16 South Africa 95.15 76.12–114.18 Swaziland 74.83 59.86–89.79 Tanzania 82.56 66.04–99.07 Uganda 65.85 52.68–79.02 Zambia 89.65 71.72–107.58 Zimbabwe 78.23 62.58–93.87
Table 3 shows the country-specific VMMC unit costs used in the study. To obtain the VMMC unit cost for each country, we vary staff costs in each country, adjusting them by after tax median monthly disposable salary [51]. The range used in the sensitivity analysis is country base case unit cost 620%. doi:10.1371/journal.pmed.1001132.t003
Following common practice in economic evaluation, we apply N Post-circumcision behavior change—a 30% post-circumcision an annual discount rate of 3% on future expenditures and savings behavior change effect is associated with MMC scale-up [54] as well as on infections averted [55]. (versus no post-circumcision behavior change, or 0%). N VMMC unit cost—country base case unit cost 620% as Base Case and Sensitivity Analysis shown in Table 3. Our analyses are limited to the scale-up of VMMC services N Lifetime ART cost—ART cost is assumed to decline over among males ages 15 to 49 y who are HIV-negative. For all time, leading to a lifetime cost of US$3,400 (versus US$7,400) models, we assume that the scale-up begins in 2011 and that the [53]. pace of scale-up is slower at first, followed by a more rapid scale- up, and then another slow period of scale-up, to allow for training and other logistic developments likely to occur in the early stages Results of implementation. The model observation period for all models is 2011 to 2025. This allows us to evaluate the 15-y impact of Base Case—80% VMMC Coverage by 2015 VMMC more comprehensively. The number of additional VMMCs required to achieve 80% For the base case, we assume the following: male circumcision coverage by 2015 is presented in Table 4. A total of 20.34 million VMMCs are required to scale up male N VMMC effectiveness—we use 60% for the protective effect of circumcision coverage to 80% by 2015 in Botswana, Lesotho, VMMC and assume that the effect is constant over the model Malawi, Mozambique, Namibia, Rwanda, South Africa, Swazi- simulation period [1,56]. land, Tanzania, Uganda, Zambia, Zimbabwe, and Nyanza N Target VMMC coverage level—male circumcision coverage is Province in Kenya. When taking the longer-term view (to 2025), increased from pre-scale-up levels to 80% of HIV-negative an additional 8.42 million VMMCs are required between 2016 males ages 15 to 49 y. and 2025, for a total of almost 29 million VMMCs in the model’s N Time to 80% scale-up—male circumcision coverage is full study period (2011–2025). The number of VMMCs needed to increased from pre-scale-up levels to 80% of HIV-negative achieve 80% male circumcision by 2015 varies by country, with males ages 15 to 49 in 5 y, with the coverage target reached by the largest number of VMMCs required in South Africa (the 2015 and maintained at 80% thereafter. country with the largest population and a high adult HIV prevalence) followed by Uganda (the country with a large and fast- Post-circumcision behavior change—no post-circumcision N growing population and low baseline VMMC prevalence), behavior change is associated with VMMC scale-up. Malawi, Zambia, and Zimbabwe (also countries low baseline N VMMC unit cost—the VMMC unit cost used for each country VMMC prevalence). is shown in Table 3. We find a strong impact of scaling up VMMC to achieve 80% Given uncertainty in model inputs, we also conduct sensitivity coverage by 2015 on the number of adult HIV infections averted. analyses, exploring various alternative scenarios. We vary VMMC A total of 430,000 HIV infections are averted in the 13 countries effectiveness, VMMC coverage level, time to scale-up, level of between 2011 and 2015, while almost 3.36 million HIV infections post-circumcision behavior change, VMMC unit cost, and lifetime are averted by 2025 (Table 4). South Africa is the country with the HIV treatment cost as follows: largest number of HIV infections averted (with more than 1 million infections averted between 2011 and 2025) followed by Zimbabwe, Zambia, and Uganda. More than 20% of new HIV VMMC effectiveness—34% and 77% (versus 60%) [1]. N infections are averted between 2011 and 2025 in all countries N Target VMMC coverage levels—50% and 100% (versus. 80%). except Mozambique, Nyanza Province (Kenya), South Africa, and N Time to 80% scale-up—1, 10, and 15 y (versus 5 y). Tanzania (the countries with the highest baseline VMMC
19 Impact and Cost of VMMC for HIV Prevention
Table 4. Impact of VMMC on HIV infections averted in base case, by country, 2011–2015 and 2011–2025.
Additional VMMCs HIV Infections HIV Infections VMMC per HIV Country Time Period (Millions) Averted (Millions) Averted (Percentage) Infection Averted
Botswana 2011–2015 0.35 0.01 12 36 2011–2025 0.49 0.06 28 8 Lesotho 2011–2015 0.38 0.01 12 28 2011–2025 0.54 0.11 37 5 Malawi 2011–2015 2.10 0.03 11 66 2011–2025 3.04 0.24 28 13 Mozambique 2011–2015 1.06 0.03 5 38 2011–2025 1.53 0.22 13 7 Namibia 2011–2015 0.33 0.00 9 109 2011–2025 0.48 0.02 25 26 Nyanza Province, Kenya 2011–2015 0.38 0.01 6 36 2011–2025 0.57 0.07 16 8 Rwanda 2011–2015 1.75 0.01 10 239 2011–2025 2.53 0.06 29 44 South Africa 2011–2015 4.33 0.14 7 30 2011–2025 5.94 1.08 20 5 Swaziland 2011–2015 0.18 0.01 12 25 2011–2025 0.27 0.06 34 5 Tanzania 2011–2015 1.38 0.02 3 56 2011–2025 1.98 0.20 9 10 Uganda 2011–2015 4.25 0.05 10 91 2011–2025 6.35 0.34 25 19 Zambia 2011–2015 1.95 0.04 12 43 2011–2025 2.87 0.34 30 8 Zimbabwe 2011–2015 1.91 0.06 13 31 2011–2025 2.17 0.57 42 4 Total 2011–2015 20.34 0.43 8 47 2011–2025 28.76 3.36 22 9
Table 4 presents the additional VMMCs and the impact of VMMC on HIV infections averted for 2011–2015 and 2011–2025 for the base case scenario. HIV infections averted are not discounted. doi:10.1371/journal.pmed.1001132.t004 prevalence). Zimbabwe is the country with the highest percentage new infections in women is also reduced. This occurs by reducing of new HIV infections averted, with 42% of new infections averted the exposure of women to HIV-infected men. That is, as HIV between 2011 and 2025. The large benefit of VMMC in incidence decreases in men following VMMC scale-up, the Zimbabwe is largely driven by the currently low prevalence of probability of women encountering infected male partners VMMC (10.3%) and high HIV prevalence (17.9%) and incidence decreases, with a consequent reduction in HIV incidence among (2.2%). women. The HIV infections averted (presented in Table 4) thus The number of VMMCs required to avert one HIV infection is represent infections averted in both men and women. Figure 2 calculated by dividing the additional number of VMMCs required illustrates the male and female HIV infections averted over time, by the number of HIV infections averted over the relevant time by country, for 2011–2025. In all countries, the cumulative period. For the period 2011–2015, the number of VMMCs per number of male HIV infections averted between 2011 and 2025 is HIV infection averted ranges from 25 in Swaziland to 239 in higher than the cumulative number of female HIV infections Rwanda, while for the full study period (2011–2025), the number averted. In the early years, the HIV infections averted occur of VMMCs per HIV infection averted ranges from four in mostly among men, but over time, the proportion of HIV Zimbabwe to 44 in Rwanda (Table 4). For the period 2011–2025, infections averted in women steadily increases, with new HIV the number of VMMCs per HIV infection averted is ten or less in infections averted in women representing almost half of the total all countries except Malawi, Namibia, Rwanda, and Uganda, each HIV infections averted in 2025. of which has a relatively low incidence of HIV. Figure 3 shows the discounted cost of scaling up VMMC to Although circumcision of HIV-infected men has not been found achieve 80% coverage by 2015 for each of the 13 priority to directly reduce HIV transmission to their female partners [57], countries for the period 2011–2015. The cost ranges from and the primary impact of increasing VMMC coverage is to US$12.53 million in Swaziland (the country with the smallest reduce the number of new HIV infections in men, the number of number of additional VMMCs required) to US$376.55 million in
20 Impact and Cost of VMMC for HIV Prevention
Figure 2. HIV infections averted in base case, by male and female, by country, 2011–2025. Figure 2 provides data for males and females for 2011–2025 for the base case scenario in which 80% coverage of VMMC is reached in 2015 and maintained thereafter, and there is no post- circumcision risk behavior change. doi:10.1371/journal.pmed.1001132.g002
South Africa (the country with the highest number of additional HIV infection averted. For 2011–2025, this value ranges from VMMCs required), with a total of US$1,520,000,000 needed to US$3,304 in Rwanda to US$7,031 in Zimbabwe. The overall net scale up VMMC coverage to 80% in the 13 countries by 2015. savings per HIV infection averted for 2011–2025 for all 13 The cost for the period 2011–2025 is shown in the first column of countries is US$6,608. Table 5. A total of over US$2,000,000,000 is needed to scale up VMMC coverage to 80% by 2015 and maintain this level of Sensitivity Analysis coverage in the 13 countries until 2025. Table 6 presents the results of the sensitivity analysis. The table Combining the cost data with the number of HIV infections provides the values for the HIV infections averted, the number of averted, we obtain the discounted cost per HIV infection averted. VMMCs per HIV infection averted, the cost per HIV infection The cost per HIV infection averted for the period 2011–2025 averted, and the net savings per HIV infection averted for 2011– ranges from US$369 in Zimbabwe, where adult HIV prevalence is 2025 for each parameter examined. The values for the base 17.9%, to US$4,096 in Rwanda, where adult HIV prevalence is scenario (80% male circumcision coverage within 5 y) are shown lower than in any of the other 13 countries studied (3.1%). The in the third column of Table 6. overall cost per HIV infection averted for 2011–2025 for all 13 The results help to confirm the internal consistency of the countries is US$809. DMPPT model and show that, in large measure, the VMMC Finally, we calculate the net savings associated with VMMC impact, cost, and savings results that we observe in the base case scale-up. The savings due to future ART costs avoided (with the are robust to changes in VMMC effectiveness, VMMC coverage, discounted value of lifetime HIV treatment costs estimated at time to scale-up, level of post-circumcision behavior change, US$7,400 per infection) minus the discounted VMMC costs VMMC unit cost, and lifetime HIV treatment cost. amount to US$16,510,000,000 for the 13 countries from 2011 to Our analysis varying VMMC effectiveness shows that even if we 2025 (Table 5). The net savings can also be combined with the assume much reduced effectiveness (34% instead of 60%), high number of HIV infections averted to obtain the net savings per numbers of infections are averted and net savings per infection
21 Impact and Cost of VMMC for HIV Prevention
Figure 3. Cost for scaling up VMMC coverage in base case, by country, 2011–2015. Figure 3 provides the discounted cost of scaling up VMMC coverage for 2011–2015 for the base case scenario in which 80% coverage of VMMC is reached in 2015 and maintained thereafter, and there is no post-circumcision risk behavior change. doi:10.1371/journal.pmed.1001132.g003 averted are obtained. The same is true if the VMMC coverage infection averted and an increase (decrease) in the net savings per target is reduced from 80% to 50% or if the time to achieve 80% HIV infection averted in countries. Assuming the lower VMMC VMMC coverage is increased to 10 or 15 y. unit cost for each country, the cost and net savings associated To assess the potential impact of post-circumcision risk behavior with scaling up VMMC coverage to 80% by 2015 and change, the model calculates the impact of risky sexual behaviors maintaining this level of coverage in the 13 countries until 2025 reverting to patterns that existed earlier in the epidemic, prior to the are US$1,620,000,000 and US$16,900,000,000 instead of scale-up of VMMC services. It should be noted that this impact is US$2,020,000,000 and US$16,510,000,000 in the base case. different from the early resumption of sexual activity before complete Assuming the higher VMMC unit cost, the cost of scaling up wound healing that can also be an issue of concern in the context of VMMC services in the 13 countries is US$2,430,000,000, expanding VMMC services. The impact is the result of behavior resulting in net savings of US$16,090,000,000. change due to a false perception that VMMC has eliminated the risk Reducing lifetime ART cost from US$7,400 to US$3,400 does of HIV transmission. As compared to no post-circumcision risk significantly reduce the net savings per infection averted in all behavior change, assuming 30% post-circumcision risk behavior countries, although it remains well above US$2,000 in almost all change results in a decrease in the HIV infections averted, an increase countries. Rwanda is the only country where such a reduction in in the numbers of VMMCs per HIV infection averted, an increase in lifetime ART cost would lead to net losses per infection averted. the cost per HIV infection averted, and a decrease in net savings. In Net savings associated with scaling up VMMC coverage is reduced all of the countries studied except Rwanda, VMMC scale-up remains to US$6,480,000,000 in the 13 countries if lifetime ART cost is cost saving, even with 30% post-circumcision risk behavior change. US$3,400. In Rwanda, the country with the lowest HIV prevalence, scaling up VMMC in the presence of this level of post-circumcision risk Discussion behavior leads to an increase in the number of new HIV infections. Predictably, decreasing (increasing) the base case VMMC unit To support decision making and planning for VMMC scale-up cost by 20% results in a decrease (increase) in the cost per HIV in Botswana, Lesotho, Malawi, Mozambique, Namibia, Rwanda,
22 Impact and Cost of VMMC for HIV Prevention
Table 5. Total cost and net savings per HIV infection averted for base case scenario, by country, 2011–2025.
Net Savings per HIV Cost of VMMC Scale-Up, Cost per HIV Infection Net Savings, 2011–2025 Infection Averted, 2011– Country 2011–2025 (US$ Millions) Averted, 2011–2025 (US$) (US$ Millions) 2025 (US$)
Botswana 32.79 693 317.14 6,707 Lesotho 34.97 442 550.57 6,958 Malawi 218.46 1,216 1,110.98 6,184 Mozambique 113.32 704 1,077.25 6,696 Namibia 35.52 2,558 67.21 4,842 Nyanza Province, Kenya 36.38 660 371.23 6,740 Rwanda 173.78 4,096 140.21 3,304 South Africa 489.47 605 5,498.55 6,795 Swaziland 17.18 406 295.88 6,994 Tanzania 140.40 950 953.74 6,450 Uganda 357.29 1,408 1,520.36 5,992 Zambia 220.42 869 1,679.41 6,623 Zimbabwe 153.84 369 2,930.64 7,031 Total 2,023.80 809 16,513.15 6,608
Table 5 provides the discounted cost of scaling up VMMC coverage, cost per HIV infection averted, and net savings per HIV infection averted for 2011–2025 for the base case scenario. HIV infections averted are discounted. doi:10.1371/journal.pmed.1001132.t005
South Africa, Swaziland, Tanzania, Uganda, Zambia, Zimbabwe, yet on the feasibility of scaling up VMMC coverage to 80%, there and Nyanza Province in Kenya, we estimate the country-specific is evidence that VMMC coverage rates of above 50% can be impact and cost of scaling up VMMC services using the DMPPT reached through VMMC scale-up efforts [56]. model developed by the USAID and UNAIDS. Our study suggests Similar to the results reported by other authors, our results that rapid scale-up of VMMC in eastern and southern Africa is suggest that post-circumcision risk behavior change is unlikely to warranted based on the likely impact on the region’s HIV completely reverse the benefits of male circumcision, with the epidemics and the resultant cost savings. possible exception of Rwanda. In the case of Rwanda, the low Model results also show that, although the primary impact of incidence of HIV (0.3%) means that risk compensation associated scaling up VMMC coverage is to reduce the number of new HIV with male circumcision could actually make the epidemic worse. infections in men, the intervention also prevents HIV infections in In the remaining countries, the negative consequences from a 30% women. While the cumulative number of male HIV infections risk compensation would be insufficient to fully reverse the benefits averted between 2011 and 2025 is higher than the cumulative of VMMC. In all countries, it is critical that VMMC programs number of female HIV infections averted, the proportion of HIV emphasize that male circumcision is not 100% protective and that infections averted in women steadily increases until HIV infections condom use and other behavioral risk reductions remain essential averted in women represent almost half of the new HIV infections [8,59]. Effective messaging in this regard is critical to ensure that averted in 2025. men and their sexual partners do not increase their sexual risk Our results show that the costs per infection averted are behaviors following VMMC scale-up. comparable to the estimated cost per infection averted of a This study has a number of limitations. Because this model was number of key HIV prevention interventions implemented in the designed primarily for advocacy purposes, certain elements are not region. The cost per HIV infection averted, which ranges from modeled, including treatment, a more fully articulated age US$369 in Zimbabwe to US$4,096 in Rwanda, is below US$1,000 structure, and force of infection, which probably leads to an in all countries except Malawi, Namibia, Rwanda, and Uganda. overestimate of the impact of male circumcision. These costs per infection averted are comparable to those of Another limitation is that we use pre-scale-up male circumcision prevention of vertical transmission (US$663 per HIV-positive coverage estimates based on men’s self-reported circumcision birth averted), voluntary counseling and testing (US$1,315 per status in household surveys. There is some evidence that HIV infection averted), and prevention of STIs (US$321– uncircumcised and partially circumcised men may report being US$1,665) [58]. circumcised in some populations [60]. Unfortunately, there is no Our results are somewhat sensitive to changes in VMMC information available on the prevalence of this phenomenon in the effectiveness, VMMC coverage targets, time to scale-up, and countries studied. In the one country where partial circumcision is VMMC cost. Nonetheless, all of the values tested for these known to be widely practiced (Lesotho), we assume baseline parameters resulted in lives saved and cost savings for all countries. VMMC prevalence to be at 0% rather than using the self-reported It is especially important to note that even if VMMC coverage is data. scaled up to 50% in 5 y instead of to 80%, high numbers of Additional limitations relate to our cost assumptions. Ideally, we infections will be averted and net savings per infection averted will would have obtained country-specific VMMC unit cost estimates be obtained in all countries except Mozambique and Tanzania— based on the MOVE model instead of only adjusting the staff costs the two countries where pre-scale-up coverage is higher than 50%. by country based on after-tax median monthly disposable salary. This is critical because even though there is no evidence available However, this model for delivering VMMC was available in only a
23 Table 6. Sensitivity analysis of key parameters (2011–2025).
Base Case ART Country Results from DMPPT Scenario VMMC Effectiveness VMMC Coverage Time to Scale-Up BC 30%a Cost per VMMC Cost = US$3,400 220% Base +20% Base 34% 77% 50% 100% 1 y 10 y 15 y Case Cost Case Cost
Botswana IA 62,773 34,504 82,029 35,620 80,877 75,952 45,226 30,171 62,773 62,773 62,773 62,773 VMMC per IA 8 14 6 8 8 7 10 14 8 8 8 8 Cost per IA (US$) 693 1,253 533 694 694 605 881 1,197 836 555 832 693 Net savings per IA (US$) 6,707 6,147 6,867 6,706 6,706 6,795 6,519 6,203 6,564 6,845 6,568 2,707 Lesotho IA 106,427 60,123 132,245 66,385 129,578 130,568 73,664 46,994 100,848 106,427 106,427 106,427 VMMC per IA 5 9 4 5 5 4 7 10 5 5 5 5 Cost per IA (US$) 442 777 356 440 454 382 579 816 466 354 530 442 Net savings per IA (US$) 6,958 6,623 7,044 6,960 6,946 7,018 6,821 6,584 6,934 7,046 6,870 2,958 Malawi IA 240,685 131,912 315,299 121,012 317,639 292,121 172,169 113,722 172,576 240,685 240,685 240,685 VMMC per IA 13 23 10 12 13 11 16 23 18 13 13 13 Cost per IA (US$) 1,216 2,210 930 1,194 1,233 1,064 1,543 2,100 1,696 973 1,459 1,216 Net savings per IA (US$) 6,184 5,190 6,470 6,206 6,167 6,336 5,857 5,300 5,704 6,427 5,941 2,184 Mozambiqueb IA 215,861 103,714 314,141 421,592 262,897 153,291 100,400 159,030 215,861 215,861 215,861 24 VMMC per IA 7 14 5 7 6 9 13 10 7 7 7 Cost per IA (US$) 704 1,440 491 714 615 900 1,235 954 563 845 704 Net savings per IA (US$) 6,696 5,960 6,909 6,686 6,785 6,500 6,165 6,446 6,837 6,555 2,696 Namibia IA 18,373 9,936 24,335 9,243 24,174 23,160 12,520 7,924 12,947 18,373 18,373 18,373 VMMC per IA 26 48 20 25 26 21 36 53 37 26 26 26 Cost per IA (US$) 2,558 4,741 1,929 2,502 2,601 2,134 3,453 4,955 3,629 2,047 3,070 2,558 Net savings per IA (US$) 4,842 2,659 5,471 4,898 4,799 5,266 3,947 2,445 3,771 5,353 4,330 842 Nyanza IA 73,420 35,776 105,322 2,382 72,961 53,834 32,309 21,649 31,567 44,636 44,636 73,420 Province, Prevention HIV for VMMC of Cost and Impact Kenya VMMC per IA 8 16 5 8 8 7 10 13 11 8 8 8 Cost per IA (US$) 660 1,337 465 661 661 585 820 1,089 933 528 793 660 Net savings per IA (US$) 6,740 6,063 6,935 6,739 6,739 6,815 6,580 6,311 6,467 6,872 6,607 2,740 Rwandac IA 56,840 32,138 72,988 33,163 72,039 68,999 40,514 26,480 (1,183) 56,840 56,840 56,840 VMMC per IA 44 79 35 43 45 38 58 81 (2,132) 44 44 44 Cost per IA (US$) 4,096 7,248 3,188 4,015 4,152 3,594 5,195 7,118 (192,148) 3,276 4,915 4,096 Net savings per IA (US$) 3,304 152 4,212 3,385 3,248 3,806 2,205 282 N.A. 4,124 2,485 (696) South Africa IA 1,083,869 553,267 1,496,933 164,840 1,677,803 1,340,315 750,846 478,788 954,426 1,083,869 1,083,869 1,083,869 VMMC per IA 5 11 4 5 6 5 7 11 6 5 5 5 Cost per IA (US$) 605 1,161 444 595 614 518 798 1,135 685 484 726 605 Net savings per IA (US$) 6,795 6,239 6,956 6,805 6,786 6,882 6,602 6,265 6,715 6,916 6,674 2,795 Table 6. Cont.
Base Case ART Country Results from DMPPT Scenario VMMC Effectiveness VMMC Coverage Time to Scale-Up BC 30%a Cost per VMMC Cost = US$3,400 220% Base +20% Base 34% 77% 50% 100% 1 y 10 y 15 y Case Cost Case Cost
Swaziland IA 56,810 31,886 71,520 33,356 70,597 69,466 39,535 25,032 53,234 56,810 56,810 56,810 VMMC per IA 5 8 4 5 5 4 6 9 5 5 5 5 Cost per IA (US$) 406 717 323 401 418 366 534 755 436 325 487 406 Net savings per IA (US$) 6,994 6,683 7,077 6,999 6,982 7,034 6,866 6,645 6,964 7,075 6,913 2,994 Tanzaniab IA 199,249 92,261 300,691 492,563 241,090 143,379 95,386 173,793 199,249 199,249 199,249 VMMC per IA 10 21 7 10 9 13 18 11 10 10 10 Cost per IA (US$) 950 2,035 633 967 838 1,190 1,600 1,090 760 1,139 950 Net savings per IA (US$) 6,450 5,365 6,767 6,433 6,562 6,210 5,800 6,310 6,640 6,261 2,450 Uganda IA 339,524 182,516 451,568 157,211 458,128 408,009 245,533 1,654,578 104,102 339,524 339,524 339,524 25 VMMC per IA 19 35 14 18 19 16 24 32 61 19 19 19 Cost per IA (US$) 1,408 2,615 1,060 1,387 1,422 1,250 1,755 2,322 4,589 1,126 1,690 1,408 Net savings per IA (US$) 5,992 4,785 6,340 6,013 5,978 6,150 5,645 5,078 2,811 6,274 5,710 1,992 Zambia IA 339,632 188,110 441,327 252,260 570,773 536,129 321,756 215,692 407,557 444,352 444,352 339,632 VMMC per IA 8 15 7 8 8 7 10 14 9 8 8 8 Cost per IA (US$) 869 1,560 671 865 874 769 1,082 1,440 948 695 1,043 869 Net savings per IA (US$) 6,531 5,840 6,729 6,535 6,526 6,631 6,318 5,960 6,452 6,705 6,357 2,531 Zimbabwe IA 565,749 311,075 691,348 320,894 686,019 699,894 375,010 231,007 520,864 565,751 565,751 565,749 matadCs fVM o I Prevention HIV for VMMC of Cost and Impact VMMC per IA 4 7 3 4 4 3 6 9 4 4 4 4 Cost per IA (US$) 369 669 300 370 390 309 523 795 401 292 437 369 Net savings per IA (US$) 7,031 6,731 7,100 7,030 7,010 7,091 6,877 6,605 6,999 7,108 6,963 3,031
Table 6 presents sensitivity analysis for the following key parameters: VMMC effectiveness, VMMC coverage, time to scale-up, level of post-circumcision behavior change, VMMC unit cost, and lifetime HIV treatment cost. aBC is post-circumcision risk behavior change. bNo models were run for the scale-up to 50% VMMC coverage scenario for Mozambique and Tanzania, since national pre-scale-up VMMC coverage in these two countries was greater than 50%. cScaling up VMMC in the presence of 30% post-circumcision risk behavior change in Rwanda resulted in an increase in HIV infections. IA, infection(s) averted. doi:10.1371/journal.pmed.1001132.t006 Impact and Cost of VMMC for HIV Prevention limited number of sites. Our VMMC unit cost, which is higher antiretroviral treatment as a result of seeking VMMC services, it is than that used in previous studies [8,11,59,61,62], reflects a better possible that the prevention benefits of VMMC could be understanding of the service delivery components required for underestimated. such a program, including waste management and logistics The past year has been marked by important advances in HIV associated with the scaling up of VMMC services [52]. It should prevention, including the finding that early treatment initiation also be noted that countries are procuring the same circumcision reduces sexual transmission by as much as 96% in stable kits and are adopting the same service delivery model that was serodiscordant couples [66]. However, the population-level impact costed in this study. If possible, we would also have adjusted of treatment is not currently well known. And coverage of those VMMC unit cost over time and scale [63,64]. However, no eligible for antiretroviral treatment under WHO guidelines information is currently available on how VMMC unit cost might remains low, with less than half of people needing treatment vary over time and with scale. Another limitation is that we do not having access to it in eastern and southern Africa in 2010 [67]. include costs associated with demand creation in our study [65]. VMMC is a one-time surgical procedure with the promise of Finally, it would have been preferable for us to obtain country- substantial impact on population-level HIV incidence. Its effects specific ART costs; however, country-specific data for the 13 accrue to both men and women, and it compares favorably with countries studied were not available. other HIV prevention programs in terms of cost-effectiveness. HIV prevalence and incidence estimates used in the model and Countries are adopting WHO/UNAIDS recommendations to how they compare to the actual present and projected prevalence incorporate VMMC within their HIV prevention portfolios. Best and incidence in each of the selected countries is another factor practices from Kenya [68], Tanzania [69], and South Africa [56] that affects the accuracy of our findings. A steeper decline in show that with government leadership, community involvement, baseline incidence, for example, will result in fewer infections and collaboration among partners, it is possible to scale up rapidly averted as a result of VMMC. Conversely, if these incidence trends by implementing service delivery models that take into consider- are overly optimistic, then the number of infections available to be ation available resources, efficiency, and quality to achieve the averted by VMMC may in turn be larger than estimated here. maximum public health impact. To ensure that high coverage We also recognize that our modeling assumes that males seeking rates are achievable, countries and their international partners out VMMC services are typical of the general male population in should allocate enough resources to take advantage of the returns the selected age range. If, for example, those seeking out VMMC on investment predicted for this biomedical HIV prevention services are in fact those who are at least risk of becoming infected modality. (perhaps VMMC clients already disproportionately use condoms), then the benefits of VMMC are likely to be overestimated. Conversely, if those who are at most risk of infection (those with a Supporting Information large number of partners and/or low condom use) are dispropor- Table S1 Model details and equations. tionately attracted to VMMC services, then the modeling (DOCX) presented in this paper may underestimate the value of VMMC scale-up. Given that VMMC client sexual behavior, relative to the Acknowledgments behavior of the general male population, remains unknown, it is not possible to determine whether our projections are underesti- The views expressed in this paper are those of the authors and do not mated or overestimated. necessarily reflect the official policy or position of the US Government. There are also a number of prevention benefits that are indirectly associated with seeking out VMMC services. These Author Contributions include the benefits of counseling and testing, STI treatment, and Conceived and designed the experiments: EN SF JR LB CH. Performed early antiretroviral treatment. For example, for those males who the experiments: EN SF LB MO VM. Analyzed the data: EN SF LB MO seek out VMMC services but find that they are already HIV- VM. Contributed reagents/materials/analysis tools: EN SF LB MO VM infected, there could be societal prevention benefits if these men JR DC CH JS NH TF. Wrote the first draft of the manuscript: MO. subsequently are able to access early treatment and reduce their Contributed to the writing of the manuscript: EN SF JR MO LB NH DC infectivity. Since we focus only on the direct benefits received by JS TF VM CH. ICMJE criteria for authorship read and met: EN SF JR those receiving VMMC and do not quantify any additional MO LB NH DC JS TF VM CH. Agree with manuscript results and benefits of counseling and testing, STI treatment, and early conclusions: EN SF JR MO LB NH DC JS TF VM CH.
References 1. Auvert B, Taljaard D, Lagarde E, Sobngwi-Tambekou J, Sitta R, et al. (2005) 6. Hallett TB, Singh K, Smith JA, White RG, Abu-Raddad LJ, et al. (2008) Randomized, controlled intervention trial of male circumcision for reduction of Understanding the impact of male circumcision interventions on the spread of HIV in HIV infection risk: The ANRS 1265 Trial. PLoS Med 2: e298. doi:10.1371/ southern Africa. PLoS ONE 3: e2212. doi:10.1371/journal. pone.0002212. journal.pmed.0020298. 7. Nagelkerke NJ, Moses S, de Vlas SJ, Bailey RC (2007) Modelling the public 2. Bailey RC, Moses S, Parker CB, Agot K, Maclean I, et al. (2007) Male health impact of male circumcision for HIV prevention in high prevalence areas circumcision for HIV prevention in young men in Kisumu, Kenya: A in Africa. BMC Infect Dis 7: 16. randomised controlled trial. Lancet 369: 643–656. 8. White RG, Glynn JR, Orroth KK, Freeman EE, Bakker R, et al. (2008) Male 3. Gray RH, Kigozi G, Serwadda D, Makumbi F, Watya S, et al. (2007) Male circumcision for HIV prevention in sub-Saharan Africa: who, what and when? circumcision for HIV prevention in men in Rakai, Uganda: A randomised trial. AIDS 22: 1841–1850. Lancet 369: 657–666. 9. Williams BG, Lloyd-Smith JO, Gouws E, Hankins C, Getz WM, et al. (2006) 4. World Health Organization, Joint United Nations Programme on HIV/AIDS The potential impact of male circumcision on HIV in sub-Saharan Africa. PLoS (2007) New data on male circumcision and HIV prevention: policy and Med 3: e262. doi:10.1371/journal.pmed.0030262. programme implications. Available: http://libdoc.who.int/publications/2007/ 10. UNAIDS/WHO/SACEMA Expert Group on Modelling the Impact and Cost 9789241595988_eng.pdf. Accessed 31 October 2011. of Male Circumcision for HIV Prevention (2009) Male circumcision for HIV 5. Alsallaq R, Abu-Raddad L (2008) Male circumcision is a leading factor behind prevention in high HIV prevalence settings: what can mathematical modelling the differential HIV prevalence in sub-Saharan Africa [poster MOPE0254]. contribute to informed decision making? PLoS Med 6: e1000109. doi:10.1371/ XVII International AIDS Conference; Mexico City, Mexico; August 2008. journal.pmed.1000109.
26 Impact and Cost of VMMC for HIV Prevention
11. Auvert B, Marseille E, Korenromp EL, Lloyd-Smith J, Sitta R, et al. (2008) 38. Namibian Ministry of Health and Social Services, Macro International I (2008) Estimating the resources needed and savings anticipated from roll-out of adult Namibia demographic and health survey 2006–07. Windhoek (Namibia): male circumcision in sub-Saharan Africa. PLoS ONE 3: e2679. doi:10.1371/ Ministry of Health and Social Services. journal.pone.0002679. 39. Lesotho Ministry of Health and Social Welfare, Lesotho Bureau of Statistics, 12. Tobian AA, Serwadda D, Quinn TC, Kigozi G, Gravitt PE, et al. (2009) Male ORC Macro (2005) Lesotho demographic and health survey 2004. Calverton circumcision for the prevention of HSV-2 and HPV infections and syphilis. (Maryland): ORC Macro. N Engl J Med 360: 1298–1309. 40. Malawi National Statistical Office, ORC Macro (2005) Malawi demographic 13. Auvert B, Sobngwi-Tambekou J, Cutler E, Nieuwoudt M, Lissouba P, et al. and health survey 2004. Calverton (Maryland): ORC Macro. (2009) Effect of male circumcision on the prevalence of high-risk human 41. Shisana O, Rehle T, Simbayi LC, Zuma K, Jooste S, et al. (2009) South African papillomavirus in young men: results of a randomized controlled trial conducted national HIV prevalence, incidence, behaviour and communication survey in Orange Farm, South Africa. J Infect Dis 199: 14–19. 2008: A turning tide among teenagers? Cape Town: HSRC Press. 14. Gray RH, Serwadda D, Kong X, Makumbi F, Kigozi G, et al. (2010) Male 42. Tanzania Commission for AIDS, Zanzibar AIDS Commission, National Bureau circumcision decreases acquisition and increases clearance of high-risk human of Statistics, Office of the Chief Government Statistician, Macro International papillomavirus in HIV-negative men: a randomized trial in Rakai, Uganda. (2008) Tanzania HIV/AIDS and malaria indicator survey 2007–08. Dar es J Infect Dis 201: 1455–1462. Salaam (Tanzania): Tanzania Commission for AIDS. 15. Serwadda D, Wawer MJ, Makumbi F, Kong X, Kigozi G, et al. (2010) 43. Stover J, Kirmeyer S (2008) DemProj: A computer program for making Circumcision of HIV-infected men: effects on high-risk human papillomavirus population projections. Washington (District of Columbia): Futures Group infections in a randomized trial in Rakai, Uganda. J Infect Dis 201: 1463–1469. International, Health Policy Initiative. 16. Wawer MJ, Tobian AA, Kigozi G, Kong X, Gravitt PE, et al. (2011) Effect of 44. United Nations Population Division (2008) World population prospects: the circumcision of HIV-negative men on transmission of human papillomavirus to 2008 revision. New York: United Nations Department of Economic and Social HIV-negative women: a randomised trial in Rakai, Uganda. Lancet 377: Affairs. 209–218. 45. Chiwevu CM (2010) Costing of male circumcision in Zambia and the impacts of 17. Gray RH, Kigozi G, Serwadda D, Makumbi F, Nalugoda F, et al. (2009) The scaling up the male circumcision programme. Washington (District of effects of male circumcision on female partners’ genital tract symptoms and Columbia): Futures Group, Health Policy Initiative, Task Order 1. vaginal infections in a randomized trial in Rakai, Uganda. Am J Obstet Gynecol 46. Forsythe S (2010) Costing of male circumcision in Namibia and the impacts of 200: 42.e41–42.e47. scaling up the male circumcision program. Washington (District of Columbia): 18. Castellsague X, Bosch FX, Munoz N, Meijer CJ, Shah KV, et al. (2002) Male Futures Group, Health Policy Initiative, Task Order 1. circumcision, penile human papillomavirus infection, and cervical cancer in 47. Kioko U (2010) Costing of male circumcision in Kenya and the impacts of female partners. N Engl J Med 346: 1105–1112. scaling up the male circumcision programme. Washington (District of 19. Morris BJ, Gray RH, Castellsague X, Bosch X, Halperin DH, et al. (2011) The Columbia): Futures Group, Health Policy Initiative, Task Order 1. strong protective effect of circumcision against cancer of the penis. Adv Urol 48. Makerere University School of Public Health (2010) Assessing the potential 2011: 812368. impact and costs of scaling up medical male circumcision services in Uganda. 20. Mosconi AM, Roila F, Gatta G, Theodore C (2005) Cancer of the penis. Crit Washington (District of Columbia): Futures Group, Health Policy Initiative, Rev Oncol Hematol 53: 165–177. Task Order 1. 21. World Health Organization, Joint United Nations Programme on HIV/AIDS 49. Ozayr M, Asmal S, Forsythe S, Njeuhmeli E (2010) Costing of male (2010) Progress in male circumcision scale-up: country implementation and circumcision in South Africa and impacts of scaling up the circumcision research update. Geneva: World Health Organization, Available: http://www. programme. Washington (District of Columbia): Futures Group, Health Policy malecircumcision.org/documents/MC_country_June2010.pdf. Accessed 25 Au- Initiative, Task Order 1. gust 2011. 50. Schutte C, Forsythe S (2010) Costing of male circumcision in Zimbabwe and 22. World Health Organization (2008) Male circumcision quality assurance: a guide impacts of scaling up the circumcision programme. Washington (District of to enhancing the safety and quality of services. Geneva: World Health Columbia): Futures Group, Health Policy Initiative, Task Order 1. Organization. 51. Numbeo (2011) Cost of living comparison between two countries [database]. 23. World Health Organization, Joint United Nations Programme on HIV/AIDS, Available: http://www.numbeo.com/cost-of-living/compare_countries.jsp. Ac- Jhpiego (2008) Manual for male circumcision under local anaesthesia, version cessed 25 August 2011. 2.5C. Geneva: World Health Organization. 52. Edgil D, Stankard PS, Forsythe S, Reed J, Rech D, et al. (2011) Voluntary 24. World Health Organization (2010) Considerations for implementing models for medical male circumcision: logistics, commodities, and waste management optimizing the volume and efficiency of male circumcision services. Geneva: requirements for scale-up of services. PLoS Med 8: e1001128. doi:10.1371/ World Health Organization, Available: http://www.malecircumcision.org/ journal.pmed.1001128. programs/documents/mc_MOVE_2010_web.pdf. Accessed 25 August 2011. 53. Schwartlander B, Stover J, Hallett T, Atun R, Avila C, et al. (2011) Towards an 25. Bollinger L, DeCormier Plosky W, Stover J (2009) Male circumcision: Decision- improved investment approach for an effective response to HIV/AIDS. Lancet Makers’ Program Planning Tool—calculating the costs and impacts of a male 377: 2031–2041. circumcision program. Washington (District of Columbia): Futures Group, 54. Drummond MF, O’Brien B, Stoddart GL, Torrance GW (1999) Methods for Health Policy Initiative, Task Order 1. the economic evaluation of health care programmes. Oxford: Oxford University 26. Brown T, Bao L, Raftery AE, Salomon JA, Baggaley RF, et al. (2010) Modelling Press. HIV epidemics in the antiretroviral era: the UNAIDS Estimation and Projection 55. Haddix AC, Teutsch SM, Shaffer PA, Dunet DO (1996) Prevention Package 2009. Sex Transm Infect 86(Suppl 2): ii3–ii10. effectiveness: a guide to decision analysis and economic evaluation. Oxford: 27. Stover J (2009) A computer program for making HIV/AIDS projections and Oxford University Press. examining the demographic and social impacts of AIDS. Washington (District of 56. Lissouba P, Taljaard D, Rech D, Dermaux-Msimang V, Legeai C, et al. (2011) Columbia): Futures Group International, Health Policy Initiative. Adult male circumcision as an intervention against HIV: an operational study of 28. Cassell MM, Halperin DT, Shelton JD, Stanton D (2006) Risk compensation: uptake in a South African community (ANRS 12126). BMC Infect Dis 11: 253. the Achilles’ heel of innovations in HIV prevention? BMJ 332: 605–607. 57. Wawer MJ, Makumbi F, Kigozi G, Serwadda D, Watya S, et al. (2009) 29. Kalichman S, Eaton L, Pinkerton S (2007) Circumcision for HIV prevention: Circumcision in HIV-infected men and its effect on HIV transmission to female failure to fully account for behavioral risk compensation. PLoS Med 4: e138. partners in Rakai, Uganda: a randomised controlled trial. Lancet 374: 229–237. doi:10.1371/journal.pmed.0040138. 58. Galarraga O, Colchero MA, Wamai RG, Bertozzi SM (2009) HIV prevention 30. Botswana National AIDS Coordinating Agency (2005) Botswana AIDS impact cost-effectiveness: a systematic review. BMC Public Health 9(Suppl 1): S5. survey II: popular report. Gaborone (Botswana): Central Statistics Office. 59. Gray RH, Li X, Kigozi G, Serwadda D, Nalugoda F, et al. (2007) The impact of 31. Zambia Central Statistical Office, Ministry of Health, Tropical Diseases male circumcision on HIV incidence and cost per infection prevented: a Research Centre, Zambia Uo (2009) Zambia demographic and health survey stochastic simulation model from Rakai, Uganda. AIDS 21: 845–850. 2007. Calverton (Maryland): Macro International. 60. Thomas AG, Tran BR, Cranston M, Brown MC, Kumar R, et al. (2011) 32. Swaziland Central Statistical Office, Macro International (2008) Swaziland Voluntary medical male circumcision: a cross-sectional study comparing demographic and health survey 2006–07. Mbabane (Swaziland): Central circumcision self-report and physical examination findings in Lesotho. PLoS Statistical Office. ONE 6: e0027561. doi:10.1371/journal.pone.002756. 33. Zimbabwe Central Statistical Office, Macro International (2007) Zimbabwe 61. Binagwaho A, Pegurri E, Muita J, Bertozzi S (2010) Male circumcision at demographic and health survey 2005–06. Calverton (Maryland): Macro different ages in Rwanda: A cost-effectiveness study. PLoS Med 7: e1000211. International. doi:10.1371/journal.pmed.1000211. 34. Institut National de la Statistique du Rwanda, ORC Macro (2006) Rwanda 62. Kahn JG, Marseille E, Auvert B (2006) Cost-effectiveness of male circumcision demographic and health survey 2005. Calverton (Maryland): ORC Macro. for HIV prevention in a South African setting. PLoS Med 3: e517. doi:10.1371/ 35. Moc¸ambique Instituto Nacional de Estatı´stica, Ministe´rio da Sau´de, Measure journal.pmed.0030517. DHS+/ORC Macro (2005) Moc¸ambique Inque´rito Demogra`fico e de Sau´de 63. Bautista-Arredondo S, Gadsden P, Harris JE, Bertozzi SM (2008) Optimizing 2003. Calverton (Maryland): ORC Macro. resource allocation for HIV/AIDS prevention programmes: an analytical 36. Kenya National Bureau of Statistics, ICF Macro (2010) Kenya demographic and framework. AIDS 22(Suppl 1): S67–S74. health survey 2008–09. Calverton (Maryland): ICF Macro. 64. Marseille E, Dandona L, Marshall N, Gaist P, Bautista-Arredondo S, et al. 37. Ugandan Ministry of Health, ORC Macro (2006) Uganda HIV/AIDS sero- (2007) HIV prevention costs and program scale: data from the PANCEA project behavioural survey 2004–2005. Calverton (Maryland): ORC Macro. in five low and middle-income countries. BMC Health Serv Res 7: 108.
27 Impact and Cost of VMMC for HIV Prevention
65. Bertrand JT, Njeuhmeli E, Forsythe S, Mattison SK, Chideya S, et al. (2011) Organization, Available: http://whqlibdoc.who.int/publications/2010/ Voluntary medical male circumcision: a qualitative study exploring the 9789241500395_eng.pdf. Accessed 25 August 2011. challenges of costing demand creation in eastern and southern Africa. PLoS 68. Mwandi Z, Murphy A, Reed J, Chesang K, Njeuhmeli E, et al. (2011) Voluntary ONE 6: e0027562. doi:10.1371/journal.pone.0027562. medical male circumcision: translating research into the rapid expansion of 66. Cohen MS, Chen YQ, McCauley M, Gamble T, Hosseinipour MC, et al. (2011) services in Kenya, 2008–2011. PLoS Med 8: e1001130. doi:10.1371/ Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med journal.pmed.1001130. 365: 493–505. 69. Mahler HR, Kileo B, Curran K, Plotkin M, Adamu T, et al. (2011) Voluntary 67. World Health Organization, Joint United Nations Programme on HIV/AIDS, medical male circumcision: matching demand and supply with quality and United Nations Children’s Fund (2010) Towards universal access: scaling up efficiency in a high-volume campaign in Iringa Region, Tanzania. PLoS Med 8: priority HIV/AIDS interventions in the health sector. Geneva: World Health e1001131. doi:10.1371/journal.pmed.1001131.
28 Impact and Cost of VMMC for HIV Prevention
Editors’ Summary Background. Every year, about 2.5 million people (mainly through 2025 and would cost US$2,000,000,000 between in sub-Saharan Africa) become infected with HIV, the virus 2011 and 2025. However, it would result in net savings that causes AIDS. There is no cure for HIV/AIDS. (because of averted treatment and care costs) of Consequently, prevention of HIV transmission is very US$16,510,000,000. important. Because the most common HIV transmission route is through unprotected sex with an infected partner, What Do These Findings Mean? These findings suggest individuals can reduce their risk of HIV infection by that rapid VMMC scale-up in eastern and southern Africa is abstaining from sex, by having only one or a few partners, warranted, given its likely impact on the region’s HIV and by using male or female condoms. There is also strong epidemics and the resultant cost savings. However, the evidence that voluntary medical male circumcision accuracy of these findings depends on the assumptions built (VMMC)—the removal of the foreskin, the loose fold of into the DMPPT and on the data fed into it. For example, skin that covers the head of the penis—reduces the there could be risk behavior changes after circumcision. That heterosexual acquisition of HIV in men by about 60%. In is, risky sexual behaviors may increase in men who have 2007, the World Health Organization (WHO) and the Joint been circumcised. However, the researchers show that, United Nations Programme on HIV/AIDS (UNAIDS) except in Rwanda, post-circumcision risk behavior change recommended that VMMC should be offered to men as is unlikely to completely reverse the benefits of VMMC. These part of comprehensive HIV risk reduction programs in modeling results also assume that men seeking out VMMC settings with generalized HIV epidemics and low levels of services are typical of the general male population, but if male circumcision. They also prioritized 13 countries in they are actually at unusually low risk of HIV infection, then eastern and southern Africa for VMMC program scale-up. the benefits of VMMC reported here are likely to be overestimated. Finally, these findings assume 80% VMMC Why Was This Study Done? The impact of VMMC scale- coverage. This may be optimistic, although results from up in terms of HIV infections and AIDS deaths averted Kenya indicate that this target is achievable. Thus, countries (epidemiologic impact) is expected to be large, and the and their international partners must allocate sufficient intervention should also reduce the costs associated with the resources to VMMC scale-up to achieve high coverage treatment, care, and support of infected individuals. rates if they are to take full advantage of the benefits However, VMMC scale-up will require substantial funding predicted here for VMMC scale-up. and considerable effort by countries—many of which have weak health systems and limited resources—to train Additional Information. Please access these websites via personnel, equip facilities, and provide the necessary the online version of this summary at http://dx.doi.org/10. commodities. To support planning for VMMC scale-up, the 1371/journal.pmed.1001132. United States Agency for International Development Health Policy Initiative has collaborated with UNAIDS to develop the N This study is part of a PLoS Collection of articles on VMMC Decision Makers’ Program Planning Tool (DMPPT), a (http://www.ploscollections.org/VMMC2011) and is further mathematical model that allows analysts and decision discussed in a PLoS Medicine Review Article by Hankins makers to estimate the epidemiologic impact and cost of et al. (http://dx.doi.org/10.1371/journal.pmed.1001127) alternative VMMC scale-up programs. In this study, the N Information is available from WHO, UNAIDS, and PEPFAR researchers use DMPPT to estimate the impact and cost of on all aspects of HIV/AIDS; the 2011WHO/UNAIDS progress scaling up adult VMMC in the 13 priority countries in eastern report on VMMC scale-up in the 13 priority countries is and southern Africa. available N NAM/aidsmap provides basic information about HIV/AIDS, What Did the Researchers Do and Find? The researchers summaries of recent research findings on HIV care and derived VMMC unit costs for each priority country based on a treatment, and information on male circumcision for the cost assessment undertaken in Zimbabwe, one of the first prevention of HIV transmission countries to scale up VMMC services using WHO’s ‘‘Models for Optimizing Volume and Efficiencies’’ (MOVE) guidelines. N Information is available from Avert, an international AIDS They fed these costs and recent epidemiologic data charity on many aspects of HIV/AIDS, including informa- (including HIV infection rates and the effectiveness of tion on all aspects of HIV prevention, and on HIV/AIDS in VMMC in preventing HIV transmission) and demographic Africa (in English and Spanish) data (including the adult population size and pre-scale-up N The Clearinghouse on Male Circumcision, a resource male circumcision prevalence) collected in each country into provided by WHO, UNAIDS, and other international bodies, the DMPPT, together with information on the lifetime costs provides information and tools for VMMC policy develop- of HIV treatment. Results from running the DMPPT model ment and program implementation, including information suggest that scaling up adult VMMC to reach 80% coverage on the DMPPT and the MOVE guidance in the 13 priority countries by 2015 would require 20.33 N Personal stories about living with HIV/AIDS are available million circumcisions to be completed between 2011 and through Avert, through NAM/aidsmap, and through the 2015. To maintain this coverage, a further 8.42 million charity website Healthtalkonline circumcisions would be required between 2016 and 2025. Such a scale-up would avert 3.36 million new HIV infections
29 30 Voluntary Medical Male Circumcision: A Framework Analysis of Policy and Program Implementation in Eastern and Southern Africa
Kim E. Dickson1*, Nhan T. Tran2, Julia L. Samuelson1, Emmanuel Njeuhmeli3, Peter Cherutich4, Bruce Dick1¤, Tim Farley1, Caroline Ryan5, Catherine A. Hankins6 1 World Health Organization, Geneva, Switzerland, 2 Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, United States of America, 3 United States Agency for International Development, Washington, District of Columbia, United States of America, 4 Ministry of Public Health and Sanitation, Nairobi, Kenya, 5 Office of the U.S. Global AIDS Coordinator, United States Department of State, Washington, District of Columbia, United States of America, 6 Joint United Nations Programme on HIV/AIDS, Geneva, Switzerland
Abstract
Background: Following confirmation of the effectiveness of voluntary medical male circumcision (VMMC) for HIV prevention, the World Health Organization and the Joint United Nations Programme on HIV/AIDS issued recommendations in 2007. Less than 5 y later, priority countries are at different stages of program scale-up. This paper analyzes the progress towards the scale-up of VMMC programs. It analyzes the adoption of VMMC as an additional HIV prevention strategy and explores the factors may have expedited or hindered the adoption of policies and initial program implementation in priority countries to date.
Methods and Findings: VMMCs performed in priority countries between 2008 and 2010 were recorded and used to classify countries into five adopter categories according to the Diffusion of Innovations framework. The main predictors of VMMC program adoption were determined and factors influencing subsequent scale-up explored. By the end of 2010, over 550,000 VMMCs had been performed, representing approximately 3% of the target coverage level in priority countries. The ‘‘early adopter’’ countries developed national VMMC policies and initiated VMMC program implementation soon after the release of the WHO recommendations. However, based on modeling using the Decision Makers’ Program Planning Tool (DMPPT), only Kenya appears to be on track towards achievement of the DMPPT-estimated 80% coverage goal by 2015, having already achieved 61.5% of the DMPPT target. None of the other countries appear to be on track to achieve their targets. Potential predicators of early adoption of male circumcision programs include having a VMMC focal person, establishing a national policy, having an operational strategy, and the establishment of a pilot program.
Conclusions: Early adoption of VMMC policies did not necessarily result in rapid program scale-up. A key lesson is the importance of not only being ready to adopt a new intervention but also ensuring that factors critical to supporting and accelerating scale-up are incorporated into the program. The most successful program had country ownership and sustained leadership to translate research into a national policy and program. Please see later in the article for the Editors’ Summary.
Citation: Dickson KE, Tran NT, Samuelson JL, Njeuhmeli E, Cherutich P, et al. (2011) Voluntary Medical Male Circumcision: A Framework Analysis of Policy and Program Implementation in Eastern and Southern Africa. PLoS Med 8(11): e1001133. doi:10.1371/journal.pmed.1001133 Academic Editor: Stephanie L. Sansom, Centers for Disease Control and Prevention, United States of America Received June 2, 2011; Accepted October 19, 2011; Published November 29, 2011 This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication. Funding: This work was funded by WHO and PEPFAR through USAID. The funders (WHO and USAID) had roles in the study design, decision to publish, and preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. Abbreviations: DMPPT, Decision Makers’ Program Planning Tool; DOI, Diffusion of Innovations; PEPFAR, United States President’s Emergency Plan for AIDS Relief; RCT, randomized controlled trial; UNAIDS, Joint United Nations Programme on HIV/AIDS; VMMC, voluntary medical male circumcision; WHO, World Health Organization * E-mail: [email protected] ¤ Current address: Department of Population, Family and Reproductive Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, United States of America
31 Lessons from Male Circumcision Program Scale-Up
Introduction assurance and improvement, human resource development, commodity security, social change communication, and monitor- In 2009, more than 25 y after HIV was first identified, 2.6 ing and evaluation [20]. million people had become infected, and there were an estimated A consensus report of six different models showed that in 33.3 million people living with HIV worldwide [1]. In the absence settings with high HIV prevalence and low circumcision of a vaccine, the next best means to combat new HIV infections is prevalence, one new HIV infection could be averted for every the implementation of evidence-based prevention strategies five to 15 circumcisions performed [21]. More detailed modeling including male and female condoms [2], antiretroviral prophylaxis was incorporated into the Decision Maker’s Program Planning to prevent vertical transmission from mother to child [3,4], harm Tool (DMPPT), and it is estimated that scaling up circumcision reduction for people who inject drugs [5], and, most recently, programs to reach 80% of adult uncircumcised men within 5 y in treating HIV-infected people in serodiscordant couples with 13 priority countries would require a total of 20.3 million antiretroviral drugs to reduce transmission to partners [6]. circumcisions to be performed and a further 8.4 million between However, the scale-up of these HIV prevention programs remains 2016 and 2025, averting an estimated 3.4 million new HIV challenging despite evidence demonstrating their effectiveness. infections and 386,000 AIDS deaths through 2025 [22]. Despite The use of male and female condoms, despite decades of convincing results demonstrating that VMMC is an effective, cost- promotion and distribution, remains suboptimal. The use of saving intervention in the fight against HIV, there have been female condoms as a prevention strategy for HIV is reportedly major challenges and barriers to implementing programs in the lower than male condom use, including in countries in sub- high priority countries. Saharan Africa, where the majority of new HIV infections are This paper analyzes the progress towards achievement of occurring through heterosexual transmission [7] and are predom- VMMC for HIV prevention program scale-up in 13 priority inantly among women. countries. We analyze the adoption of VMMC as an additional Evidence demonstrating the effectiveness of voluntary medical HIV prevention strategy. We further explore the factors that may male circumcision (VMMC) in preventing HIV sexual transmis- have expedited or hindered the adoption of VMMC policies and sion was first released in 2005 from the South Africa (Orange strategies as well as initial program implementation in the 13 Farm) randomized controlled trial (RCT) [8]. This was followed priority countries to date. We also explore the factors that may by results in 2006 from RCTs in Uganda (Rakai District) and have influenced subsequent program scale-up. Challenges en- Kenya (Kisumu) [9,10]. All three RCTs confirmed that male countered and lessons learned are highlighted for application to circumcision performed by well-trained and equipped medical countries still in the early stages of VMMC scale-up as well as to providers is safe and reduces the risk of heterosexual acquisition of other potential HIV prevention strategies such as topical (vaginal HIV infection among men by as much as 60%. These RCT results and rectal microbicides) and oral pre-exposure prophylaxis. confirmed decades of evidence from observational studies suggesting male circumcision’s strong protective effect for men Methods against HIV [11]. Male circumcision also has a strong protective effect against other sexually transmitted infections in men and in The actual numbers of VMMCs performed in priority countries women [12–16]. Although there is no conclusive evidence that per calendar year since 2008 were recorded and the totals used to medical male circumcision has a direct effect on women’s risk of classify countries into five adopter categories according to the HIV infection [17], a systematic review, largely based on Diffusion of Innovations (DOI) framework. Country progress observational studies, estimated an overall 20% lower HIV towards achievement of the goal of circumcising 80% of eligible incidence in female partners of circumcised men, compared with men was also calculated [22]. partners of uncircumcised men [18]. The DOI and ExpandNet frameworks were selected to analyze Following the release of the results of the RCTs, in 2007 the the status of VMMC programming in the 13 priority countries as World Health Organization (WHO) and the Joint United Nations they are complimentary theories that both refer to the process by Programme on HIV/AIDS (UNAIDS) convened an international which innovations are disseminated and taken to scale. The DOI consultation of stakeholders from a range of disciplines to review and ExpandNet frameworks highlight the importance of the the body of evidence from the three trials and the wealth of earlier nature of the innovation itself and the sociopolitical context or ecological and observational studies [19]. The consultation environment in which diffusion takes place (see Figure 1). Whereas resulted in a firm endorsement of the evidence from the three the DOI theory addresses the diffusion of any innovation, the trials, and the formulation of eleven key conclusions and WHO ExpandNet framework explicitly addresses the adoption recommendations for the implementation and scale-up of VMMC and scale-up of public health programs and services in the health programs in countries and settings with generalized high- sector. Both frameworks stress that innovations that are Credible, prevalence HIV epidemics and low levels of male circumcision Observable, and Relevant; have Relative advantages; are Easy to [19]. install and understand; and are Compatible and Testable WHO and UNAIDS identified 13 countries in southern and (‘‘CORRECT’’) are most likely to be successfully adopted and eastern African as high-priority countries for the implementation scaled up [23]. and rapid scale-up of VMMC programs: Botswana, Kenya, The DOI theory defines diffusion as the process by which an Lesotho, Malawi, Mozambique, Namibia, Rwanda, South Africa, innovation is communicated through certain channels over time Swaziland, Tanzania, Uganda, Zambia, and Zimbabwe. These among members of a social system [24]. The DOI theory posits countries have been working towards the implementation of that regardless of the setting, initially there are a few individuals VMMC programs using operational guidance developed by who adopt innovations, known as the ‘‘innovators’’—these are WHO and UNAIDS. The guidance emphasizes ten essential those who are willing to adopt new ideas before they are components of program implementation: leadership and partner- mainstream. Following this, another small proportion, the ‘‘early ships, situation analysis, advocacy, enabling policy and regulatory adopters,’’ of the population follow. The ‘‘early majority’’ is the environments, national strategy and operational plans, quality next group to adopt the innovation. By this time, more than half of
32 Lessons from Male Circumcision Program Scale-Up
Figure 1. Comparison of Diffusion of Innovation and ExpandNet frameworks. A comparison of the DOI and ExpandNet frameworks is shown. Whereas the DOI describes the process through which innovations are adopted and diffused through the population, the ExpandNet framework specifically addresses the diffusion and scale-up of public health interventions. As highlighted in the figure, the ExpandNet framework reflects the earlier thinking of the DOI and incorporates the elements of context as environment, communication channels as scale-up strategy, and adoption by the population as adoption by the organization implementing the intervention. doi:10.1371/journal.pmed.1001133.g001 the population has adopted the innovation. The ‘‘late majority’’ is (no progress) to 3 (maximum progress). A total scale-up score, a the next group to accept new interventions; the ‘‘laggards’’ are the composite indicator of progress in program scale-up, was last to adopt innovations, they represent those who are least likely calculated for each country. Additionally, further analyses were to accept innovation (see Figure 2). While the DOI theory refers to undertaken in order to understand if there were differences in the the adoption pattern of individuals within a community, we contributions of the six key elements to the overall DOI adoption applied these concepts to analyze the adoption of VMMC policies, status, since one can argue that some elements are better strategies, and initial program implementation by countries. The predictors of adoption and scale-up than others. To do this, the DOI framework was used to define the adoption status of each association between the DOI classification ‘‘adopter status’’ and country to comparatively assess progress towards scale-up among the scores on the individual elements of scale-up was determined the 13 priority countries. using a crude estimation of association based on an R-squared In order to facilitate analysis using the DOI framework, six analysis. program components based on key elements defined as essential The ExpandNet framework is based on the theory underpin- for the operationalization of VMMC programs described in the ning the DOI model and supplemented with experiences apply- WHO/UNAIDS operational guidance were used to assess overall ing the model in low-resource settings [25]. The ExpandNet country progress towards VMMC program scale-up [20]. The six framework views scaling up as an open system that draws on five elements selected were those for which progress towards scale interacting elements: the innovation itself, the resource team, the up over a 3-y period could be objectively quantified. A country scale-up strategy, the user organization, and the environment. We was assigned a score for each key element, ranging from 0 used the ExpandNet framework to explore the factors that may
33 Lessons from Male Circumcision Program Scale-Up
Figure 2. Diffusion of Innovation adoption status. The typical distribution of categories of adopters according to the DOI theory is shown in this figure. The very first adopters or innovators typically represent 2.5% of the population. The next group of adopters, called early adopters, represent about 13.5% of the population and are followed by the early majority, who represent approximately 34% of the total population. The next 34% of the population to adopt are the late majority; this group is followed by the remaining 16% of the population, who are the laggards, the last to adopt a new innovation. doi:10.1371/journal.pmed.1001133.g002 have influenced subsequent program scale-up, highlighting factors already achieved 61.5% of the DMPPT target. None of the other specific to VMMC. countries—including the early adopters—appear to be on track to achieve their targets. Results Table 3 shows the scoring scheme for each of the six key elements derived from the WHO/UNAIDS operational guidance, The ‘‘CORRECT’’ attributes of VMMC are highlighted in reflecting also the time when the different milestones were Table 1. Male circumcision, an old procedure but an innovation achieved (earlier completion resulting in higher scores). Table 4 for HIV prevention, has many of the ‘‘CORRECT’’ attributes shows the scores for each element by country and reflects the needed to enhance scalability. VMMC is Credible, with consistent progress in scaling up key program elements in all of the priority evidence of reduced HIV rates from numerous Observational countries. The total scale-up scores calculated for each of the 13 studies and RCTs conducted by respected researchers in Africa, countries ranged from a low of 4 (Mozambique) to a high of 17 where the intervention is Relevant, as it provides a new solution to (Kenya) out of a maximum score of 18. To date, all countries have address a high-burden public health problem; VMMC has conducted a situation analysis to assess the acceptability of Relative advantages, as it is a one-time intervention conferring introducing VMMC programs with the support of a WHO toolkit lifetime reduction in risk of HIV infection. Medical male [27]. At least seven countries had conducted their situation circumcision is not particularly Easy to install, as it is a surgical analyses in 2008, within a year of the release of the WHO/ intervention that requires training and expertise, with complex UNAIDS recommendations. Beyond the conduct of the situation social and cultural barriers that need to be overcome as programs analysis, progress towards the other key elements of scale-up varies are scaled up. However, it is a one-time intervention, provides a significantly among the 13 countries. rare opportunity to reach men, and it is cost saving [22,26]. In particular, large variation is seen in the leadership scores, Medical male circumcision is Compatible with existing national with only Botswana, Kenya, Rwanda, and Swaziland having priorities for HIV prevention in the priority countries. In all clearly identifiable prominent national champions for VMMC. priority countries, VMMC services have been or are being pilot We found that although almost all countries had a dedicated Tested, providing critical information for moving to scale. national focal person for VMMC in place, Botswana and Kenya Table 2 shows that between 2008 and 2010, an estimated total had identified theirs within the first year of scale-up. Twelve out of of 559,528 VMMCs for HIV prevention have been done in the 13 the 13 countries had nationally approved policies for VMMC priority countries, with a large majority (417,974) done in 2010. scale-up by the end of 2010, with Botswana and Kenya having Kenya has carried out the largest number (232,287) of VMMCs, their policies in place within a year of the WHO/UNAIDS followed by South Africa (145,475) and then Zambia (81,849). recommendations. By the end of 2010, only Mozambique and Table 2 also shows the classification of countries according to DOI Uganda did not have nationally approved scale-up strategies. adopter status. The innovator (8% of countries), Kenya, started Figure 3 summarizes the results of the R-squared analysis for performing VMMCs in or before 2008 and reached more than each of the six key elements of scale-up identified in this study. The 10,000 VMMCs in 2008. Early adopters (23% of countries), South figure shows each of the 13 countries according to their DOI Africa, Zambia, and Swaziland, started in or before 2008. The classification and their scores on the individual elements of early majority (38% of countries), Botswana, Zimbabwe, Tanza- VMMC program scale-up (Table 4), together with a crude nia, Namibia, and Mozambique started in 2009. The late majority estimate of association (R-squared) computed by assigning a linear (15% of countries), Uganda and Rwanda, started in 2010 and did score to the DOI classes. The association values can range from 0 more than 1,000 VMMCs. The laggards (15% of countries), to 1, with an R-squared value of 1 indicating perfect association. Malawi and Lesotho, did fewer than 1,000 VMMCs in 2010. The For this analysis, the association values ranged from 0.16 to 0.57. distribution of the total scale-up scores reflecting adoption status As the results suggest, conducting a pilot program (R-squared = among the 13 countries in this study compares well with the 0.57) may be the most important predictor of DOI class, followed distribution proposed by the DOI theory. Also highlighted in by establishing a male circumcision focal point (R-squared = Table 2, only Kenya appears to be on track towards achievement 0.27), developing a national male circumcision policy (R-squared of the DMPPT-estimated 80% coverage goal by 2015, having = 0.27), and developing a national implementation strategy
34 Lessons from Male Circumcision Program Scale-Up
Table 1. ExpandNet innovation characteristics.
Innovation Element Key Questions for Scale-Up Male Circumcision
Credible 1. Have results of pilot testing the innovation been Ecological studies show lower prevalence of HIV infection in documented? 2. How sound is the evidence? 3. Is countries with high rates of circumcision. further evidence/better documentation needed? 4. Has the innovation been tested in the type of setting where it will be scaled up? Epidemiological studies show that circumcised men have a consistently lower incidence of HIV than uncircumcised men, even after adjusting for differences in sexual behavior. Three independent RCTs showed that circumcision reduced the risk of HIV infection in young men. Research was conducted by credible researchers in directly relevant settings in African countries. Male circumcision is not a new procedure, but is an innovation for HIV prevention, with potential for significant impact in countries with generalized HIV epidemics and low prevalence of circumcision. Observable How observable are results? Results from the epidemiological studies and RCTs are unequivocal in demonstrating lower HIV incidence in circumcised men. Impact of program scale-up on incidence of HIV infection will take some time to be realized. Impact on AIDS and AIDS-related mortality will be even more distal. Modeling, costing, and impact studies indicate that VMMC is cost saving and will benefit both men and women. Relevant Does the innovation address a felt need, VMMC addresses the persistent problem of finding ways to persistent problem, or policy priority? prevent HIV in generalized heterosexual epidemics. It is directly relevant in southern and eastern African countries that have the greatest HIV incidence and burden of infection. Relative advantages 1. Does the innovation have relative advantage over VMMC is only a one-time intervention, resulting in lifelong existing practices? 2. Is it more cost-effective than lower risk of HIV infection in men. existing practices or alternatives? VMMC programs are a rare opportunity to reach young men through health services and provide good sexual and reproductive health and HIV risk reduction counseling. VMMC has been demonstrated to be highly cost-effective and cost saving for all priority countries. The potential impact is substantially greater than other HIV prevention interventions. Ease to install 1. What degree of change from current norms, practices, VMMC is a challenging intervention to implement since it and levels of resources is implied in the innovation? 2. requires surgical skills that are in short supply in the Africa region. What is the level of technical sophistication needed to introduce the innovation? 3. Are major additional human or financial resources and commodities needed to introduce the innovation? The number of circumcisions necessary to achieve rapid impact on the HIV epidemic is large, with consequently potentially large implications for human resources, facilities, and supplies. VMMC is a straightforward minor outpatient surgical procedure, but must be performed by adequately trained and equipped teams. Compatible 1. Is the innovation compatible with current values or VMMC is consistent with already existing national priorities services of the user organization? 2. Will it be for comprehensive HIV prevention. difficult to maintain the basic values of the innovation as expansion proceeds? 3. Will changes in logistics need to be made to accommodate the innovation? 4. Which components will need local adaptation to be relevant for changes in local context? There are a wide range of sociocultural factors that need to be considered when scaling up VMMC programs. Countries need to ensure that VMMC is promoted in a culturally sensitive way and does not introduce stigma associated with circumcision status. The implications of VMMC for women also need to be taken into account when scaling up programs. Testable Can the user organization test the innovation Pilot projects have been set up in all 13 priority countries and tailored in stages without fully adopting it? to local contexts. The pilots have provided information for subsequent program scale-up.
35 Lessons from Male Circumcision Program Scale-Up
Table 1. Cont.
Innovation Element Key Questions for Scale-Up Male Circumcision
Since VMMC scale-up requires substantial infrastructure and human and financial resources, incremental approaches to scale-up have been used. Best combination for service delivery scale-up is yet to be determined, as well as how to balance supply and demand creation.
doi:10.1371/journal.pmed.1001133.t001
(R-squared = 0.26). The role of having a national champion (R- Kenya, and the early adopters (South Africa, Zambia, and squared = 0.17) and conducting a situation analysis (R-squared Swaziland) initiated VMMC program implementation soon after = 0.16) was shown to have a weaker predictive value on the the release of the recommendations by adopting national policies likelihood of VMMC adoption and program scale-up. and strategies and starting to perform VMMCs as part of pilot Specific examples of how the ‘‘user organization,’’ ‘‘resource programs in 2008. However, only Kenya appears to be on track to team,’’ scaling-up strategy, and environment influenced scale-up achieving 80% coverage by 2015. To date, VMMC programs in are highlighted in Box 1. The type of policy document developed priority countries have reached only approximately 3% of the varies by country: some countries have stand-alone policy target coverage level of 80% of eligible African men proposed by documents, while others have VMMC incorporated into other Njeuhmeli et al. [22]. Clearly an accelerated pace of VMMC HIV prevention policies. Governments have been engaged in the service delivery is needed to take programs to scale and to setting up of pilot sites to test the feasibility of scale-up in all maximize the impact of the programs. countries. Although the rate of adoption of VMMC programs varies among the 13 priority countries, the initiation of program Discussion adoption and initial implementation as a whole have been carried out relatively quickly compared to other public health interven- We found that the DOI theory was most useful in analyzing the tions. Although there is much more research and scientific adoption and initial implementation of VMMC programs, while discovery now than ever before, the uptake of innovations does the ExpandNet theory helped to explore the factors that have not seem to be much faster than it was 100 y ago. A review of nine facilitated or hindered scale-up. The analysis shows that although landmark clinical procedures suggested that, on average, it takes a progress in VMMC program implementation has been made in all minimum of 6.3 y for research evidence to reach reviews, papers, 13 countries within the first 3 y of the release of the WHO/ and text books. They estimated that a further 9.3-y transition UNAIDS recommendations, much more needs to be done if period is needed to implement the evidence from scientific countries are to take programs to scale and achieve their targets of publications [28]. ExpandNet case studies of scaling up reproduc- circumcising 80% of eligible men by 2015. The innovator country, tive health interventions indicate that about 10 y is required to
Table 2. Service delivery statistics.
Estimated Number Achievement DOI Adopter Status of VMMCs to Reach towards 80% Countries Number of Male Circumcisions Done in Each Calendar Year Classification 80% Coverage Coverage 2008 2009 2010 Total
Kenyaa 11,663 80,719 139,905 232,287 Innovator 377,788 61.5% South Africa 5,190 9,168 131,117 145,475 Early adopter 4,333,134 3.4% Zambia 2,758 17,180 61,911 81,849 Early adopter 1,949,292 4.2% Swaziland 1,110 4,336 18,869 24,315 Early adopter 183,450 13.3% Botswana 0 5,424 5,773 11,197 Early majority 345,244 3.2% Zimbabwe 0 2,801 11,176 13,977 Early majority 1,912,595 0.7% Tanzania 0 881 28,562 29,443 Early majority 1,373,271 2.1% Namibia 0 224 1,763 1,987 Early majority 330,218 0.6% Mozambique 0 100 7,633 7,733 Early majority 1,059,104 0.73 Uganda 0 0 9,052 9,052 Late majority 4,245,184 0.2% Rwanda 0 0 1,694 1,694 Late majority 1,746,052 0.1% Malawi 0 0 300 300 Laggard 2,101,566 ,0.1% Lesotho 0 0 219 219 Laggard 376,795 0.1% Total 20,721 120,833 417,974 559,528 20,333,693 2.68%
These data were compiled by the PEPFAR Male Circumcision Technical Working Group and largely reflect data collated from sites funded by this agency. aNyanza Province only. doi:10.1371/journal.pmed.1001133.t002
36 Lessons from Male Circumcision Program Scale-Up
Table 3. VMMC key elements of program scale-up scoring key.
Leadership: Prominent Leadership: National Strategy Situational National National VMMC Policy or and Operational/ Pilot/Demonstration Analysis Completed Champion Dedicated Focal Similar Guidance Implementation Sites: Government Score (Full or Selective) Engaged Point in Place Approved Plan Approved Involvement
3CompletedallorInfluential national National VMMC task Formal policy or Approved by Pilots set up with some elements leader/advocate force constituted guidance, either end of 2008 government engagement before the end engaged for that meets regularly separate or integrated by 2008 of 2008 VMMC in 2007 by end of 2008 into other national policy, approved by end 2008 2CompletedallorInfluential national National VMMC task Policy or equivalent Approved during Pilots set up with some elements leader/advocate force constituted by approved during 2009 2009 government engagement before the end engaged for end of 2009 by 2009 of 2009 VMMC in 2009 1CompletedallorInfluential national National VMMC task Draft policy, not yet Draft or approved Pilots set up with some elements leader/advocate force constituted by approved or completed, during 2010 government engagement before the end engaged for end of 2010 during 2010 by 2010 of 2010 VMMC in 2010 0 None, or initial No leader/advocate No national VMMC No policy or policy None No government steps in progress engaged in VMMC task force guidance involvement or no early in the process established pilot programs
doi:10.1371/journal.pmed.1001133.t003 scale up from pilot testing to nationwide expansion [25,29]. policies/strategies for HIV prevention that included VMMC and Whereas in the case of VMMC, within 4 mo of the results of the had begun to initiate service delivery. However, although the three RCTs being released (in December 2006), WHO/UNAIDS uptake of VMMC RCT evidence was relatively quick, it took had endorsed VMMC as a safe and effective means of HIV almost two decades from the first cohort data [30] to the RCT prevention and published recommendations for the implementa- meta-analysis, and implementation research is still insufficient. tion and scale-up of VMMC programs [19]. Less than a year later, Potential predictors of innovation and early adoption of the pilot programs were initiated in the priority countries. Within 3 y VMMC programs identified by this analysis include having a of the release of the WHO/UNAIDS recommendations, the VMMC focal person, establishing a national policy, and having an majority of the 13 priority countries had established national operational strategy, as well as having a pilot or demonstration site
Table 4. Country progress with scaling- up VMMC programs in focal countries (December 2010).
Total Scale-Up Country Key Elements of VMMC Program Scale-Up Score) Leadership: Situational Prominent Leadership: National Strategy Pilot/ Analysis National National VMMC Policy or and Operational/ Demonstration Completeda Champion Dedicated Focal Similar Guidance Implementation Sites: Government (Full or Selective) Engaged Point in Place Approveda Plan Approveda Involvement
Botswana 3 3 3 3 2 2 16 Kenya 3 3 3 3 2 3 17 Lesotho 3 0 1 1 1 1 7 Malawi 1 0 1 1 1 1 5 Mozambique 3 0 0 0 0 1 4 Namibia 3 0 2 1 1 2 9 Rwanda 1 1 1 1 1 2 7 South Africa 2011127 Swaziland 3 1 2 2 2 3 13 Tanzania 2 0 1 1 2 2 8 Uganda 2 0 1 1 0 1 5 Zambia 2 0 2 2 2 2 10 Zimbabwe 3 0 1 2 1 1 8
Score range is 0 (lowest) to 3 (highest). aThe grading is based on the date of publication or official launch of these documents. doi:10.1371/journal.pmed.1001133.t004
37 Lessons from Male Circumcision Program Scale-Up
Figure 3. Association of scale-up element scores and Diffusion of Innovation adoption status. The correlation between each of the six elements of scale-up and DOI adoption status is shown in this figure. The scores obtained (ranging from 0 to 3) for each element by each country is shown (on the vertical axis) in relation to the adoption status (shown on the horizontal axis). Having a pilot program appears to be the strongest predictor of adoption status; this can be seen in the linear clustering of the countries. Conversely, having conducted a situational analysis appears to the least predictive of adoption status; the clustering of the countries is less linear and appears more random. BOT, Botswana; KEN, Kenya; LES, Lesotho; MAL, Malawi; MOZ, Mozambique; NAM, Namibia; RWA, Rwanda; SAF, South Africa; SWA, Swaziland; TAN, Tanzania; UGA, Uganda; ZAM, Zambia; ZIM, Zimbabwe. doi:10.1371/journal.pmed.1001133.g003
38 Lessons from Male Circumcision Program Scale-Up with government involvement. These are important elements in challenges of the Kenya program, including human resource confirming country ownership of the program. To create an constraints, inadequate infrastructure, and shortages of equipment environment for increasing country ownership and country-to- and supplies, as well as difficulties with data management [40]. country learning, preparatory multi-stakeholder meetings were Despite challenges, the innovative responses that the Kenya held in Kenya, Lesotho, Swaziland, Tanzania, and Zambia in Ministry of Health implemented in collaboration with other 2006 [31–36]. Apart from Lesotho, all the other countries that partners facilitated the translation of the national policies and held early stakeholder consultations are classified within the early programs into service delivery. Program managers facing majority, possibly also highlighting the significant role of the shortages in human resources and inadequate infrastructure would environment. not, without strong political support, be able to address these The DOI and the ExpandNet frameworks both postulate that implementation challenges to the scale-up of an innovation. the sociopolitical context and cultural relevance of an innovation While this analysis has not focused on the differences between are also critical factors influencing the widespread adoption of an country strategies for VMMC scale-up, the choice of strategy innovation. The DOI theory also emphasizes that opinion leaders undoubtedly had an impact on VMMC program implementation directly affect the adoption of an innovation; this explains why and subsequent scale-up. For example, the scale-up strategy of some innovations are quickly adopted in one setting but fail to take Botswana has differed from that of other countries that off in others. It is therefore interesting to note that from our acknowledge the need for a phase of vertical programs—the analysis, the role of having a national champion engaged early on ‘‘catch-up’’ phase—to rapidly expand access to safe VMMC in the process does not appear to be a predictor of adoption status. services in addition to a strategy for integration [19]. The However, despite the low association scores that having a national Botswana strategy has focused on the integration of VMMC champion was shown to have in this analysis, we also know from within existing health services; this is perhaps a reflection of the our program implementation experience that former Botswana country’s experience with scaling up ART. This may explain why President Festus Mogae and Kenya’s Prime Minister Raila Odinga Botswana has not performed as many VMMCs as the early were prominent leaders that championed VMMC programs in adopter countries even though it had a VMMC focal person, their respective countries [37]. In addition, in 2009, South Africa policy, and strategy early on in the process. The timing of when acquired a government that prioritized the strengthening of HIV countries conducted a situation analysis does not appear to have a programs, and on World AIDS Day in December 2009, the new bearing on the process of adoption of VMMC programs. president of South Africa publicly announced the government’s However, for this analysis we did not take into account the type determination to move the AIDS agenda forward, saying, ‘‘Let the of situation analysis that was conducted or the process for politicization and endless debates about HIV and AIDS stop’’ and, dissemination and utilization of the situation analysis findings. regarding the need to struggle against AIDS as they had done with This analysis is not intended to criticize progress in any apartheid, ‘‘We have no choice but to deploy every effort, mobilize particular country; rather, it is an attempt to identify elements every resource, and utilize every skill our nation possesses.’’ critical to success and underscore some of the challenges to scale- [38,39] This commitment at the highest level undoubtedly up. While medical male circumcision has many of the ‘‘COR- contributed to the significant number of VMMCs (131,117) RECT’’ characteristics, it is difficult to scale up the intervention, performed in South Africa in 2010. Botswana had strong political particularly as a result of human resources limitations in terms of support from their former President Festus Mogae; however, his both quantity and quality [42]. Also challenging are logistics and presidency ended at the end of March 2008. This may explain supply management: successful scale-up will in part depend on the why Bostwana has reached only 3.2% of its 2015 DMPPT target definition and accessibility of commodities essential for VMMC despite strong momentum in the initial phases. programming and the appropriate allocation of resources to In Kenya, consistent political support and ongoing community support commodity procurement and supply chain logistics [43]. consultation have allowed implementation challenges to be This study suggests that the adoption of a health services addressed as they arise [40]. Kenya established a national VMMC innovation—the development of national policies and strategies, task force that engaged all the key partners and was replicated at and the initiation of pilot programs—and program scale-up are the provincial level. In addition, in 2007, Kenya initiated distinct processes. This is clearly illustrated by the limited progress community mobilization activities that engaged community made towards scale-up by the early adopter countries. Thus, leaders and other key stakeholders in VMMC program imple- although the DOI theory was useful for predicting the adoption of mentation and generated the demand for services. These activities VMMC as an innovation for HIV prevention, it was less useful as helped to overcome some of the initial political and cultural a model to describe what is actually needed to scale up VMMC tensions and to accelerate service delivery [41] This further services. The ExpandNet framework helps to provide some highlights the importance of the sociopolitical context and insights into the critical components required for scale-up. environment as well as sustained leadership at all levels (described in the ExpandNet framework) for moving ‘programs from initial Limitations adoption to scale-up. As suggested by the results of this study, it The VMMC scale-up process has not been well documented in might therefore be that the consistency of political support and countries and therefore data were obtained from limited sources. leadership, at all levels, is more important than just the initial The total scale-up score has not been used or validated in other engagement. This is due to the fact that while adoption is a one- programs and is reliant on assessments by a number of key time event, the scaling up of an innovation such as VMMC is an informants who have been closely involved in VMMC scale-up at iterative process carried out over an extended period of time that global and national levels. These individuals were interviewed and requires continued political support and, in many instances, the the information triangulated in an attempt to limit the potential input of decision makers to resolve implementation challenges as bias. We limited our predictors of adoption status to the elements they arise. This ongoing need for political support and the defined in the WHO/UNAIDS operational guidance, and yet involvement of decision makers is highlighted in the Kenya there may be other important program elements that are not well experience. An evaluation of the first year of VMMC program captured in the guidance. We have attempted to draw out some of implementation in Kenya provides insights into some of the these other factors in the discussion but recognize that there may
39 Lessons from Male Circumcision Program Scale-Up
Box 1. ExpandNet VMMC Components User organization: the organizations and programs policy and strategy development. adopting the innovation N The recommendations address the sociocultural, gender, N National Ministries of Health in all 13 priority countries and human rights issues that countries need to consider as have taken ownership and are leading the roll out in most well as health service issues, and therefore provide a countries. comprehensive framework for policy development. N All countries have coordination structures (VMMC task N The UN provided tools and guidance for scale-up, forces) that are functioning to varying degrees. These task including operations guidance, legal regulatory tool, forces are partnerships between the Ministries of Health clinical manual for practice, and training, quality assurance, and the implementing partners. Kenya and South Africa monitoring, and evaluation. also have provincial task forces. N Funding from donor organizations was made available and N Human resource constraints—lack of personnel at national coordinated to enhance capacity of Ministries of Health and facility level. and provider organizations in country. N Partners in countries available to provide technical support for scale-up. Environment: the social, cultural, political, and economic context within which scaling up takes place N Coordinated international leadership and advocacy sup- ported country action. Notable political champions in Botswana and Kenya. N Strong partnerships between governments and non- Political changes in Botswana and South Africa affected N governmental organizations have facilitated program initial trajectories both negatively and positively. scale-up in Kenya and, recently, Tanzania. N Gaining political support – it has been a process to get political buy-in in some countries N Global and national advocacy has moved some ‘‘early Scaling up strategy: the means by which the adopter’’ countries; peer pressure is working to bring the innovation is communicated, disseminated, ‘‘late majority’’ on board. Preparatory stakeholders meet- transferred, or otherwise promoted ings were held in 2006 in five countries (Lesotho, Kenya, N Policy development was diverse across countries with Tanzania, Swaziland, and Zambia) before the release of the differing types of policy instruments, e.g., Botswana has no WHO/UNAIDS recommendations. Regional consultations separate policy but strategy with policy elements; Zambia were held and national stakeholders meetings held in all sent Information note to Cabinet; Kenya developed policy countries after the release of the recommendations. guidelines; dedicated policies were developed in Lesotho, N Cultural context: issue raised of conflation with female Namibia, South Africa, Swaziland, Uganda, and Zimbabwe. genital mutilation (which occurs in some parts of Kenya N Country strategies developed that include objectives, and Uganda). In Lesotho, Malawi, and South Africa, cultural target population, numbers of men to be reached, costs, issues have been a challenge—the role of traditional service delivery strategies, resource mobilization, monitor- providers has caused much debate and tensions. Lack of ing, and evaluation. traditional male circumcision in Swaziland facilitated N DMPPT used to estimate cost, impact, pace of scale-up, adoption. and to develop or revise strategies. N Legal issues: few countries have laws governing practice of N Different scale-up strategies have influenced program traditional circumcisers. Task shifting, while successful in implementation. Most countries have ‘‘catch-up’’ strategies Kenya, is a challenge in other countries. to reach adult men—Kenya, Swaziland, Zimbabwe, Zam- bia—however, implementation varies; Kenya has gone Resource team: those involved in the development ahead with focused campaigns to achieve numbers, while and testing of the innovation and/or seeking to Botswana is focusing on integrated service delivery. promote its wider use N Demand creation—matching services to demand is difficult. N Developing countries look to WHO to provide norms and standards, therefore the timely release of the WHO/ N Communicating partial protection and risk compensation UNAIDS recommendations provided guidance for national are challenging.
be other elements of program adoption that are not easily circumcision sites have been set up by PEPFAR implementing quantifiable. partners working closely with government. Finally, the quantita- Although the scores have captured the time when the different tive analyses included in this study were based on a total sample of milestones were recorded and published, these only indirectly 13 countries. As such, although the results are indicative of reflect the timing and pace of scale-up. The date of publication experiences of these specific countries, the ability to generalize may have been several months after the work was initiated and/or these findings to other contexts is limited. completed, so the impact of the element may have been realized. The number of male circumcisions performed in priority countries Conclusion may be underestimated as the data largely reflects male Three years after the WHO/UNAIDS recommendations to circumcisions done through programs funded by the United expand, promote, and integrate VMMC into comprehensive HIV States President’s Emergency Plan for AIDS Relief (PEPFAR). prevention packages, VMMC has been adopted as a national HIV However, in many of the priority countries, initial male prevention strategy and implementation has been initiated in all of
40 Lessons from Male Circumcision Program Scale-Up the priority countries. Policies, national scale-up strategies, and pilot Acknowledgments projects have been put in place, and by the end of 2010, approximately 559,528 VMMCs had been performed in the priority We acknowledge the contributions of male circumcision focal persons in countries, and yet this represents only about 3% of what is needed to each of the priority countries. We also acknowledge Jason Reed for his contribution to the writing and review of versions of this paper. achieve country-derived targets. The variability in progress in scale- up of male circumcision is evident, and the two diffusion frameworks suggest that the adoption of VMMC as a HIV prevention innovation Author Contributions does not guarantee scale-up. A key lesson is the importance of not Conceived and designed the experiments: KD NT JS. Analyzed the data: only being ready to adopt a new intervention but also ensuring that KD NT JS EN. Wrote the first draft of the manuscript: NT KD. those factors that accelerate and sustain program implementation are Contributed to the writing of the manuscript: KD NT JS EN PC BD TF built and maintained. The most successful national program CR CH. ICMJE criteria for authorship read and met: KD NT JS EN PC exhibited country ownership and sustained leadership at all levels, BD TF CR CH. Agree with manuscript results and conclusions: KD NT JS EN PC BD TF CR CH. in addition to the adoption of a national policy and strategy to translate the research into a viable program.
References 1. Joint United Nations Programme on HIV/AIDS (2010) Global report: 20. World Health Organization/Joint United Nations Programme on HIV/AIDS UNAIDS report on global AIDS epidemic 2010. Geneva: Joint United Nations (2008) Operational guidance for scaling up male circumcision services for HIV Programme on HIV/AIDS. prevention. Geneva: World Health Organization. 2. Weller S, Davis-Beaty K (2002) Condom effectiveness in reducing he- 21. UNAIDS/WHO/SACEMA Expert Group on Modelling the Impact and Cost terosexual HIV transmission. Cochrane Database Syst Rev 2002. pp of Male Circumcision for HIV Prevention (2009) Male circumcision for HIV CD003255. prevention in high HIV prevalence settings: what mathematical models 3. DeCock K, Fowler M, Mercier E, deVincenzi I, Saba J, et al. (2000) Prevention contribute to informed decision making. PLoS Med 9: e:1000109. of mother-to-child HIV transmission in resource-poor countries: translating doi:10.1371/journal.pmed.1000109. research into policy practice. JAMA 9: 1175–1182. 22. Njeuhmeli E, Forsythe S, Reed J, Opuni M, Bollinger L, et al. (2011) Voluntary 4. World Health Organization (2010) Antiretroviral drugs for treating medical male circumcision: modeling the impact and cost of expanding male pregnant women and preventing HIV infection in infants: recommendations circumcision for HIV prevention in eastern and southern Africa. PLoS Med 8: for a public health approach—2010 version. Geneva: World Health e1001132. doi:10.1371/journal.pmed.1001132. Organization. 23. Glaser E, Abelson H, Garrison K (1983) Putting knowledge to use: facilitating 5. World Health Organization (2009) WHO, UNODC, UNAIDS technical guide the diffusion of knowledge and the implementation of planned change. San for countries to set targets for universal access to HIV prevention, treatment and Francisco: Jossey-Bass. care for injecting drug users. Geneva: World Health Organization. 24. Rogers E (1995) Diffusion of innovations, 4th ed. New York: Free Press. 6. Stover K (2011) Treating HIV-infected people with antiretrovirals significantly 25. Simmons R, Fajans P, Ghiron L (2007) Scaling up health service delivery: from reduces transmission to partners: findings result from NIH-funded international pilot innovations to policies and programmes. Geneva: World Health study. NIH News. Bethesda (Maryland): United States National Institutes of Organization. Health May 12. 26. Bollinger L, Stover J (2009) The potential cost and impact of expanding male 7. Peters A, Jansen W, van Driel F (2010) The female condom: the international circumcision in 14 countries. Geneva: World Health Organization. denial of a strong potential. Reprod Health Matters 18: 119–128. 27. World Health Organization (2009) Male circumcision: situation analysis toolkit. 8. Auvert B, Taljaard D, Lagarde E (2005) Randomized controlled intervention Geneva: World Health Organization. trial of male circumcision for reduction of HIV infection risk: The ANRS 1265 28. Balas E, Boren S (2000) Managing clinical knowledge for health care trial. PLoS Med 2: e298. doi:10.1371/journal.pmed.0020298. improvement. In: Bemmel J, McCray A, eds. Yearbook of medical informatics. 9. Gray RH, Kogozi G, Serwadda D, Makumbi F, Watya S, et al. (2007) Male Stuttgart: Schattauer Publishing. pp 65–70. circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial. 29. World Health Organization (2009) Practical guidance for scaling up health Lancet 369: 657–666. service innovations. Geneva: World Health Organization. 10. Bailey R, Moses S, Parker C (2007) Male circumcision for HIV prevention in 30. Halperin DT, Bailey RC (1999) Male circumcision and HIV infection: 10 years young men in Kisumu, Kenya: a randomised controlled trial. Lancet. pp and counting. Lancet 354: 1813–1815. 643–656. 31. Muula AS (2006) On evidence in support of male circumcision in HIV 11. Weiss H, Quigley M, Hayes R (2000) Male circumcision and risk of HIV prevention: what next? PLoS Med 3: e66. doi:10.1371/journal.pmed.0030066. infection in sub-Saharan Africa: a systematic review and meta-analysis. AIDS 32. Joint United Nations Programme on HIV/AIDS (2006) Report on the male 14: 2361–2370. circumcision and HIV prevention country consultation meeting. Esibayeni 12. Tobian A, Serwadda D, Quinn T, Kigozi G, Gravitt P, et al. (2009) Male Lodge, Matsapha. 26-27 September 2006 Geneva: Joint United Nations circumcision for the prevention of HSV-2 and HPV infections and syphilis. New Programme on HIV/AIDS. Engl J Med 13: 1298–1309. 33. Joint United Nations Programme on HIV/AIDS (2006) Male circumcision and 13. Auvert B, Sobngwi-Tambekou J, Cutler E, Nieuwoudt M, Lissouba P, et al. HIV prevention: Tanzania country stakeholder consultation. Dar es Salaam (2009) Effect of male circumcision on the prevalence of high-risk human Protea Hotel. 14-15 September 2006 Geneva: Joint United Nations Programme papillomavirus in young men: results of a randomized controlled trial conducted on HIV/AIDS. in Orange Farm, South Africa. J Infect Dis 199: 14–19. 34. Joint United Nations Programme on HIV/AIDS (2006) Male circumcision 14. Gray R, Serwadda D, Kong X, Makumbi F, Kigozi G, et al. (2010) Male consultative meeting report. Protea in Lusaka, Zambia. 11-12 September 2006 circumcision decreases acquisition and increases clearance of high-risk human Geneva: Joint United Nations Programme on HIV/AIDS. papillomavirus in HIV-negative men: a randomized trial in Rakai, Uganda. 35. Joint United Nations Programme on HIV/AIDS (2006) Kenya stakeholder J Infect Dis 201: 1455–1462. consultation on male circumcision in the context of HIV prevention. Mayfair 15. Serwadda D, Wawer M, Makumbi F, Kong X, Kigozi G, et al. (2010) Holiday Inn, Nairobi, Kenya. 7-8 September 2006 Geneva: Joint United Circumcision of HIV-infected men: effects on high-risk human papillomavirus Nations Programme on HIV/AIDS. infections in a randomized trial in Rakai, Uganda. J Infect Dis 201: 1463–1469. 36. Joint United Nations Programme on HIV/AIDS (2006) Male circumcision and 16. Wawer M, Tobian A, Kigozi G, Kong X, Gravitt P, et al. (2011) Effect of HIV prevention: Lesotho country consultation meeting report. Lesotho Sun Hotel, circumcision of HIV-negative women: a randomised trial in Rakai, Uganda. Lesotho. 25 July 2006 Geneva: Joint United Nations Programme on HIV/AIDS. Lancet 377: 209–218. 37. Dickson K (2010) Male circumcision for HIV prevention: progress in scale-up 17. Wawer M, Makumbi F, Kigozi G, Serwadda D, Watya S, et al. (2009) [presentation]. 17th Conference on Retroviruses and Opportunistic Infections; Circumcision in HIV-infected men and its effect on HIV transmission to female San Francisco, California; 16–19 February 2010. partners in Rakai, Uganda. Lancet 374: 229–237. 38. Dugger CW (2009) Breaking with past, South Africa issues broad AIDS policy. 18. Weiss H, Hankins C, Dickson K (2009) Male circumcision and risk of HIV The New York Times December 2: A6. infection in women: a systematic review and meta-analysis. Lancet Infect Dis 9: 39. Smith D (2009) South Africa’s Jacob Zuma promises to take HIV test in AIDS 669–677. policy reversal. The Guardian December 1: 16. 19. World Health Organization/Joint United Nations Programme on HIV/AIDS 40. National AIDS and STI Control Programme (2010) Progress report on Kenya’s (2007) WHO/UNAIDS technical consultation on male circumcision and HIV voluntary medical male circumcision programme, 2008-09: summary. Nairobi: prevention: research implications for policy and programming. Conclusions and Government of Kenya, Ministry of Public Health and Sanitation, National recommendations. Geneva: World Health Organization. AIDS and STI Control Programme.
41 Lessons from Male Circumcision Program Scale-Up
41. Herman-Roloff A, Llewellyn E, Obiero W, Agot K, Ndinya-Achola J, et al. needs of scale-up in southern and eastern Africa. PLoS Med 8: e1001129. (2011) Implementing voluntary medical male circumcision for HIV prevention doi:10.1371/journal.pmed.1001129. in Nyanza Province, Kenya: lessons learned during the first year. PLoS ONE 6: 43. Edgil D, Stankard PS, Forsythe S, Reed J, Rech D, et al. (2011) Voluntary e18299. doi:10.1371/journal.pone.0018299. medical male circumcision: logistics, commodities, and waste management 42. Curran K, Njeuhmeli E, Mirelman A, Dickson K, Adamu T, et al. (2011) requirements for scale-up of services. PLoS Med 8: e1001128. doi:10.1371/ Voluntary medical male circumcision: strategies for meeting the human resource journal.pmed.1001128.
42 Lessons from Male Circumcision Program Scale-Up
Editors’ Summary Background. Every year, more than 2.5 million people program scale-up soon after the release of the 2007 (mostly in sub-Saharan Africa) become infected with HIV, the recommendations and had started VMMC scale-up pilot virus that causes AIDS. There is no cure for HIV/AIDS and no programs in 2008 but were far from achieving their VMMC HIV vaccine. Consequently, global efforts to combat HIV/ targets. Having a VMMC focal person, establishing a national AIDS are concentrating on evidence-based prevention policy, having an operational strategy, and establishing a strategies such as voluntary medical male circumcision pilot program all predicted early adoption of VMMC scale-up. (VMMC). Circumcision—the removal of the foreskin, a loose fold of skin that covers the head of the penis—reduced HIV What Do These Findings Mean? These findings show transmission through sexual intercourse by 60% in men that, three years after the WHO/UNAIDS recommendation to in trials undertaken in sub-Saharan Africa, so in 2007, the integrate VMMC into comprehensive HIV prevention World Health Organization (WHO) and the Joint United programs, VMMC scale-up activities had been initiated in Nations Programme on HIV/AIDS (UNAIDS) recommended all the priority countries but that progress towards the 80% implementation of VMMC programs in countries with a coverage target was variable and generally poor. generalized HIV epidemic and low levels of male Importantly, they show that early adoption of VMMC as a circumcision. They also identified 13 countries in southern national program had not necessarily resulted in rapid and eastern Africa as high priority countries for rapid VMMC program scale-up. Although these findings may not be scale-up. Mathematical modeling suggests that 20.3 million generalizable to other settings, they suggest that countries circumcisions by 2015 and 8.4 million circumcisions between endeavoring to scale up VMMC (or other HIV prevention 2016 and 2025 are needed to reach 80% VMMC coverage in strategies) must not only be ready to adopt VMMC but must these countries. If this coverage is achieved, it will avert also ensure that all the factors critical to supporting and about 3.4 million new HIV infections through 2025. accelerating scale-up are incorporated into the scale-up program. Finally, these findings show that the most Why Was This Study Done? Despite convincing evidence successful national programs are those that involve that VMMC is an effective, cost-saving intervention in the country ownership of the program and that have sustained fight against HIV/AIDS, national VMMC scale-up programs in leadership at all levels to facilitate the translation of research the priority countries are currently at very different stages. A into national policies and programs. better understanding of the challenges faced by these programs would help countries still in the early stages of Additional Information. Please access these websites via VMMC scale-up implement their national programs and the online version of this summary at http://dx.doi.org/10. would facilitate implementation of other HIV prevention 1371/journal.pmed.1001133. strategies. In this study, the researchers use the Diffusion of Innovations (DOI) theory to analyze progress towards VMMC N This study is part of a PLoS Collection of articles on VMMC scale-up in the priority countries and to identify the factors (http://www.ploscollections.org/VMMC2011) and is further that may have expedited or hindered program scale-up. This discussed in a PLoS Medicine Review Article by Hankins theory seeks to explain how, why, and at what rate new et al. (http://dx.doi.org/10.1371/journal.pmed.1001127) ideas and technology spread through cultures. It posits that N Information is available from WHO, UNAIDS, and PEPFAR a few individuals (‘‘innovators’’) adopt new ideas before they on all aspects of HIV/AIDS become mainstream ideas. A few more individuals—the N NAM/aidsmap provides basic information about HIV/AIDS, ‘‘early adopters’’—follow the innovators. The ‘‘early majority’’ summaries of recent research findings on HIV care and is the next group to adopt the innovation, followed by the treatment, and information on male circumcision for the ‘‘late majority’’ and the ‘‘laggards.’’ prevention of HIV transmission What Did the Researchers Do and Find? The researchers N Information is available from Avert, an international AIDS used the annual number of VMMCs performed in the priority charity on many aspects of HIV/AIDS, including informa- countries since 2008 to classify the countries into DOI tion on aspects of HIV prevention, and on HIV/AIDS in adopter categories. They calculated a total scale-up score for Africa (in English and Spanish) each country based on six key elements of program scale-up N The Clearinghouse on Male Circumcision, a resource (such as whether and when a VMMC policy had been provided by WHO, UNAIDS, and other international bodies, approved). Finally, they analyzed the association between provides information and tools for VMMC policy develop- the DOI adopter category and the scores for the individual ment and program implementation scale-up elements to determine which elements predict N Wikipedia has a page on Diffusion of Innovations theory adoption and VMMC scale-up. By the end of 2010, about (note: Wikipedia is a free online encyclopedia that anyone 560,000 VMMCs had been completed, less than 3% of the can edit; available in several languages) target coverage for the priority countries. Kenya, the only N Personal stories about living with HIV/AIDS are available DOI innovator country, had completed nearly two-thirds of through Avert, through NAM/aidsmap, and through the the VMMCs needed to reach its target coverage and was the charity website Healthtalkonline only country on track to reach its target. The early adopters (South Africa, Zambia, and Swaziland) had initiated VMMC
43 44 Voluntary Medical Male Circumcision: A Cross-Sectional Study Comparing Circumcision Self-Report and Physical Examination Findings in Lesotho
Anne Goldzier Thomas1*, Bonnie Robin Tran1, Marcus Cranston2, Malerato Cecilia Brown3, Rajiv Kumar4, Matsotetsi Tlelai5 1 Department of Defense HIV/AIDS Prevention Program, Naval Health Research Center, San Diego, California, United States of America, 2 United States Air Force, Department of Defense, Keesler Air Force Base, Mississippi, United States of America, 3 United States President’s Emergency Plan for AIDS Relief—Lesotho, United States Embassy, Maseru, Lesotho, 4 Indian Armed Forces, Pune, India, 5 Lesotho Defence Force, Maseru, Lesotho
Abstract
Background: Overwhelming evidence, including three clinical trials, shows that male circumcision (MC) reduces the risk of HIV infection among men. However, data from recent Lesotho Demographic and Health Surveys do not demonstrate MC to be protective against HIV. These contradictory findings could partially be due to inaccurate self-reported MC status used to estimate MC prevalence. This study describes MC characteristics among men applying for Lesotho Defence Force recruitment and seeks to assess MC self-reported accuracy through comparison with physical-examination-based data.
Methods and Findings: During Lesotho Defence Force applicant screening in 2009, 241 (77%) of 312 men, aged 18–25 y, consented to a self-administered demographic and MC characteristic survey and physician-performed genital examination. The extent of foreskin removal was graded on a scale of 1 (no evidence of MC) to 4 (complete MC). MC was self-reported by 27% (n = 64/239) of participants. Of the 64 men self-reporting being circumcised, physical exam showed that 23% had no evidence of circumcision, 27% had partial circumcision, and 50% had complete circumcision. Of the MCs reportedly performed by a medical provider, 3% were Grade 1 and 73% were Grade 4. Of the MCs reportedly performed by traditional circumcisers, 41% were Grade 1, while 28% were Grade 4. Among participants self-reporting being circumcised, the odds of MC status misclassification were seven times higher among those reportedly circumcised by initiation school personnel (odds ratio = 7.22; 95% CI = 2.29–22.75).
Conclusions: Approximately 27% of participants self-reported being circumcised. However, only 50% of these men had complete MC as determined by a physical exam. Given this low MC self-report accuracy, countries scaling up voluntary medical MC (VMMC) should obtain physical-exam-based MC data to guide service delivery and cost estimates. HIV prevention messages promoting VMMC should provide comprehensive education regarding the definition of VMMC.
Citation: Thomas AG, Tran BR, Cranston M, Brown MC, Kumar R, et al. (2011) Voluntary Medical Male Circumcision: A Cross-Sectional Study Comparing Circumcision Self-Report and Physical Examination Findings in Lesotho. PLoS ONE 6(11): e27561. doi:10.1371/journal.pone.0027561 Editor: Stephanie L. Sansom, Centers for Disease Control and Prevention, United States of America Received June 24, 2011; Accepted October 19, 2011; Published November 29, 2011 This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication. Funding: Our program is supported by PEPFAR. DoD/DHAPP is a USG PEPFAR implementing agency. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The views expressed in this article are those of the authors and do not necessarily reflect the official policy or position of the Department of the Navy, Department of Defense, or the U.S. Government. Competing Interests: The authors have declared that no competing interests exist. We were unable to contact Rajiv Kumar, listed as an author of this paper, to confirm his involvement in the latest version. The corresponding author, Anne Goldzier Thomas, has therefore supplied the information regarding his contribution to the manuscript and his competing interests and it is correct to the best of her knowledge. The authors are not aware of any other potential conflict of interest associated with this manuscript. Abbreviations: LDF, Lesotho Defence Force; LDHS, Lesotho Demographic and Health Survey; MC, male circumcision; OR, odds ratio; VMMC, voluntary medical male circumcision; WHO, World Health Organization * E-mail: [email protected]
Introduction Many countries are planning or already delivering VMMC as a component of comprehensive HIV prevention services. However, With an estimated 2.6 million incident HIV infections in 2009 some countries may be hesitant to scale up VMMC because of and 33.3 million prevalent infections worldwide, the need for nationally representative survey results demonstrating higher HIV effective HIV prevention has never been more critical [1]. In prevalence among those who report being circumcised [9–11]. MC observational studies [2–4] and randomized clinical trials [5–7], impact modeling shows that reaching 80% MC coverage in a 5-y male circumcision (MC) has been shown to decrease HIV time frame will have the most substantial impact on HIV incidence. acquisition in men. Responding to these findings, the World Health These models rely on nationally representative MC prevalence data Organization (WHO) published recommendations supporting [12]. However, MC prevalence estimates available to planners and voluntary medical male circumcision (VMMC) for HIV prevention policy makers, such as those in the Demographic Health Survey or in countries with high HIV and low MC prevalence [8]. other nationally representative cross-sectional surveys, are based on
45 Lesotho Male Circumcision Prevalence self-reported MC status. Several studies comparing MC status LDF nurses and doctors were trained on the study protocol classification in diverse settings, as determined by self-reporting and using standardized materials, including how to correctly classify physical examination, have shown large reporting discrepancies the different grades of MC using the grading scale [22] (Table 1) between these different data collection modalities [13–15]. and a graphic assessment tool. LDF nurses briefed the applicants The Joint United Nations Programme on HIV/AIDS estimates about the study and conducted the informed consent process. that 24.5% of those aged 15–49 y living in the Kingdom of Study participants were not provided any HIV prevention Lesotho—which has a population of 1.9 million—are infected with education or information regarding MC prior to completing a HIV [1]. The 2004 Lesotho Demographic and Health Survey brief self-administered survey about their age, birthplace, marital (LDHS) found that 48% of men aged 15–59 y self-identified as status, education level, and religious affiliation. In addition, being circumcised. Contrary to most studies correlating lower HIV participants were asked about their circumcision status with the prevalence with higher prevalence of MC [16], the 2004 LDHS question ‘‘Are you circumcised?’’ The response choices were ‘‘yes’’ reported 23% HIV prevalence among men reporting MC and or ‘‘no.’’ If an individual answered ‘‘yes,’’ he was prompted on the 15% for those reporting no MC [9]. Similarly, the 2009 LDHS survey to provide age at circumcision, town/area of circumcision, reported 21% HIV prevalence among men reporting MC and district of circumcision, attendance at an initiation school (where a 16% among those reporting no MC [10]. These data, showing traditional circumciser would provide MC), the circumcision higher HIV prevalence in those self-reporting MC, are similar to setting (hospital/medical clinic, church or place of worship, findings from other countries, including Malawi [11]. These study initiation school, home, or other), and the circumcision provider results could be explained by misclassification of MC status due to (doctor or medical personnel, religious leader, initiation school self-report, or lack of adjustment for confounders such as MC personnel, family member/relative, or other). occurring after sexual debut [17–19], traditional MC practices During the physical examination, with a study nurse present, a that promote the spread of HIV, such as the reuse of unsterilized medical doctor examined the penis for extent of circumcision, MC cutting instruments on multiple males [20,21], or males using a four-point scale to classify the foreskin from completely having sex just after being circumcised, before the wound has uncircumcised (Grade 1) to completely circumcised (Grade 4) completely healed [17]. Thus, further studies exploring the (Table 1). After the physical examination, the physician placed concordance of self-reported circumcision status with physical study materials into a sealed envelope, which was then collected by examination are needed to better understand the factors associated the study coordinator. All study materials were stored securely and with discrepant reporting and the magnitude of inaccuracy. made available only to study personnel. This paper presents findings from a cross-sectional study Of the 241 consented participants, data on two were missing the estimating the prevalence of MC and exploring factors associated variable of interest, self-reported MC status, and were excluded with discrepant reports of MC status in a sample of young adult from the analysis. Descriptive statistics were computed for the men undergoing physical examination during the Lesotho Defence remaining 239 men. Frequency counts and percentages were Force (LDF) recruitment process. tabulated for categorical variables, and means and standard deviations were calculated for continuous variables. To assess Methods population sample representativeness, district comparisons of the From March to April 2009, all men undergoing physical participant population were made with the 2006 Lesotho census examination for voluntary recruitment into the LDF were invited and the 2004 LDHS, which used the 1996 Lesotho census for the to participate in the study. The applicants were aged 18 y or older population sampling frame. and came from all ten districts of Lesotho. Pearson’s chi-square tests (for categorical variables) were used to Human subjects participated in this study after giving their free determine any significant difference (p,0.05) in grade of MC by and informed consent. This research has been conducted in circumcision characteristics. Fisher’s exact tests were used when compliance with all applicable Federal Regulations governing the expected cell frequencies were less than five. A one-way analysis of Protection of Human Subjects in Research. Institutional review variance was used to determine significant differences in means boards in the United States (Naval Health Research Center, San between two or more groups. Diego, California) and Lesotho (Ministry of Health, Lesotho) Prevalence of MC by self-report was calculated by dividing the approved the study prior to data collection. Potential participants number of men who answered ‘‘yes’’ to the question whether they received written and verbal information about the purpose and were circumcised by the total number of men who answered the methods of the study and were given an opportunity to ask any question. Prevalence of complete MC as determined by physical questions about their possible participation. All men undergoing examination was calculated by dividing the number of men who entry physicals (n = 312) were invited to participate in the study, were classified by the physician as having a Grade 4 circumcision and 241 (77.2%) agreed to participate and provided written by the total number of men who underwent a physical informed consent. examination.
Table 1. Grading criteria for extent of male circumcision.
Grade Description
1 Foreskin covers one-half or more of the glans; completely uncircumcised 2 Foreskin is past the sulcus, but covers less than one-half of the glans 3 Foreskin is not past the sulcus, but can be extended past the sulcus to cover one-half of the glans without compressing the glans 4 Foreskin is completely absent; completely circumcised
doi:10.1371/journal.pone.0027561.t001
46 Lesotho Male Circumcision Prevalence
Simple logistic regression analysis was used to examine the Sixty-four participants (26.8%) reported that they were associations of demographic and MC characteristics with inaccu- circumcised (Table 3). The mean reported age at the time of rate reporting of MC status among participants who self-reported circumcision was 17.1 y, which did not differ significantly by grade being circumcised. Participants’ responses were classified as of MC (p = 0.50). In addition, there was no significant difference in ‘‘inaccurate’’ if they reported they were circumcised (answered mean age at MC among those circumcised in a traditional (15.9 y) ‘‘yes’’ to the question ‘‘Are you circumcised?’’) and were versus medical (17.7 y) setting (p = 0.14; data not shown). Among subsequently classified as Grade 1, 2, or 3 by a physician. those who reported being circumcised, nearly half of the Participants’ responses were classified as ‘‘accurate’’ if they circumcisions were performed by a medical professional (n = 30/ reported they were circumcised and were subsequently classified 61; 49.2%) and half (n = 29/61; 47.5%) by initiation school as Grade 4 by a physician. A sub-analysis was also conducted in personnel. Similarly, approximately half (n = 30/62; 48.4%) of the which participants with MC Grades 2 and 3 were excluded, as self-reported MCs were performed in a medical setting, and half some may argue that those with Grades 2 and 3 are circumcised, (n = 31/62; 50.0%) were performed in initiation school settings. as they do have evidence of some level of circumcision. Of those who reported being circumcised (n = 64), physical All statistical tests were two-tailed and were performed using examination revealed that 23.4% had no evidence of MC (Grade SAS statistical software, version 9.2 (SAS Institute). 1), 26.6% had evidence of partial MC (Grades 2 and 3), and 50.0% were classified as having complete foreskin removal (Grade Results 4). The grade of MC differed significantly by MC provider (p,0.001). Among the MCs reported to be performed by medical Demographic characteristics of study participants (n = 239) are personnel, 3.3% were classified as Grade 1, while 13.3% were presented in Table 2. The mean age was 21.5 y, with a range of classified as Grade 2, 10.0% were classified as Grade 3, and 73.3% 18–25 y. Most (88.2%) of the men were single; 9.6% reported were classified as Grade 4. The MCs reported to be performed by being married, and 2.2% were cohabiting. The majority (86.5%) initiation school personnel were classified as follows: Grade 1 of study participants had completed secondary/high school, as (41.4%), Grade 2 (24.1%), Grade 3 (6.9%), and Grade 4 (27.6%). compared with 28.0% of the 2004 LDHS male population. The association of grade of MC and the setting of the MC was also Religious affiliation was mixed and similar to the 2004 LDHS found to be statistically significant (p,0.001), with MCs performed sample. Participants were from all ten districts in the country, with at an initiation school more likely to be reported as Grade 1 or 2 a distribution representative of the 2006 Lesotho census and the than those performed in a medical setting (Table 3). The three 2004 LDHS (see Table S1). districts with the highest percentages of discrepant self-reported Of 239 participants, 175 (73.2%) reported that they were not MC status were Butha-Buthe (21.4%), Qacha’s Nek (14.3%), and circumcised (Table 3). Of these men, physical examination Quthing (14.3%) (see Table S2). revealed that 95.4% showed no evidence of circumcision, and Overall, the prevalence of MC as determined by self-report was 4.0% were partially circumcised (Grades 2 and 3) or had naturally 26.8% (n = 64/239). The prevalence of Grade 4 MC (complete shorter foreskins. Of this group that reported not being MC) as determined by physical examination was 13.8% (n = 33/ circumcised, one participant (0.6%) was classified as having a 239) (Table 3). The difference between the self-reported MC complete circumcision (Grade 4). prevalence and the physician-classified prevalence of Grade 4 MC is 13.0%. Results of simple logistic regression analysis examining factors Table 2. Demographic characteristics of study participants significantly associated with inaccurate MC reporting among (n = 239). participants who self-reported MC are presented in Table 4. Age, current marital status, education, and age at MC were not . Demographic Characteristic Value significantly associated with inaccurate reporting of MC (p 0.05). MC provider and setting were significantly associated with Age in years 21.561.4 inaccurate reporting of MC. The odds of inaccurate reporting of Current marital status MC were seven times higher among participants who reported Married 22 (9.6) that their MC was performed by initiation school personnel Living together 5 (2.2) than among those who reported that their MC was performed by a doctor or medical personnel (odds ratio [OR] = 7.22; 95% Not in union 202 (88.2) CI = 2.29–22.75). Similarly, the odds of inaccurate reporting of Education MC was almost six times higher among participants who reported Vocational/tech primary 4 (1.7) that their MC was conducted at an initiation school than among Secondary/high 205 (86.5) those who reported that their MC was conducted at a hospital or Vocational/tech secondary 14 (5.9) medical clinic (OR = 5.70; 95% CI = 1.90–17.14). In the sub- College 14 (5.9) analysis excluding participants with MC Grades 2 and 3, significant, elevated odds of inaccurate MC reporting were also Religion observed among those who reported their MC was performed by Roman Catholic 115 (48.3) initiation school personnel (OR = 33.0; 95% CI = 3.68–296.23) or Lesotho Evangelical 66 (27.7) conducted at an initiation school (OR = 32.7; 95% CI = 3.72– Anglican 23 (9.7) 287.21) (data not shown). Other Christian 29 (12.2) None/other 5 (2.1) Discussion
All values are n (percent), except for age, which is given as mean 6 standard The results from this cross-sectional study of MC among male deviation. LDF applicants provide further evidence that categorizing MC doi:10.1371/journal.pone.0027561.t002 status by self-report is highly prone to error. Upon physical
47 Lesotho Male Circumcision Prevalence
Table 3. Self-reported male circumcision characteristics by grade of circumcision as determined by physical examination.
b MC Characteristic Total, n (%)a Extent of Circumcision by Physical Exam, n (%) p-Value Grade 1 Grade 2 Grade 3 Grade 4
Self-reported circumcision status (n =239) ,0.001c Circumcised 64 (26.8) 15 (23.4) 11 (17.2) 6 (9.4) 32 (50.0) Not circumcised 175 (73.2) 167 (95.4) 5 (2.9) 2 (1.1) 1 (0.6) Age at MC in years, mean ± SD (n = 60)d 17.1 (4.5) 18.4 (2.6) 16.8 (3.5) 15 (5.2) 16.9 (5.3) 0.50e Provider of MC (n =61)d ,0.001c Doctor/medical personnel 30 (49.2) 1 (3.3) 4 (13.3) 3 (10.0) 22 (73.3) Initiation school personnel 29 (47.5) 12 (41.4) 7 (24.1) 2 (6.9) 8 (27.6) Religious leader 1 (1.6) 0 (0) 0 (0) 0 (0) 1 (100) Other 1 (1.6) 1 (100) 0 (0) 0 (0) 0 (0) Setting of MC (n =62)d ,0.001c Hospital/medical clinic 30 (48.4) 1 (3.3) 4 (13.3) 4 (13.3) 21 (70.0) Initiation school 31 (50.0) 14 (45.2) 6 (19.4) 2 (6.5) 9 (29.0) Home 1 (1.6) 0 (0) 0 (0) 0 (0) 1 (100)
Due to rounding, percentages may not add up to 100. aColumn percentages are presented. bRow percentages are presented. cResults of chi-square analysis. dMissing participant responses. eResults of one-way analysis of variance. SD, standard deviation. doi:10.1371/journal.pone.0027561.t003
examination, only 50% of participants self-reporting circumcision showed evidence of complete circumcision, and another 27% Table 4. Univariate associations of demographic and male showed evidence of only partial circumcision. The prevalence of circumcision characteristics with inaccurate self-reporting of self-reported MC (27%) in this study is also much lower than the MC among participants who self-reported ‘‘yes’’ to MC prevalence found in the 2004 and 2009 LDHS studies (48% and (n = 64). 52%, respectively). The discrepancies in self-reported MC status found in this study suggest that data from population-based surveys evaluating associations between MC and risk of HIV Characteristic Inaccurate Self-Reporting of MC infection should be interpreted with caution. The magnitude of OR (95% CI) p-Value inaccurate self-reported MC status could be sufficient to explain the apparent correlation of higher rates of HIV with higher rates Age in years 0.88 (0.60–1.28) 0.51 of MC, as seen in the Lesotho and Malawi Demographic Health Current marital status Surveys. However, other factors could also explain this observa- Not in union 1.0 0.78 tion, including lack of adjustment for confounders, such as MC Married 0.58 (0.13–2.67) occurring after sexual debut and traditional MC practices that Living together 0.96 (0.06–16.21) may promote the transmission of HIV. This study has some limitations. The LDF recruit applicant Education population was similar to the 2004 LDHS population in terms of Secondary/high 1.0 0.70 district of birth and religion. Thus, for MC practices and MC- Vocational/tech secondary 0.47 (0.08–2.75) related sexual practices related to local cultural or religious College 0.93 (0.06–15.6) affiliation, the LDF applicant study population is representative of Age at MC in years 1.02 (0.91–1.14) 0.77 the national population. But the LDF applicants were more Provider of MCa educated and of a more restricted age range than the general population. However, this does not appear to have biased the Doctor/medical personnel 1.0 0.001 study results, as the average MC prevalence in the LDHS is similar Initiation school personnel 7.22 (2.29–22.75) across all men aged 20 y and older (57.9%). However, higher Setting of MCb education may be associated with increased medical MC as Hospital/medical clinic 1.0 0.002 compared with traditional MC, because of an increased Initiation school 5.70 (1.90–17.14) appreciation for the medical risks of MC performed in traditional settings. Alternatively, increased educational attainment might be Results of simple logistic regression analysis. a marker of increased family income to pay for a medical aExcludes participants who reported religious leader (n = 1) or other (n = 1) as provider of MC. circumcision. It is unclear whether medical MC is ultimately more bExcludes participants who reported home as setting of MC (n = 1). expensive than traditional MC, since traditional MC may include doi:10.1371/journal.pone.0027561.t004 costs for hosting community celebrations as well as payment for
48 Lesotho Male Circumcision Prevalence the circumciser and initiation school [17]. Another limitation to The associations of demographic and MC characteristics with this study is that the physicians were not blinded to the inaccurate reporting of MC status in this study were also participants’ reported MC status, nor were the physicians’ examined. The only factors shown to be significantly associated assessments of the MC grades validated, which may have resulted with inaccurate reporting were having MC performed by initiation in misclassification. Although we cannot exclude the possibility of school personnel or conducted at an initiation school. This MC grade misclassification, the grading scheme was relatively association was significant even when those with Grade 2 and 3 simple and the rates of misclassification were most likely low. MCs were removed from the analysis. In many countries, Despite these limitations, this study has many strengths. Among including Lesotho, South Africa, Malawi, Namibia, Kenya, and them is the short time frame that was required for the study to be Uganda, non-medical MC is commonly performed during implemented; the data was collected in approximately four weeks attendance at a traditional initiation school [26], and undergoing during the LDF applicant examinations. This short time frame is MC in this setting is a rite of passage to manhood. Underscoring especially important as educational campaigns and news of MC as this point, the word in Sesotho for going through the initiation an HIV prevention modality become more common. Similarly, process, ‘‘lebollo,’’ is very similar to the word for circumcision. study personnel may have implemented the protocol with higher Thus, a male who has attended an initiation school may report conformity during this short time frame. Offering study partici- that he has been circumcised even if the foreskin was only cut, or pation to men already undergoing physical examination including only part of the foreskin was removed. These results provide genital exam may have reduced non-respondent bias. The compelling evidence that specific VMMC communication cam- regional and religious representativeness of the study population paigns must include factual information describing or graphically are other key strengths, as these demographic characteristics are representing the penile foreskin, so that VMMC is understood to often related to MC practices. The higher educational status of the mean the complete removal of the foreskin and is not conflated LDF applicants may also be a strength, as these individuals may be with traditional MC practices, such as those conducted at more likely to provide an accurate report of their circumcision initiation schools. status. As nations with high HIV prevalence begin to act on WHO Discrepancies in self-reported MC status have important recommendations for VMMC programs, the need for accurate implications for planning VMMC scale-up, communication, and MC prevalence data becomes even more critical. Thus, until education for HIV prevention, as reliance on self-reported MC further research can document improved methods for obtaining status may underestimate the volume of surgical intervention accurate self-reported MC data, all assessments of MC and HIV required and not accurately identify those individuals for which the prevalence, as well as projections for VMMC interventions, should intervention is indicated. While 26.8% of the participants self- be informed by physical-exam-based data. reported being circumcised, the prevalence of complete MC as estimated by physical examination was only 13.8%. This means that if self-reported data were used to estimate the need for MC in this Supporting Information population, the need would be underestimated by 13.0%. If these Table S1 Study population: district of residence findings are applied to Lesotho as a whole, the difference between (n = 239). the 52% self-reported MC prevalence found in the 2009 LDHS and (DOCX) the 14% MC prevalence found in this study population increases the national need for VMMC by 38%. The prevention effect of Table S2 District of male circumcision by grade of VMMC in reducing HIV incidence in Lesotho may also be vastly circumcision as determined by physical examination underestimated. Furthermore, these data show that while most of (n = 64). the discrepant results were among those who reported undergoing (DOCX) traditional MC, a substantial percentage of the physician-performed VMMCs were incomplete (n = 8/30; 26.6%), which suggests that Acknowledgments additional surgical training may be necessary to ensure adequate VMMC results to improve HIV prevention outcomes [23]. The authors wish to thank the male applicants to the LDF who participated in this study and all of the physicians and nurses who helped This study adds to the evidence that determining MC status facilitate data collection. We also thank the staff at PSI Lesotho for their through self-report is prone to result in misclassification [24,25]. Even help with training study staff and supporting the logistics of the study. In among participants who reported not being circumcised, about 5% addition, we would like to thank Naomi Bock, Emmanuel Njeuhmeli, Jason were partially or completely circumcised. The reasons for inaccurate Reed, and Carolyn Williams for their support in the design and self-report may be that (1) survey tools and methods do not currently implementation of the study. Finally, we thank Braden Hale, Stephanie allow for more than dichotomous (yes/no) categorization and do not Brodine, and Melanie Bacon for their thoughtful reviews and suggestions capture all aspects of MC (such as the level of foreskin removal or for the paper. nuances of traditional circumcision), (2) there is a misunderstanding of the meaning of medical circumcision as compared with traditional Author Contributions circumcision, or (3) there is a desire to maintain secrecy about Conceived and designed the experiments: AGT MC MCB MT. Performed initiation rites. Future studies seeking to improve MC self-report may the experiments: MCB RK MT. Analyzed the data: AGT BRT. benefit from the addition of partial MC categories, along with Contributed reagents/materials/analysis tools: AGT MC RBT. Wrote graphics depicting all four grades of male circumcision. the paper: AGT BRT MC MCB RK MT.
References 1. Joint United Nations Programme on HIV/AIDS (2010) Global report: 3. Fink AJ (1986) A possible explanation for heterosexual male infection with UNAIDS report on the global AIDS epidemic 2010. 10.11E | JC1958E AIDS. N Engl J Med 315: 1167. 10.11E | JC1958E Geneva: Joint United Nations Programme on HIV/ 4. Moses S, Bradley JE, Nagelkerke NJ, Ronald AR, Ndinya-Achola JO, et al. AIDS. (1990) Geographical patterns of male circumcision practices in Africa: 2. Alcena V (1986) AIDS in third world countries. N Y State J Med 86: 446. association with HIV seroprevalence. Int J Epidemiol 19: 693–697.
49 Lesotho Male Circumcision Prevalence
5. Auvert B, Taljaard D, Lagarde E, Sobngwi-Tambekou J, Sitta R, et al. (2005) 16. Weiss HA, Quigley MA, Hayes RJ (2000) Male circumcision and risk of HIV Randomized, controlled intervention trial of male circumcision for reduction of infection in sub-Saharan Africa: a systematic review and meta-analysis. AIDS HIV infection risk: the ANRS 1265 Trial. PLoS Med 2: e298. doi:10.1371/ 14: 2361–2370. journal.pmed.0020298. 17. Bailey RC, Egesah O, Rosenberg S (2008) Male circumcision for HIV 6. Bailey RC, Moses S, Parker CB, Agot K, Maclean I, et al. (2007) Male prevention: a prospective study of complications in clinical and traditional circumcision for HIV prevention in young men in Kisumu, Kenya: A settings in Bungoma, Kenya. Bull World Health Organ 86: 669–677. randomised controlled trial. Lancet 369: 643–656. 18. Kelly R, Kiwanuka N, Wawer MJ, Serwadda D, Sewankambo NK, et al. (1999) 7. Gray RH, Kigozi G, Serwadda D, Makumbi F, Watya S, et al. (2007) Male Age of male circumcision and risk of prevalent HIV infection in rural Uganda. circumcision for HIV prevention in men in Rakai, Uganda: A randomised trial. AIDS 13: 399–405. Lancet 369: 657–666. 19. Peltzer K, Kanta X (2009) Medical circumcision and manhood initiation rituals 8. World Health Organization, Joint United Nations Programme on HIV/AIDS in the Eastern Cape, South Africa: a post intervention evaluation. Cult Health (2007) WHO/UNAIDS technical consultation on male circumcision and HIV Sex 11: 83–97. prevention: research implications for policy and programming: new data on 20. Kepe T (2010) ‘Secrets’ that kill: crisis, custodianship and responsibility in ritual male circumcision and HIV prevention: policy and programme implications: male circumcision in the Eastern Cape Province, South Africa. Soc Science Med conclusions and recommendations. Geneva: World Health Organizations. 70: 729–735. 9. Lesotho Ministry of Health and Social Welfare, Lesotho Bureau of Statistics, 21. Brewer DD, Potterat JJ, Roberts JM, Jr., Brody S (2007) Male and female ORC Macro (2005) Lesotho Demographic and Health Survey 2004. Calverton circumcision associated with prevalent HIV infection in virgins and adolescents (Maryland): MOH, BOS, and ORC Macro. 10. Lesotho Ministry of Health and Social Welfare, ICF Macro (2010) Lesotho in Kenya, Lesotho, and Tanzania. Ann Epidemiol 17: 217–226. Demographic and Health Survey 2009. Calverton (Maryland): MOH and ICF 22. Wynder EL, Licklider SD (1960) The question of circumcision. Cancer 13: Macro. 442–445. 11. Malawi National Statistical Office, ORC Macro (2005) Malawi Demographic 23. Kigozi G, Wawer M, Ssettuba A, Kagaayi J, Nalugoda F, et al. (2009) Foreskin and Health Survey 2004. Calverton (Maryland): NSO and ORC Macro. surface area and HIV acquisition in Rakai, Uganda (size matters). AIDS 23: 12. Njeuhmeli E, Forsythe S, Reed J, Opuni M, Bollinger L, et al. (2011) Voluntary 2209–2213. medical male circumcision: modeling the impact and cost of expanding male 24. Lagarde E, Dirk T, Puren A, Reathe RT, Bertran A (2003) Acceptability of male circumcision for HIV prevention in eastern and southern Africa. PLoS Med 8: circumcision as a tool for preventing HIV infection in a highly infected e1001132. doi:10.1371/journal.pmed.1001132. community in South Africa. AIDS 17: 89–95. 13. Weiss HA, Plummer ML, Changalucha J, Mshana G, Shigongo ZS, et al. (2008) 25. Templeton DJ, Mao L, Prestage GP, Jin F, Kaldor JM, et al. (2008) Self-report is Circumcision among adolescent boys in rural northwestern Tanzania. Trop a valid measure of circumcision status in homosexual men. Sex Transm Infect Med Int Health 13: 1054–1061. 84: 187–188. 14. Urassa M, Todd J, Boerma JT, Hayes R, Isingo R (1997) Male circumcision and 26. Wilcken A, Keil T, Dick B (2010) Traditional male circumcision in eastern and susceptibility to HIV infection among men in Tanzania. AIDS 11: 73–80. southern Africa: a systematic review of prevalence and complications. Bull 15. Risser JM, Risser WL, Eissa MA, Cromwell PF, Barratt MS, et al. (2004) Self- World Health Organ 88: 907–914. assessment of circumcision status by adolescents. Am J Epidemiol 159: 1095–1097.
50 Voluntary Medical Male Circumcision: Logistics, Commodities, and Waste Management Requirements for Scale-Up of Services
Dianna Edgil1*, Petra Stankard2, Steven Forsythe3, Dino Rech4, Kristin Chrouser5, Tigistu Adamu5, Sameer Sakallah6, Anne Goldzier Thomas7, Jennifer Albertini8, David Stanton1, Kim Eva Dickson9, Emmanuel Njeuhmeli1 1 United States Agency for International Development, Washington, District of Columbia, United States of America, 2 Gillings School of Global Public Health, University of North Carolina, Chapel Hill, North Carolina, United States of America, 3 Futures Institute, Glastonbury, Connecticut, United States of America, 4 Centre for HIV and AIDS Prevention Studies, Johannesburg, South Africa, 5 Jhpiego, Baltimore, Maryland, United States of America, 6 Supply Chain Management Systems, Arlington, Virginia, United States of America, 7 Department of Defense HIV/AIDS Prevention Program, Naval Health Research Center, San Diego, California, United States of America, 8 United States Agency for International Development, Mbabane, Swaziland, 9 World Health Organization, Geneva, Switzerland
Abstract
Background: The global HIV prevention community is implementing voluntary medical male circumcision (VMMC) programs across eastern and southern Africa, with a goal of reaching 80% coverage in adult males by 2015. Successful implementation will depend on the accessibility of commodities essential for VMMC programming and the appropriate allocation of resources to support the VMMC supply chain. For this, the United States President’s Emergency Plan for AIDS Relief, in collaboration with the World Health Organization and the Joint United Nations Programme on HIV/AIDS, has developed a standard list of commodities for VMMC programs.
Methods and Findings: This list of commodities was used to inform program planning for a 1-y program to circumcise 152,000 adult men in Swaziland. During this process, additional key commodities were identified, expanding the standard list to include commodities for waste management, HIV counseling and testing, and the treatment of sexually transmitted infections. The approximate costs for the procurement of commodities, management of a supply chain, and waste disposal, were determined for the VMMC program in Swaziland using current market prices of goods and services. Previous costing studies of VMMC programs did not capture supply chain costs, nor the full range of commodities needed for VMMC program implementation or waste management. Our calculations indicate that depending upon the volume of services provided, supply chain and waste management, including commodities and associated labor, contribute between US$58.92 and US$73.57 to the cost of performing one adult male circumcision in Swaziland.
Conclusions: Experience with the VMMC program in Swaziland indicates that supply chain and waste management add approximately US$60 per circumcision, nearly doubling the total per procedure cost estimated previously; these additional costs are used to inform the estimate of per procedure costs modeled by Njeuhmeli et al. in ‘‘Voluntary Medical Male Circumcision: Modeling the Impact and Cost of Expanding Male Circumcision for HIV Prevention in Eastern and Southern Africa.’’ Program planners and policy makers should consider the significant contribution of supply chain and waste management to VMMC program costs as they determine future resource needs for VMMC programs. Please see later in the article for the Editors’ Summary.
Citation: Edgil D, Stankard P, Forsythe S, Rech D, Chrouser K, et al. (2011) Voluntary Medical Male Circumcision: Logistics, Commodities, and Waste Management Requirements for Scale-Up of Services. PLoS Med 8(11): e1001128. doi:10.1371/journal.pmed.1001128 Academic Editor: Stephanie L. Sansom, Centers for Disease Control and Prevention, United States of America Received May 26, 2011; Accepted October 19, 2011; Published November 29, 2011 This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication. Funding: Funding for the execution of this costing analysis was provided by PEPFAR through The Partnership for Supply Chain Management Project (SCMS) and Health Policy Initiative Costing Task Order (HPI CO TO) both funded by the United States Agency for International Development (USAID). Technical staff from USAID, US Centers for Disease Control and Prevention, and US Department of Defense was involved in study design, data collection and analysis, decision to publish, and preparation of the manuscript. Competing Interests: During the writing of this paper, PS had a short-term (4 month) consulting contract with PSI, an international non-governmental organization implementing medical male circumcision programs. PS’s consulting scope of work focused on tuberculosis programming and did not include any activities associated with their medical male circumcision programs. PS did not directly report to any individual working with these circumcision programs. All other authors have declared that no competing interests exist. Abbreviations: MOVE, the World Health Organization report ‘‘Considerations for Implementing Models for Optimizing the Volume and Efficiency for Male Circumcision Services’’; PEPFAR, United States President’s Emergency Plan for AIDS Relief; PFSCM, Partnership for Supply Chain Management; STI, sexually transmitted infection; UNAIDS, Joint United Nations Programme on HIV/AIDS; VMMC, voluntary medical male circumcision; WHO, World Health Organization * E-mail: [email protected]
51 Supply Chain and Logistics for Male Circumcision
Introduction prevention, emergency care, and general surgical theater equip- ment) [13]. The meeting did not address supplies for HIV testing, Based on evidence from randomized controlled clinical trials STI screening or treatment, or equipment for waste management. conducted in Africa indicating that voluntary medical male In 2010, the WHO’s release of ‘‘Considerations for Implementing circumcision (VMMC) significantly reduces male participants’ risk Models for Optimizing the Volume and Efficiency for Male of acquiring HIV infection, the global HIV prevention community Circumcision Services’’ (MOVE) clarified the need for standard- has supported the scale-up of VMMC programs in 14 countries [1– ized commodities and bundled supplies and equipment to achieve 3]. Evidence suggests that reaching a goal of 80% VMMC coverage efficiency [14]. in 5 y and sustaining it thereafter would avert more than 3.6 million Given the limitations of previous costing exercises for VMMC adult HIV infections in the next 15 y and benefit as many as 20.3 programs, the purpose of this paper is to present, to our knowledge million adult HIV-negative men for HIV prevention purposes [4]. for the first time, the full supply chain costs associated with the Reaching this population with comprehensive VMMC services will scale-up of a VMMC program in one southern African country, have substantial implications for the coordination of human Swaziland. This information will also improve the standard list resources [5], commodities, and infrastructure. of commodities outlined by PEPFAR and WHO by adding Supply chain considerations and commodity costs carry signif- commodities for waste management, counseling and testing, and icant financial implications for all public health programs, but these temporary infrastructure for VMMC programming. costs are particularly relevant for highly technical programs planned To inform this analysis, we collected costing information for on the scale anticipated for VMMC services. The minimum supply chain and waste management activities and commodities package of services that should compose all VMMC for HIV associated with the planned implementation of Swaziland’s 1-y prevention programs—as delineated by the World Health Organi- VMMC scale-up campaign—known as the ASI [15]. This zation (WHO)—encompasses an array of activities beyond information was used to calculate the total cost per VMMC for circumcision surgery, including (1) HIV testing and counseling, (2) the Swazi program with either a planned target of 152,000 VMMCs sexually transmitted infection (STI) screening and treatment, (3) or a reduced target of 75,000 VMMCs. In our analysis of costs condom provision and promotion, and (4) counseling on risk required for the ASI, we outline key commodity and supply chain reduction and safer sex [6]. Thus, program needs are broad, and inputs and offer future considerations that may serve to inform resources must support all major components of a logistics system VMMC program planning and policy development in coming years. involving human capital and/or material inputs, procurement of commodities, freight management, warehousing, in-country distri- Swaziland VMMC Program Service Provision Model bution, and waste management. The Swazi campaign goal is to increase VMMC coverage for As a result, costing studies have been undertaken in Namibia HIV prevention purposes from 8% to 80% among 15- to 49-y-old [7], Kenya [8], South Africa [9], Uganda [10], Zambia [11], and males by the end of 2011 to avert nearly 100,000 HIV infections; Zimbabwe [12] to help governments fully appreciate the initial reaching this target equates to circumcising approximately financial investments required to launch VMMC initiatives. 152,000 men [15]. In line with WHO MOVE recommendations To date, these studies have estimated the cost of a single male for efficient service delivery, the Swazi service provision strategy circumcision to be, on average, approximately US$53, ranging includes the use of prepackaged VMMC kits with disposable from US$19 in Uganda [10] to US$73 in Zambia [11]. However, surgical instruments [14]. Optimum implementation of the service until this study, no quantification of VMMC program costs, to our delivery model chosen by Swaziland enables a service delivery knowledge, has included the full range of supply chain and team—composed of a surgeon assisted by five nurses and four commodity considerations. For example, the supply chain costing counselors—to deliver services to an average of 40 clients per day, data included by Njeuhmeli et al. [4] uses costing data collected with a maximum throughput of 80 clients per day. in Zimbabwe in 2009 that estimated the per circumcision supply Consistent with MOVE recommendations and to meet the chain costs at between US$59.71 and US$71.59 ([12]; Supply target of 152,000 VMMCs, it was determined that services would Chain Management Systems–Zimbabwe, personal communi- be delivered by a varying number of teams per month (to account cation). While the Zimbabwe exercise went further than its for demand seasonality and scale-up and scale-down periods), with predecessors in looking at the cost of procurement and distribution up to 35 teams operating simultaneously at 27 sites during peak of commodities, it failed to account for all necessary supply chain months in 2011 [14]. Seven of these sites (comprising 14 teams) are costs associated with VMMC programs, including warehousing considered fixed and are integrated into existing health facilities. and waste management. Gaps in data present difficulties for both The remaining sites, all but one of which comprise one team each policy makers and implementers as they plan for the rapid scale-up (the exception has two teams), are considered mobile, and are of adult VMMC. contained within tented facilities, which, for the purpose of The Partnership for Supply Chain Management (PFSCM), meeting unexpected demand, can be moved to a different location. supported by the United States President’s Emergency Plan for At the midpoint of the VMMC campaign in Swaziland, the rate AIDS Relief (PEPFAR), began assisting countries that were scaling of scale-up is less than would be expected to allow the program to up VMMC programs to develop, strengthen, and manage reliable, reach its target of 152,000 VMMCs. To reflect a lower rate of cost-effective, and sustainable supply chains as early as 2009. service delivery, with a target of reaching 75,000 men over 1 y, the Further, in an attempt to harness economies of scale and stream- number of operational sites was reduced to 20, with a single team line logistical issues, PEPFAR hosted an expert consultation in operating at each site. 2009 on VMMC supply chain issues [13]. The PFSCM provides technical and programmatic support on As a result of that meeting, a standard list of VMMC procurement, logistics, and waste management to the Government commodities was put forward by international stakeholders. The of the Kingdom of Swaziland in its effort to implement the ASI bundling of commodities into surgical instrument kits was program. In support of the Swaziland campaign, and with input recommended (depending upon surgical method and disposa- from other stakeholders, PFSCM developed a full list of commodities bility of items), as well as standardized equipment (for infection for the ASI campaign. This included the four modules already developed
52 Supply Chain and Logistics for Male Circumcision by PEPFAR and WHO, as well as commodities modules for (1) HIV circumcisions achieved. This process assumes that the utilization of counseling and testing and STI treatment and (2) temporary infrastructure the VMMC facilities is at a sufficient volume of clients to allow for necessary for the delivery of mobileVMMCservices(fullcontentsof 152,000 circumcisions to be performed over the course of 12 mo. modules and line item costs are listed in Table S1). Furthermore, the Per VMMC costs were also re-analyzed using a reduced target of module on infection prevention was expanded to include all commodities 75,000 VMMCs to reflect the current rate of approximately 6,500 necessary for waste management for VMMC programs. These modules VMMCs per month. were added to the four modules already developed by PEPFAR [13] and WHO [14]. Variable Commodity Costs In addition to assisting with the development of the complete list of We calculated consumables costs based on high-volume price commodities, PFSCM was asked to provide technical assistance on quotes secured by PFSCM. The calculated cost of freight for all the development of a supply chain to deliver necessary commodities commodities is 6% of the total cost. In addition to the cost of for the ASI campaign. The structure of the supply chain for this consumables, we calculated a 5% surcharge for procurement program included quantification, forecasting, procurement, freight services provided by the Arlington-based staff of PFSCM. This forwarding, customs negotiation, warehousing, implementation of a cost covers the effort required to source suppliers, compile tender supply chain management information system for stock keeping and documents, negotiate with suppliers, and assess the quality of ordering, distribution of commodities when needed to VMMC sites, proposals and actual products. and monitoring of supplies at the site level. Three staff at the organization’s headquarters in Arlington, Virginia, performed the Fixed Commodity Costs quantification, forecasting, and procurement activities. The ware- We derived unit costs for fixed-cost items for all sites from price housing and distribution of commodities to VMMC sites was quotes obtained by PFSCM based on the commodities outlined in contracted through the local private sector market. With this Table 1. These items include those on the standard list of essential contract, PFSCM retained the use of sufficient warehouse space and equipment as well as infrastructure commodities needed to racking for all commodities for the ASI program over an 18-mo establish a mobile site. period, allowing for program scale-up and scale-down. As well, the local contractor provided trucks and drivers to accommodate both high- and low-demand scenarios for the in-country distribution of Supply Chain Costs supplies. PFSCM employed three full-time staff in Swaziland to Next are supply chain management costs, including warehous- oversee management of the warehouse, the management informa- ing, distribution, and other logistical services required to deliver tion system, the contractor, and monitoring of the sites. products to sites, as well as staffing costs. For the program, PFSCM PFSCM also provided technical assistance to support the hired local staff for in-country logistics and waste management Government of the Kingdom of Swaziland Ministry of Health in positions. The salary for the logistics advisor position was increasing the capacity and sustainability of the current waste calculated based on local salaries. Personnel costs for a locally disposal system while integrating programmatic VMMC waste hired logistics advisor were at rates exceeding local government disposal needs. At the national level, this included the revision of rates, but 50% lower than expatriate rates, to account for the need national health care waste management guidelines and policies to ensure the services of skilled staff. and supporting personnel with training, supervision, and tools for The calculated cost of non-facility-based overheads is 12% of in- record keeping and documentation. At the facility level, waste country operating costs, to account for technical assistance and management costs include transporting medical waste from the travel on the part of the PFSCM Project Management Office. The VMMC sites to the regional hospitals for final disposal; regularly cost of in-country supply chain and logistics staff was divided by monitoring and evaluating the plan to ensure that practices are the total number of targeted male circumcisions (152,000) or the maintained properly to minimize risk, damage, and disease revised target (75,000) for a single year of the VMMC scale-up transmission; procuring required waste disposal commodities program. (including incinerators) to ensure proper waste storage, handling, In addition to the costs identified above, we added 2% of the and disposal; and developing training curricula for facility staff total direct operating costs to cover overhead costs that were responsible for waste management [16]. neither facility- nor labor-based. For example, the 2% overhead would cover communication costs and other program-manage- Methods ment-related expenses.
Unit Cost Analysis for the Calculation of Commodities, Waste Management Operational Costs Logistics, and Waste Management Costs in Swaziland Waste management costs are both fixed and variable. PFSCM In addition to the programmatic costs associated with the performed an assessment of the national waste management operation of the supply chain, we assessed the fixed and variable system prior to developing the waste management plan for the cost of all quantified commodities, organized into six modules: (1) Swazi VMMC program. Based on the volume and weight of the VMMC kit and consumables, (2) infection prevention and waste commodities needed for each VMMC using fully disposable kits, management, (3) equipment for male circumcision sites, (4) PFSCM assessed the needs of the waste disposal system in emergency medical supplies, (5) HIV counseling and testing and Swaziland, using the combined weight of the consumables other STI management, and (6) temporary infrastructure. (Full required for an individual surgery to produce an estimate of contents of modules and line item costs are listed in Table S1.) approximately 0.5 kg of biologically contaminated waste to be The following sections detail the assumptions we made in generated per procedure. This level was more than the national calculating the unit cost for the ASI service delivery model. waste disposal system could process. In this analysis, staffing for waste management included the Service Utilization provision of technical assistance by two PFSCM headquarters staff In calculating the unit cost of a single VMMC, start-up or fixed and the training and support of two local staff (a driver and a waste costs, including equipment costs, are divided by the number of handler) for transportation of medical waste. As well, cost sharing
53 Supply Chain and Logistics for Male Circumcision
Table 1. Supply chain costs.
Category Sub-Category Target of 152,000 VMMCs Target of 75,000 VMMCs
Personnel and in-country operation costs Staff $246,720.00 $246,720.00 Office operation costs $37,100.00 $37,100.00 Supply chain costs Warehouse and cross docking in Matsapa $2,400.00 $2,400.00 Domestic distribution to 27 sites $160,245.00 $160,245.00 24-h security coverage at warehouses $26,190.00 $26,190.00 In-bound cartage JNB–MTSa plus customs clearance $45,045.00 $45,045.00 International freight $431,792.22 $257,474.41 Commodity insurance $11,939.05 $7,119.17 Fuel costs for site operations $105,840.00 $105,840.00 Fuel costs for incineration $109,500.00 $109,500.00 Distribution of waste from five regions/40 $160,245.00 $160,245.00 sites to region waste center Incinerator maintenance $39,250.00 $39,250.00 Fees Non-facility-based technical assistance overhead $21,488.00 $21,488.00 Non-facility, non-labor overhead $2,095.00 $2,095.00 Procurement fee $359,826.85 $214,562.01
All costs given in US dollars. aJohannesburg Grand Central Airport, South Africa, to Manzini-Matsapha International Airport, Swaziland. doi:10.1371/journal.pmed.1001128.t001 of incinerator operation and staff time at regional hospitals was from the manufacturer’s external packaging. This reduces the total included in the final cost. VMMC procedure time by removing the need to unwrap various commodities. Removal from the original packaging does destroy Results the items’ sterility, necessitating that the entire kit be sterilized using ethylene oxide gas before use. This service is provided by the Using the expanded list of WHO-recommended VMMC kit packager and included in the total cost of the kit. The market commodities, we estimated the commodity procurement, supply price for an ethylene-oxide-sterilized VMMC kit with a standard chain, and waste management costs associated with providing consumables pack and disposable forceps-guided instrument set VMMC services to 152,000 adult men in 1 y in Swaziland to be is currently US$17. US$8,956,213.46. This is the equivalent of US$58.92 per VMMC. Module 1 also includes bulk consumables that supplement For a reduced target of 75,000 VMMCs in 1 y, the total cost is those found in the VMMC kit. These include pharmaceuticals calculated as US$5,518,393.50, the equivalent of US$73.57 per necessary for the procedure (paracetamol, lignocaine, and VMMC. betadine) but that were removed from the kit itself due to expiration concerns and importation restrictions. Additional Module 1: VMMC Kit and Consumables sutures, gloves, and gauze also are included, in case those in Module 1 contains the ethylene-oxide-sterilized, fully disposable the kit are not adequate. The cost of these consumables is kit for performing the forceps-guided method of VMMC and estimated to be US$8.26 per VMMC. supplementary consumable medical supplies. These commodities are required for the performance of each surgery, representing a variable cost of US$25.17 per circumcision. In the Swaziland Module 2: Infection Prevention and Waste Management context with a target of 152,000 VMMCs, Module 1 costs The second module contains standard infection prevention contribute a total of US$3,839,144.08 to the commodity and items as well as those required for waste management. Previously, supply chain costs of the program. For a target of 75,000 VMMCs, based on WHO recommendations, this module contained only Module 1 contributes US$1,902,572.04 to the total cost of the infection prevention commodities. However, given the importance Swazi VMMC program, representing a per VMMC cost of of effective waste management and its potential impact on the US$25.18. broader health system, this estimate expands the module to Aligned with prior PEPFAR and WHO guidance, the VMMC include waste management commodities. kit comprises a consumables pack and an instrument set [14]. The Most infection prevention commodities are consumable and consumables pack contains a standardized set of disposable items represent variable costs, depending on the number of circumci- necessary for the performance of one male circumcision. This sions performed. Waste management commodities represent includes gauze, a scalpel, syringes, gloves, aprons, sutures, needles, largely fixed costs that require one-time procurement at the start surgical tape, alcohol swabs, and an O drape. The instrument set of program operations. In Swaziland, for a target of 152,000 contains all necessary surgical instruments for the performance of VMMCs, these commodities total US$1,427,853.90 and contrib- one circumcision. ute US$9.39 to the cost of each VMMC. To reach a target of Items within the kit are packed in the order in which they will be 75,000 VMMCs, the total cost of commodities for Module 2 is required for the surgical procedure. Before being placed in the kit, US$819,862.91, contributing US$10.93 to the cost of each consumable items and any necessary instruments are removed VMMC. Of these costs, US$211,872 is fixed.
54 Supply Chain and Logistics for Male Circumcision
Infection prevention commodities, while critical to program US$28,556.46, of which US$28,360 is fixed. This module implementation, are largely low cost and include items such as contributes US$0.38 to the cost of each VMMC. surgical masks and caps, protective eyewear, surgical scrub solution and alcohol hand wash, buckets for instrument cleaning, Module 5: HIV Counseling and Testing and Other STI and bleach. Management In contrast, waste management commodity costs can be This module includes HIV test kits, lancets, and pipettes, as well significant. Items such as sharps bins and biohazard bags are as pharmaceuticals for the treatment of STIs. These costs are used to collect three types of medical waste associated with a variable, and the quantity required will be determined by the rate VMMC program—sharps waste (two needles, one surgical blade, of consumption. For example, HIV tests should be recommended and one to two sutures), human tissue (foreskin), and general to all VMMC clients, making this cost dependent on expected medical waste (used gauze, gloves, etc.). This waste is transported demand. Confirmatory test kits should be ordered based on the to an incinerator, where it is turned into an ash by-product. HIV prevalence in the target population. Similarly, pharmaceu- For Swaziland, the ASI program will require the provision of tical costs for STI management will depend on the prevalence of upgrades to the current waste disposal facilities and the various STIs in the population. procurement of two additional incinerators, as well as the In Swaziland, the total cost of this module is US$486,833.70, or generators and fuel necessary to operate these incinerators. This US$3.20 per VMMC to reach a target of 152,000 VMMCs. To reach increases the cost of Module 2 by US$105,470 (US$0.69 per atargetof75,000VMMCs,thismodulecostsUS$243,416.85, or VMMC). US$3.25 per VMMC. Metal waste, generated by the use of either disposable or reusable instruments, cannot be incinerated and must be buried or Module 6: Temporary Infrastructure recycled. These disposal methods do not include additional Finally, two key commodities are included as a separate module commodity costs, but do require that the instruments be and represent costs associated with the mobile delivery of VMMC disinfected prior to disposal. In the case of the ASI program, services in Swaziland. The first comprises mobile medical disinfected metal waste will be donated to a metal foundry in (marquees) and industrial tents where VMMC mobile services Swaziland. Transportation costs to the foundry are included in the will operate. The second is generators, which will be used to power cost of metal waste disposal. diathermy machines and other electrical equipment at mobile sites. Also included are costs associated with fuel to operate the Module 3: Equipment for Male Circumcision Sites generators. The third module includes equipment necessary for the set-up We believe that these are important costs to assess, given that of one operating theater with four beds and one recovery area. many VMMC programs are considering mobile delivery of This includes such items as an operating table, lamp, instrument services. While the delivery method remains nearly identical in tray, intravenous stand, diathermy machine, and recovery chair. these mobile sites, there are key commodity considerations. In Small equipment items, such as a stethoscope and glucometer, are Swaziland, where tent structures will be used for 23 out of 27 sites also included. These items represent fixed costs, requiring (the remaining sites will use existing structures), these commodities procurement only once during program implementation. Gluc- contribute US$921,380 to the total cost of VMMC service ometer strips may be considered an exception, as they will require delivery. This is the equivalent of US$6.06 per circumcision. replacement over the lifetime of a program. Reducing the number of sites providing VMMC services to 20 For a program reaching a target of 152,000 VMMCs, with results in an overall reduction in the total cost of this module to service provision at 27 sites and 35 teams, the total cost of this US$835,130, with a per VMMC cost of US$11.14. Despite the module is US$469,202.33, or US$3.08 per VMMC, US$2,100 of reduction of the VMMC target by half, the program still requires which is variable costs. For a reduced target of 75,000 VMMCs the use of 20 sites to facilitate access to services over the entire achieved through service provision at 20 sites with 20 teams, this country. This results in a near doubling of the per VMMC costs in module contributes US$267,953.96 (US$1,050 variable costs) to this module despite a reduction in volume. the total cost of the Swazi VMMC program, or US$3.57 per VMMC. Supply Chain Costs The procurement, freight, and distribution costs associated with Module 4: Emergency Medical Supplies these commodities are significant. After estimating costs asso- The final WHO-recommended module, emergency medical ciated with each step of the supply chain, the total cost of these supplies, includes items necessary in the case of a medical services amounts to US$1,759,676.52. Costs included in the supply emergency. These include jump bags with necessary medical chain calculation are found in Table 1. This is reduced to equipment (such as glucometers, oxygen cylinders, and sphygmo- US$1,435,273.99 with a reduction in target to 75,000 VMMCs. manometers), emergency pharmaceuticals (atropine, adrenaline, The reduction is determined by the calculated procurement dextrose, and sodium chloride), and an emergency trolley to hold service fee as a percentage of total commodity costs. these items. With the exception of the pharmaceuticals, these costs Logistics costs include both international freight and in-country are fixed. Items in the jump bag may require occasional distribution to 27 service delivery sites. We also include com- replacement, but this cost will be minimal and dependent on the modity insurance and clearance costs based on current market number of adverse events in a program. These replacement costs rates, costs associated with warehousing commodities in Swazi- are not included in this calculation. land, and the costs of moving commodities from a central In Swaziland, for a program targeting 152,000 VMMCs through warehouse to individual surgical sites. Finally, we include the cost service provision at 35 sites, the total cost of this module is of transporting waste to incineration sites and performing US$50,022.92, or US$0.33 per VMMC. Of these costs, US$49,630 maintenance on incinerators. is fixed. For a VMMC program reaching 75,000 men through Last, we assess additional costs associated with in-country service provision at 20 sites, the total cost of this module is operation costs, particularly personnel. We include a waste
55 Supply Chain and Logistics for Male Circumcision management supervisor, as well as resources for monitoring and This paper provides a quantification that represents the best evaluation and technical assistance, given the embryonic state of international guidance on commodity procurement. For this waste management services in Swaziland. We include five reason, the analysis takes into account commodities not previously additional staffing positions for other supply chain and waste considered, including HIV counseling and testing, STI treatment, management services, as well as expenses associated with an office temporary infrastructure, and waste management; the last is for these staff. Staffing and other operation costs amount to particularly important, given the historical lack of attention to this US$283,820. area and its potential impact on strengthening the overall health system [7–12]. So, too, is temporary infrastructure given the Total Cost of Commodities and Supply Chain increasing understanding by program planners and policy makers Management that outreach services are an integral component of all VMMC Based on the above modules, we estimate the total cost of programs. commodities, supply chain, and waste management for the The data presented here represent rapid scale-up of VMMC in Swaziland VMMC program reaching a target of 152,000 men a low-resource setting with minimal supply chain systems. These to be US$8,956,213.46. Approximately US$2,609,202.66 of these assumptions impact the calculations provided and limit their costs is fixed, while the remaining US$6,334,010.80 is variable, generalization to other country settings. Various factors need to be based on the number of VMMCs performed, the number of considered before this calculation is used to inform costing medical staff, and the number of surgical sites. Fixed and variable estimations in other settings. costs are outlined by module in Table 2. First, better-resourced countries may incur lower program costs The total cost of commodities, supply chain, and waste because of existing infrastructure and supply chain functions (e.g., management is US$58.92 per male circumcision. warehousing, waste management). For example, according to the For a reduced target of 75,000 VMMCs, the total cost is estimates reported here, commodities make up a significant calculated as US$5,518,393.50, which equals a per VMMC cost of proportion of the total unit cost (US$47.34–US$49.50). Consum- US$73.57. Fixed and variable costs for this reduced target are also ables contribute more than two-thirds of the supply chain costs for included in Table 2. VMMC programs. These costs may be reduced below the estimates presented here by using existing infrastructure (i.e., Discussion existing operating tables) and strategically placing pooled orders, which take advantage of economies of scale and allow for the This report represents the first attempt, to our knowledge, to negotiation of better pricing on bulk commodities. Similar use of quantify fully the supply chain and waste management costs existing infrastructure could reduce costs associated with the associated with VMMC programming. Previous VMMC program procurement of temporary infrastructure, which adds at least costing studies generally have been performed at the point of US$6.06, and as much as US$11.14, to the cost of each surgery service provision and thus have failed to capture supply chain and (e.g., tents, generators). Additionally, the use of in-country as waste management costs. However, given the significant contri- opposed to expatriate staff can result in considerable savings. In bution to total program costs made by commodities, it is critical the case of Swaziland, staffing costs can be reduced by at least 50% that these costs be considered during implementation planning to through hiring skilled local staff. provide an accurate indication of the resources required. The implementation timeline may also alter costs. The costs For example, based on previous costing studies, which indicated associated with the ASI program in Swaziland assume the an average cost of US$50 per VMMC, the initial resource needs generation of demand for 152,000 or 75,000 VMMCs in 12 mo; assessment for the ASI program underestimated the actual costs by a lengthier timeline needed to meet these targets would increase approximately US$8 million. For program planners, an underes- costs for rentals, warehousing and distribution contracting, timate of this magnitude would gravely impact successful staffing, and some formerly bulk-procured commodities. Countries program implementation. Moreover, given Joint United Na- planning to perform a larger number of VMMCs over a longer tions Programme on HIV/AIDS (UNAIDS) estimates of a period of time will see some cost savings as certain flat-rate costs US$10,000,000,000 funding gap for HIV programs globally are distributed across more surgeries (e.g., surgical beds, diathermy [17], accurate program costs are critical to determining the most machines). efficient and effective use of resources to impact the global Program planners will also need to consider the fact that the epidemic. number of VMMC sites will impact fixed costs and change
Table 2. Fixed and variable costs, by module.
Module Target of 152,000 VMMCs Target of 75,000 VMMCs Fixed Variable Fixed Variable
VMMC kits and consumables None $3,839,144.08 None $1,902,572.04 Infection prevention and waste management $211,871.93 $1,215,981.98 $211,871.93 $607,990.99 Equipment for male circumcision sites $469,202.33 $2,100.00 $266,903.96 $1,050 Emergency medical supplies $49,630.00 $392.92 $28,360.00 $196.46 Counseling and testing and other STI management $1,000.00 $485,833.70 $500 $243,416.85 Temporary infrastructure $921,380.00 None $835,130.00 None
All costs given in US dollars. doi:10.1371/journal.pmed.1001128.t002
56 Supply Chain and Logistics for Male Circumcision resource needs. These costs will increase with the number of sites Finally, in Swaziland, the costs of waste disposal were based on selected by program planners. The selection of the appropriate the use of incineration for biological waste disposal. For other number of sites for a particular campaign would be determined by countries, costs will be impacted for programs choosing alternative such factors as volume over a given time frame, geography, size, disposal methods. As mentioned previously, countries can adapt difficulty of transport, and population density. However, given that cost estimates to reflect the reality of their infrastructure (which many VMMC programs are expected to move quickly to full scale might vary from site to site). As in the case of Swaziland, and reach target saturation within 2–3 y of initiation, many of the improvements made to the health care waste management system costs of increased targets simply would be multiplied by the for the disposal of VMMC programmatic waste can greatly number of teams and/or sites at a commensurate increased cost. improve waste management abilities throughout the health care Alternatively, the volume through an individual site (assuming a system. maximum capacity of 80 VMMCs per team per day based on the MOVE model) is associated with variable costs, which would Lessons Learned increase as volume increases, while the fixed costs associated with VMMC programs implemented at the scale and pace required each site would decrease as volume increases. to stem HIV transmission and save lives as quickly as possible are The largest single cost in consumables is that of the disposable somewhat at odds with the global health community’s focus on $ $ kit, at US 1,275,000–US 2,584,000. Prices for forceps-guided strengthening health systems. The WHO MOVE recommenda- disposable kits were negotiated by PFSCM for the aggregate group tions attempt to mitigate demands on the health care system by of instruments (rather than by line item), including packaging and task shifting, creating efficient patient flows, and simplifying the sterilization. We did this to take advantage of economies of scale, complex list of commodities needed for a successful program. Still, not to be directive regarding VMMC method. The disposable kit much like vaccination campaigns, the implementation of VMMC seems relatively expensive but does help to improve efficiency and programs often involves the creation of a parallel supply chain for avoids the need for sterilization equipment and staff training. In VMMC commodities outside of the national essential medicines fact, through PFSCM, the pooled procurement of VMMC kits supply chain; this parallel supply chain is generally managed by already has resulted in a reduction in market price from US$23 to the implementing partner providing VMMC services. These US$17 per kit. Further reductions in prices are expected as commodities are varied (requiring the management of over 50 additional VMMC kit vendors enter the market. As of March different consumable products alone), demand large well-ordered 2011, 6 mo after the procurement of VMMC kits for Swaziland, storage facilities, and must be distributed to a range of dedicated that cost had already dropped to US$15 per kit. and mobile clinics with seasonal operating schedules and locations. While the Swazi ASI chose to use the forceps-guided VMMC In the case of Swaziland, and given the lack of precedent in the technique, some programs have chosen to use other methods of VMMC (i.e., sleeve resection or dorsal slit). These techniques literature, supply chain and logistics needs to support the require different and/or additional instruments, require more time ambitious program implementation plan were vastly underesti- per procedure, and are less commonly used in VMMC programs. mated and under-budgeted. Commodities included in the initial Program planners choosing to use one of these methods can expect action plan were limited to VMMC kits and HIV rapid test kits; an increase in commodity costs for the prepackaged kit, which is the plan for the management of the supply chain in Swaziland was currently priced at US$19 per kit. As well, increasing the time limited to procurement, delivery to a small warehouse for required per surgery results in a decrease in the volume of clients temporary cross docking, and near immediate distribution to serviced per team at the site level, resulting in increased fixed costs. VMMC service provision sites. The staffing plan included only a As discussed, some programs may opt for reusable rather than logistics advisor and an administrative assistant. disposable instruments. However, reusable instrument ‘‘lifetimes’’ The exercise documented here helped program planners vary significantly because of variations in instrument quality and address this challenge, vastly improving estimation and budgeting differences in cleaning and maintenance practices. In addition, for the Swaziland program. Several key lessons learned from this there are costs for reusable instrument use (in equipment purchase, process will be useful to other VMMC programs in Africa to maintenance, training, human resource time expenditures, ensure efficient resource allocations. cleaning solutions, and instrument turnaround time), so one First, proper assessment of programmatic needs can prevent cannot assume that reusable instruments are more cost-effective in programs from being under-resourced. For example, because over every context. Programs with pre-existing autoclave facilities and two-thirds of the total supply chain budget is devoted to the staff that can be expanded (or fully utilized, if not currently at procurement of nearly 100 different commodities, the act of capacity) are the most likely to derive cost savings from using quantifying program needs allowed for a more realistic budget to reusable instruments. Some programs might choose a combi- be created (an increase from US$2.5 million to US$8.9 million). nation of reusable instruments at sites with autoclaves and Without this, the VMMC program in Swaziland would have been disposable kits at less well-equipped mobile and outreach sites under-resourced, stalling before a single procedure was performed. [14]. Such variable patterns of program design have significant Second, assessment of needs and cost estimation also ensure that cost impacts that are difficult to quantify and cannot be reflected existing resources are properly allocated. In Swaziland, the waste fully in this report. management system assessment conducted prior to launch of the In addition, this report assumes the use of a diathermy machine program was critical to ensuring appropriate resource allocation. for cauterization, which changes costs associated with instru- The assessment was conducted out of a recognition that because ments by decreasing the number of sutures required. Programs most African waste management programs are unable to meet considering the use of diathermy machines should consider costs current waste disposal demands [18], the effort to reach national associated with additional staff training required to operate these targets through VMMC programs will quickly overwhelm national machines safely, as such costs are not reflected in this paper. Only waste disposal systems. As a result of the assessment, the costs programs with a reliable electricity supply, stable voltage, and estimated for the management of VMMC-program-generated demand for high-volume VMMC services are likely to derive waste were significantly reduced, and the estimated number of significant benefit from the investment in diathermy machines. incinerators needed decreased from six to two. This reduction in
57 Supply Chain and Logistics for Male Circumcision anticipated waste management costs allowed for reallocation of into account the additional costs associated with such repairs in funding resources to other programmatic priorities. their specific context. Additionally, costs for resupply of emergency Finally, despite extensive needs assessment and cost estimation, drugs in case of adverse events are not included. program implementation can still result in unexpected supply Another limitation to this paper is that the data for commodities chain costs associated with the implementation of a VMMC are derived from current market prices for quantities calculated at program. In the Swaziland context, funding delays, the reality of an estimated rate of consumption, rather than expenditure lead times associated with any procurement process, and slow reporting. It is quite likely that prices of certain commodities will community engagement led to a late start to program activities be impacted by supply/demand issues both locally and interna- (start delayed from January to April). This meant that at the tionally. Moreover, if there is a wide variation in the specific midpoint of the campaign, communication efforts had generated commodities ordered by program planners, the potential econo- less demand for service than anticipated. This has significantly mies of scale can be lost or diluted. impacted supply chain and logistics needs on the ground. This variation in commodity costs is a particular limitation for Commodities were initially ordered to support a program the cost estimates provided for the performance of 75,000 meeting a target of 152,000 VMMCs over a 1-y period. With a VMMCs. Because PFSCM negotiated prices based on an initial lower than expected demand for services, the supply chain for volume of 152,000 VMMCs, the line item costs used for this these commodities has required a great deal of change. From a estimate reflect this higher bulk pricing. It is possible that if 75,000 cross-docking operation that quickly moved commodities to the VMMCs had been the program target at the outset, rather than a field for immediate consumption, at the country level, the Swazi function of decreased demand, commodity costs would have been VMMC supply chain has evolved into a complicated operation higher because of lower volumes during pricing negotiations. This involving warehouse management, distribution, oversight of may have resulted in a further increase in the per VMMC price. commodities consumption and waste disposal practices at the Also, this paper does not take into account the impact of scale- service delivery site, forecasting of needs, and reverse logistics up strategies (such as numbers of sites and VMMC volumes per of program-generated waste. Ultimately, from an initial plan of site over time) or the need for and cost of different commodities hiring two staff, five were hired to manage the supply chain, over time. However, where possible, we have outlined fixed and in addition to the expansion of the scopes of contracts for variable costs associated with VMMC scale-up. These calculations warehousing and distribution of supplies. should be assessed carefully based on local context to avoid over- In addition, for VMMC kits (US$17 per kit) procured in or underestimation of costs. For example, if a site is not operating January 2011 for a 1-y program, a 2-y shelf life was rapidly being at full capacity, high fixed costs will increase the per procedure cost spent. At the current rate of service provision, the program was at that facility. projected to reach 75,000 VMMCs at the end of the 1-y period. The implementation time period (scale-up, scale-down) is not Given this forecast, it was decided that of 100,000 kits currently in included as a variable in the calculations in this paper. The time country, 40,000 should be sold internationally. This required over which services are delivered can be leveraged to maximize packaging, transportation, and customs brokerage, but allowed a cost efficiencies. For example, as VMMC coverage goals are met, savings of nearly US$1 million in programmatic costs that might the demand drops and low-volume sites can be closed. The have been wasted had the commodities expired in the warehouse. remaining usable commodities at these sites (beds, lights, This implementation experience speaks to an important lesson instruments) can be used to replace at active sites those com- learned regarding the necessity of staging the importation of modities that are worn out or in need of repair, thus reducing commodities into a country based on real consumption data. A repair/replacement costs. balance exists between staging orders and a desire to save money Finally, this report estimates costs associated with VMMCs by pooling small orders into a much larger one; for untested performed according to current WHO guidelines [19]. Research VMMC programs, this desire should be tempered by the into novel technologies, including circumcision devices and uncertainty of demand in country. methods of local anesthesia administration, are ongoing through- out Africa, and these technologies will likely be more broadly Limitations available within the next 2–3 y. These technologies have the While we include line item costs in this paper for one VMMC potential to transform the VMMC service provision model and scale-up program, commodity and service costs can vary may drive commodity prices lower, resulting in additional cost significantly, depending on location, order volume, and other savings for the intervention. The prospective use of such market forces. The widely variable setting in which VMMC scale- technologies, the costs of training staff in performing new up is being undertaken has such a significant impact on line item techniques, and the need to revise VMMC kit composition values as to make them almost impossible to estimate for broad and other associated commodities should be considered by applicability. However, this paper provides a framework for supply program planners in the transition between the old and new chain and waste management cost considerations. In addition, procedures for future VMMC programs as new information varying quality of such items can impact cost (and usable lifetime) becomes available. significantly, and it is difficult to differentiate specific commodities in terms of quality (aside from field testing, for which only Conclusion anecdotal data are available). In addition, although certain commodities are listed as fixed As VMMC programs mature, further efforts regarding costing costs (operating table, trolleys, instrument stands), there may be should be undertaken to improve on the current estimates of the additional costs not included here, such as the cost of maintenance cost of VMMC. Experience from implementing the ASI VMMC and repair. Such costs are difficult to estimate and will depend on program in Swaziland indicates that supply chain and waste such factors as the volume of clients, climate, quality of routine management add an additional approximately US$60 per maintenance, and total time used. Due to the significant variability VMMC, which adds nearly 100% in additional costs per of such costs from country to country, programs will need to make procedure to program costs. This analysis provides a framework procurement decisions based on their particular settings and take to inform program planning with a more accurate representation
58 Supply Chain and Logistics for Male Circumcision of the critical supply chain and waste management costs of exercise in Zimbabwe; the US Department of Defense, for providing cost VMMC services. estimates to the paper; and the Government of the Kingdom of Swaziland Ministry of Health, for its role in developing the ASI implementation plan, on which the commodities analyzed here were based, and its partnership Supporting Information with PFSCM in developing the plan to increase the capacity of the national Table S1 Commodities and price list. waste management system. We would like to acknowledge Jason Reed and (DOCX) Naomi Bock for their contribution to the writing and review of versions of the paper. Acknowledgments Author Contributions Thanks to Orange Farm and Population Services International for Conceived and designed the experiments: DE EN. Analyzed the data: DE information related to the first VMMC kits and modules developed by PS SF SS EN. Wrote the first draft of the manuscript: DE PS. Contributed PFSCM; PEPFAR, WHO, and UNAIDS, for convening the expert to the writing of the manuscript: DE PS SF DR KC TA SS AGT JA DS consultation in May 2009; the PEPFAR Male Circumcision Technical KED EN. ICMJE criteria for authorship read and met: DE PS SF DR KC Working Group, which contributed to the development of the standard list of commodities; PFSCM, for obtaining current market prices for TA SS AGT JA DS KED EN. Agree with manuscript results and commodities for the ASI program; Zimbabwe Ministry of Health and conclusions: DE PS SF DR KC TA SS AGT JA DS KED EN. Child Welfare, for sharing the cost for VMMC derived from the costing
References 1. Auvert B, Taljaard D, Lagarde E, Sobngwi-Tambekou J, Sitta R, et al. (2005) Epidemiology and Biostatistic at Makere University, Ugandan Ministry of Randomized, controlled intervention trial of male circumcision for reduction of Health (2011) Assessing the potential impact, costs of scaling-up voluntary HIV infection risk: The ANRS 1265 Trial. PLoS Med 2: e298. doi:10.1371/ medical male circumcision services in Uganda. Washington (District of journal.pmed.0020298. Columbia): United States Agency for International Development Health Policy 2. Bailey RC, Moses S, Parker CB, Agot K, Maclean I, et al. (2007) Male Initiative. circumcision for HIV prevention in young men in Kisumu, Kenya: A 11. United States Agency for International Development Health Policy Initiative, randomised controlled trial. Lancet 369: 643–656. Joint United Nations Programme on HIV/AIDS, Zambian Ministry of Health 3. Gray R, Kigozi G, Serwadda D, Makumbi F, Watya S, et al. (2007) Male (2011) Costing of voluntary medical male circumcision services and the impact circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial. of accelerated scale-up in Zambia. Washington (District of Columbia): United Lancet 369: 657–666. States Agency for International Development Health Policy Initiative. 4. Njeuhmeli E, Forsythe S, Reed J, Opuni M, Bollinger L, et al. (2011) Voluntary 12. United States Agency for International Development Health Policy Initiative, medical male circumcision: modeling the impact and cost of expanding male Joint United Nations Programme on HIV/AIDS, Zimbabwe Ministry of Health circumcision for HIV prevention in eastern and southern Africa. PLoS Med 8: and Child Welfare (2011) Costing of voluntary medical male circumcision and e1001132. doi:10.1371/journal.pmed.1001132. the impact of accelerated scale-up in Zimbabwe. Washington (District of 5. Curran K, Njeuhmeli E, Mirelman A, Dickson K, Adamu T, et al. (2011) Columbia): United States Agency for International Development Health Policy Voluntary medical male circumcision: strategies for meeting the human resource Initiative. needs of scale-up in southern and eastern Africa. PLoS Med 8: e1001129. 13. The United States President’s Emergency Plan for AIDS Relief (2009) PEPFAR doi:10.1371/journal.pmed.1001129. Male Circumcision Partners’ Meeting: commodities and improved coordination 6. World Health Organization, Joint United Nations Programme on HIV/AIDS of male circumcision for HIV prevention. Washington (District of Columbia): (2007) New data on male circumcision and HIV prevention: policy and The United States President’s Emergency Plan for AIDS Relief. programme implications. Geneva: World Health Organization. 14. World Health Organization (2010) Considerations for implementing models for 7. United States Agency for International Development Health Policy Initiative, optimizing the volume and efficiency of male circumcision services. Geneva: Namibia Ministry of Health and Social Services (2011) Costing of voluntary World Health Organization. medical male circumcision services and the impacts of accelerated scale-up in Namibia. Washington (District of Columbia): United States Agency for 15. Swaziland Male Circumcision Task Force, Government of the Kingdom of International Development Health Policy Initiative. Swaziland (2010) Male circumcision for HIV prevention accelerated saturation 8. United States Agency for International Development Health Policy Initiative, initiative action plan for Swaziland. Mbabane: Swaziland Male Circumcision Joint United Nations Programme on HIV/AIDS, Kenya Ministry of Public Task Force. Health and Sanitation (2011) Estimating the costs and impacts of expanding 16. Partnership for Supply Chain Management (2010) Male circumcision waste voluntary medical male circumcision services in Kenya. Washington (District of management plan. Arlington (Virginia): Partnership for Supply Chain Columbia): United States Agency for International Development Health Policy Management. Initiative. 17. Joint United Nations Programme on HIV/AIDS (2010) UNAIDS report on the 9. United States Agency for International Development Health Policy Initiative, global AIDS epidemic. Geneva: Joint United Nations Programme on HIV/ Joint United Nations Programme on HIV/AIDS, South Africa National AIDS. Department of Health (2011) Costing of voluntary medical male circumcision 18. United Nations Educational, Scientific and Cultural Organization (2009) Africa service and impacts of accelerated scale-up in South Africa. Washington (District review report on waste management. Paris: United Nations Educational, of Columbia): United States Agency for International Development Health Scientific and Cultural Organization. Policy Initiative. 19. World Health Organization, Joint United Nations Programme on HIV/AIDS, 10. United States Agency for International Development Health Policy Initiative, Jhpiego (2008) Manual for male circumcision under local anaesthesia. Geneva: Joint United Nations Programme on HIV/AIDS, Ugandan Department of World Health Organization.
59 Supply Chain and Logistics for Male Circumcision
Editors’ Summary Background About 33 million people (including 22.5 maintenance of incinerators. They also estimated the staffing million in sub-Saharan Africa) are currently infected with costs of supply chain and waste management services. From HIV, the virus that causes AIDS. Although antiretroviral drugs these component costs, the researchers estimate that, keep HIV in check, there is no cure for HIV/AIDS. overall, the costs of supply chain and waste management, Consequently, prevention of HIV transmission is extremely including procurement of commodities and associated labor, important. Because HIV is usually spread through add US$58.92 if 152,000 men are circumcised and US$73.57 if unprotected sex with an infected partner, individuals can 75,000 men are circumcised to the previously estimated cost reduce their risk of becoming infected with HIV by of performing one adult male circumcision through the abstaining from sex, by having only one or a few partners, Swaziland ASI VMMC program. and by always using male or female condoms. In addition, trials in sub-Saharan Africa have shown that male What Do These Findings Mean? This study suggests that, circumcision—the removal of the foreskin, the loose fold of for the Swaziland ASI VMMC program, procurement, supply skin that covers the head of the penis—reduces the risk of chain, and waste management costs nearly double the HIV infection in men by 60%. In 2007, the World Health previously estimated cost per VMMC procedure. That is, the Organization (WHO) and the Joint United Nations supply chain and waste management costs for this program Programme on HIV/AIDS (UNAIDS) recommended that are nearly as high as the costs of the equipment and staff voluntary medical male circumcision (VMMC) should be needed to do the circumcisions. Because these costs were part of HIV prevention programs in regions with a not taken into account during the planning stages of generalized HIV epidemic and a low level of male Swaziland’s ASI VMMC program, the initial needs circumcision. Together with the United States President’s assessment for this program underestimated the actual Emergency Plan for AIDS Relief (PEPFAR), WHO, and UNAIDS costs by about US$8 million. Although the magnitude of this also prioritized 14 countries in eastern and southern Africa underestimate cannot be generalized to other settings, this for VMMC program scale-up. Mathematical models suggest analysis emphasizes the importance of considering the that, if 80% VMMC coverage is reached by 2015 (which will contribution of supply chain and waste management to entail performing 20.33 million circumcisions between 2011 costs when determining the future resource needs of VMMC and 2015) and sustained thereafter, VMMC programs in programs. Moreover, it provides a framework to help these priority countries will avert more than 4 million HIV program planners and policy makers estimate the costs infections among adults between 2009 and 2025. involved in the scale-up of VMMC programs in other priority countries. Why Was This Study Done? Successful VMMC scale-up will depend on the commodities that are essential for VMMC Additional Information Please access these websites via services being accessible and on the appropriate allocation the online version of this summary at http://dx.doi.org/10. of resources to support VMMC programs (which, in addition 1371/journal.pmed.1001128. to circumcision, include HIV testing and counseling, sexually transmitted infection screening and treatment, condom N This study is part of a PLoS collection of articles on VMMC provision and promotion, and counseling on risk reduction (http://www.ploscollections.org/VMMC2011) and is further and safer sex). To help program planners and policy makers, discussed in a PLoS Medicine Review Article by Hankins costing studies have been undertaken in several African et al. (http://dx.doi.org/10.1371/journal.pmed.1001127). countries. These studies considered the costs of a standard N Information is available from WHO, UNAIDS, and PEPFAR list of commodities prepared by PEPFAR, WHO, and UNAIDS on all aspects of HIV/AIDS and estimated that, on average, one male circumcision costs N NAM/aidsmap provides basic information about HIV/AIDS, about US$53. However, these studies did not include the summaries of recent research findings on HIV care and costs of the supply chain, waste management, HIV treatment, and information on male circumcision for the counseling and testing, treatment of sexually transmitted prevention of HIV transmission infections, or the temporary infrastructure needed to deliver Information is available from Avert, an international AIDS mobile VMMC services. Here, the researchers estimate these N charity on many aspects of HIV/AIDS, including informa- hitherto ignored costs for the Accelerated Saturation tion on all aspects of HIV prevention, and on HIV/AIDS in Initiative (ASI; Soka Uncobe [Circumcise and Conquer] in Africa and in Swaziland (in English and Spanish) SiSwati), a one-year program to circumcise 152,000 men in Swaziland. N The Clearinghouse on Male Circumcision, a resource provided by WHO, UNAIDS, and other international bodies, What Did the Researchers Do and Find? The researchers provides information and tools for VMMC policy develop- used current market prices of goods and services to calculate ment and program implementation the fixed and variable costs of various aspects of the VMMC N Personal stories about living with HIV/AIDS are available commodity supply chain such as procurement, international through Avert, through NAM/aidsmap, and through the freight, in-country distribution to service delivery sites, and charity website Healthtalkonline warehousing, and of various aspects of waste management, such as the transportation of waste to incinerators and the
60 Voluntary Medical Male Circumcision: A Qualitative Study Exploring the Challenges of Costing Demand Creation in Eastern and Southern Africa
Jane T. Bertrand1*, Emmanuel Njeuhmeli2, Steven Forsythe3, Sarah K. Mattison1, Hally Mahler4, Catherine A. Hankins5 1 Tulane School of Public Health and Tropical Medicine, New Orleans, Louisiana, United States of America, 2 United States Agency for International Development, Washington, District of Columbia, United States of America, 3 Futures Institute, Glastonbury, Connecticut, United States of America, 4 Jhpiego, Dar es Salaam, Tanzania, 5 Joint United Nations Programme on HIV/AIDS, Geneva, Switzerland
Abstract
Background: This paper proposes an approach to estimating the costs of demand creation for voluntary medical male circumcision (VMMC) scale-up in 13 countries of eastern and southern Africa. It addresses two key questions: (1) what are the elements of a standardized package for demand creation? And (2) what challenges exist and must be taken into account in estimating the costs of demand creation?
Methods and Findings: We conducted a key informant study on VMMC demand creation using purposive sampling to recruit seven people who provide technical assistance to government programs and manage budgets for VMMC demand creation. Key informants provided their views on the important elements of VMMC demand creation and the most effective funding allocations across different types of communication approaches (e.g., mass media, small media, outreach/ mobilization). The key finding was the wide range of views, suggesting that a standard package of core demand creation elements would not be universally applicable. This underscored the importance of tailoring demand creation strategies and estimates to specific country contexts before estimating costs. The key informant interviews, supplemented by the researchers’ field experience, identified these issues to be addressed in future costing exercises: variations in the cost of VMMC demand creation activities by country and program, decisions about the quality and comprehensiveness of programming, and lack of data on critical elements needed to ‘‘trigger the decision’’ among eligible men.
Conclusions: Based on this study’s findings, we propose a seven-step methodological approach to estimate the cost of VMMC scale-up in a priority country, based on our key assumptions. However, further work is needed to better understand core components of a demand creation package and how to cost them. Notwithstanding the methodological challenges, estimating the cost of demand creation remains an essential element in deriving estimates of the total costs for VMMC scale-up in eastern and southern Africa.
Citation: Bertrand JT, Njeuhmeli E, Forsythe S, Mattison SK, Mahler H, et al. (2011) Voluntary Medical Male Circumcision: A Qualitative Study Exploring the Challenges of Costing Demand Creation in Eastern and Southern Africa. PLoS ONE 6(11): e27562. doi:10.1371/journal.pone.0027562 Editor: Stephanie L. Sansom, Centers for Disease Control and Prevention, United States of America Received June 28, 2011; Accepted October 19, 2011; Published November 29, 2011 This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication. Funding: This work was funded by the President’s Emergency Plan for AIDS Relief (PEPFAR) through the United States Agency for International Development (USAID) and UNAIDS. Technical staff from USAID and UNAIDS was involved in study design, decision to publish, and preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. Abbreviations: MC, male circumcision; MCHIP, Maternal and Child Health Integrated Program; SMS, short message service; UNAIDS, Joint United Nations Programme on HIV/AIDS; USAID, United States Agency for International Development; VMMC, voluntary medical male circumcision * E-mail: [email protected]
Introduction country governments in eastern and southern Africa considered scaling up VMMC services: Botswana, Kenya, Lesotho, Malawi, By 2007, three clinical trials had provided compelling evidence Mozambique, Namibia, Rwanda, South Africa, Swaziland, that voluntary medical male circumcision (VMMC) reduces Tanzania, Uganda, Zambia, and Zimbabwe. female-to-male HIV transmission by about 60% [1–3]. Following VMMC service implementation and scale-up success depend on a review of these results in 2007, the World Health Organization both the supply of, and the demand for, high-quality safe services and the Joint United Nations Programme on HIV/AIDS for those who would most benefit. Elements of supply include (UNAIDS) recommended that VMMC services be expanded as voluntary HIV testing and counseling offered prior to the surgery, part of a comprehensive HIV prevention package, especially in MC counseling, the surgical procedure; routine follow-up visits settings with high HIV prevalence, low levels of male circumcision and other contacts (e.g., short message service [SMS] messaging to (MC), and generalized heterosexual epidemics [4]. With the facilitate return in case of adverse events), and emergency services support of the international donor community, the following 13 in case of complications. Facility-level costs associated with the
61 Costing Demand Creation for MMC Scale-Up supply of VMMC services include direct and indirect labor, Using purposive sampling, we recruited seven key informants to material, equipment, and overhead (including maintenance and participate in the study. We conducted interviews by telephone in utilities). Additional costs include those for training, policy early 2011; these lasted an average of 1 h. We obtained written development, and supply chain management (including waste responses when phone conversations were not possible. The disposal). In collaboration, the United States Agency for discussion guide appears in Text S1. The informants were International Development (USAID) and UNAIDS have assisted communication specialists working for agencies providing demand countries in estimating the direct costs of scaling up VMMC creation technical assistance to African governments actively services by supporting facility-based cost data collection to involved in VMMC scale-up and administering specific budgets populate the Decision Makers’ Program Planning Tool, which for demand creation. The agencies for whom they were working estimates both the cost and HIV impact of various programming were the Academy for Educational Development/C-Change in options [5,6]. Such costing studies have been conducted in Kenya, Kenya; the Center for Communication Programs, Johns Hopkins Namibia, South Africa, Uganda, Zambia, and Zimbabwe, and are Bloomberg School of Public Health in South Africa; Jhpiego, an underway in Tanzania and Malawi. affiliate of Johns Hopkins University in Tanzania; and Population ‘‘Demand’’ refers to the level of interest or motivation to access Services International in Kenya and Zambia. services and obtain the procedure on the part of the population prioritized for VMMC. Demand creation aims to influence the Results attitudes, intentions, and, ultimately, the decisions of men to seek VMMC services. Demand creation comprises a constellation of Demand Creation Packages and Channel Mix for activities designed to increase awareness of VMMC among both Resource Allocation men and women (including guardians of adolescents), provide With respect to the potential elements of a demand creation factual information regarding its benefits, inform the population package and recommended resource allocation across communi- about where services can be obtained, dispel myths, influence cation channels (i.e., channel mix), key informants emphasized the community norms regarding MC, and encourage men and difficulty of employing a ‘‘one size fits all’’ approach to demand adolescent boys to seek these services, thus increasing the flow of creation, given the spectrum of media environments ranging from clients to them. We are unaware of any attempts to date to urban South Africa to rural Tanzania. Nonetheless, they understand and estimate the cost of demand creation for VMMC addressed the question: ‘‘What do you consider to be the ideal [7–10]. resource allocation for communication for male circumcision?’’ There is a growing literature of acceptability studies that Their responses were divided among mass media, print materials attempt to measure how receptive men in different countries are to (small media), and outreach/mobilization. The recommended MC, the reasons for their interest, and the barriers to getting budget allocation for mass media relative to all forms of circumcised [11–16]. Westercamp et al. [15], summarizing 13 communication channels ranged from 20% to 60% (with qualitative studies conducted in nine countries in sub-Saharan responses almost evenly distributed across this range). Most Africa, found that 65% of men (range, 29%–87%) were willing to wanted some, if not all, mass media funding to go to radio, with the exception of South Africa, where more than 80% of the become circumcised, 69% (47%–79%) of women favored population watches television. There was strong support for the circumcision for their partners, and 71% (50%–90%) of men use of ‘‘targeted radio’’_i.e., regional or community stations near and 81% (70%–90%) of women were willing to circumcise their VMMC sites that broadcast specific information related to the sons. Whereas acceptability studies provide useful insights into the services provided in that area (e.g., the date and time when attitudes, perceptions, and values of men and women in different services will be available at a particular location). One informant countries, specific factors that influence the demand for VMMC strongly advocated local radio but recommended a relatively small likely vary by setting, requiring local information gathering to allocation because ‘‘local radio is so inexpensive in our country.’’ inform demand creation programs. Television, which was allocated less than 10% in all budgets, The aim of this paper is to propose an approach for estimating was not considered essential for several reasons. One informant the costs of demand creation for VMMC scale-up by addressing mentioned that it was useful for creating awareness about the the following questions. (1) What are the elements of a benefits of VMMC at the start of the scale-up but was not an standardized package for demand creation? (2) What methodo- effective mechanism for prompting men to make a decision to logical challenges exist and must be taken into account in undergo VMMC or go to a VMMC site. This same informant estimating the costs of demand creation? (3) Taking these would consider using television in a similar situation if the budget challenges into account, what are the recommended steps in for VMMC communication was generous. Other limitations of costing VMMC demand creation? Drawing on the principles and television that were mentioned were its inability to reach some experience of demand creation in other areas of public health, we rural areas and/or the frequent shut-downs of electricity. By explore adaptations to the field realities of VMMC programs in contrast, another informant was far more positive about the eastern and southern Africa. potential of television, especially drama series, for social modeling (i.e., exposing men to desirable models of behavior, including Methods VMMC). He noted that television drama on its own would be insufficient and that additional advertising would be required, In any type of multi-country costing exercise, it is important to particularly in environments where there is a diversity of media develop a standard list of items to be costed, even if these countries available to consumers. Others commented that television can be do not incur all of the expenses and/or incur other costs not on the useful and inexpensive if a news program does a feature on the list. As no such standard package appears to exist for VMMC VMMC program or a television talk show invites staff to serve as demand creation, we conducted a key informant study aimed at guests. This television coverage comes at no expense to the identifying (1) elements in a standard demand creation package VMMC program and increases its visibility. and (2) the percentage of the demand creation budget to allocate There was little enthusiasm for billboards; only one informant to different types of communication (i.e., channel mix). would give them 10% of the budget (for use in urban areas only)
62 Costing Demand Creation for MMC Scale-Up and the others would allocate zero. The perceived limitation is that organizations. Efficient use of resources requires training billboards do not lend themselves to communicating a lot of existing outreach personnel to discuss VMMC in the broader information on a complex topic. Moreover, they generally are not context of HIV prevention and paying them when they give available in rural areas. One informant cited wall hangings outside talks on VMMC. of shops as a viable alternative to billboards in rural areas. N Targeting educational institutions (e.g., schools, universities) Programs that opt to use billboards can increase their effectiveness allows programs to reach adolescents and young men in these by including a link to service delivery, such as a helpline number. settings–the age groups that to date have proven most Informants allocated the remaining support (ranging from 30% receptive to VMMC. to 60%) to outreach. Several informants included the costs of print N Integrating VMMC messages into the broader context of HIV materials (small media) in outreach, presumably because they are prevention (and not as an isolated topic) is essential. The an essential element contributing to effective outreach. Programs messages must convey that VMMC is only partially protective often had at least two types of brochures: one for the prospective against HIV transmission and that behavioral risk-reduction client and one for the female partner. Two informants cited a third strategies remain essential to safer sex strategies to protect type of brochure: in one case, it was for adolescents and their oneself and one’s sexual partners (e.g., correct and consistent guardians, and in another, a flyer for clients to read at VMMC condom use, reducing one’s number of partners). Messaging sites while waiting for the procedure. Other print materials may be intensified during campaign periods, especially to alert included posters—for use in service delivery locations to announce the population to dates and locations where teams will be the next date or ongoing availability of service—and flyers. One performing VMMC. program in South Africa had discovered that the cost of small Creating demand for VMMC among men over the age of 25 media can be prohibitive, depending on the size of the intended N has proven more difficult in some settings than attracting population. adolescents and younger men to VMMC services. In SMS, a form of text messaging, was mentioned by three key estimating future costs, it is essential to take into account the informants but none of the respondents included it in the additional costs of strategies tailored to reach this ‘‘harder to hypothetical allocation of funds across channels. The experience reach’’ segment of the population. in the Iringa Region of Tanzania, illustrated a potentially effective use of this medium. Although the Iringa Region is rural, many N Seasonal variability is also a factor. This type of variability clients and prospective clients own cell phones. At a very low cost, takes two forms: (1) natural, caused by locally occurring events the VMMC program has worked with the local phone provider to independent of VMMC programs (e.g., school holiday periods, make a series of text messages available to the public. If a person cooler temperatures), and (2) programmatically induced, texts ‘‘TOHARA’’ (‘‘circumcision’’ in Swahili), he/she receives resulting from efforts by governments to intensify the delivery two messages on the benefits of VMMC. Texting ‘‘WAPI’’ of services in hopes of stimulating demand during certain (‘‘where’’ in Swahili) sends a message on where and at what times periods. In some cases, the intensified programs are scheduled VMMC services will be available. For clients who are circumcised, to occur with naturally occurring events (e.g., school holidays). texting ‘‘BAADA’’ (‘‘after’’ in Swahili) triggers 8–10 messages over The Rapid Results Initiative in Kenya illustrates this case. a 6-wk period, including messages on wound care, healing, and N Collaboration between partners (government, technical assis- abstinence. At 6 wk, men receive a message that it is now tance agencies, and local community organizations) is essential permissible to have sex, but that they should use condoms and for ensuring both initial success and sustained effects of stick with one partner. These messages are free of charge to the VMMC demand creation efforts. Currently, the nature and user. The phone company reported that the VMMC program had extent of this collaboration varies by country (e.g., high in the largest number of users of any ‘‘company’’ that year. The key Kenya and Tanzania, lower in Zambia), but collaboration informant from South Africa referred to the ‘‘combination remains an important factor in most programs. approach’’ of alerting cell phone owners to information on a N Collaboration also extends to the divide between VMMC and website (which they then would access through their cell phone) traditional MC. A key informant based on South Africa shed and using the messaging service to deliver follow-up reminders for light on the importance of collaboration, although admitted it men who opt to be circumcised. He stated that the use of cell was still a work in progress. The informant relayed that there phone technology for the promotion of HIV prevention and was an existing partnership in the Eastern Cape where trained VMMC needs to be integrated into social mobilization activities medical professionals aided in traditional MC when medical through which phone numbers can be gathered, because sending emergencies necessitated it. Further communication between unsolicited SMS messages generally is illegal. the two sectors is underway, but according to the informant, other activities have taken priority. Key Considerations in Designing and Costing Successful N Rather than individual program branding, a strong national Demand Creation Programs VMMC symbol, such as a logo that links the media messages Key informants provided other useful insights for the design and to actual VMMC service delivery sites, would facilitate costing of successful demand creation programs based on their personal action by men who have been motivated to seek experiences with VMMC demand creation. services. N Interpersonal communication is essential for demand creation for VMMC (e.g., community health workers, community Challenges in Estimating the Cost of Demand Creation to mobilizers, small media [brochures]). Mass media can create Reach Universal Coverage awareness of VMMC, but interpersonal communication serves Given the lack of a standard package of VMMC demand as the catalyst to action. creation activities, country programs have opted to use different N Programs should take advantage of existing community channels to reach prospective clients, with a strong focus on organizations and partner programs, piggybacking the interpersonal communication strategies using mobilizers or peer VMMC content onto the existing outreach efforts of trusted educators. Table S1 outlines the types of mass and interpersonal
63 Costing Demand Creation for MMC Scale-Up channels used to date. The lack of a standard or recommended cost-effectiveness (i.e., the cost for demand creation per VMMC package of communication activities to serve as a starting point for performed). costing demand creation activities for VMMC is by no means Even when the ‘‘cost per unit of change’’ (that is, costs related to unique. In virtually every health domain, effective communication demand creation activities per VMMC performed) is estimated in programs aimed at creating demand are those tailored to the one population or setting, this estimate would not be generalizable specific behavior change or demand creation objective, taking into to other populations, given variations in the supply of and demand account context-specific factors. for VMMC services. In geographical areas with a high latent Even with a standard package of VMMC demand creation demand for VMMC (e.g., a community with many ‘‘early activities, costs would vary by country or program, based on adopters’’), it would be less costly to create demand for one factors such as geographical area to be covered, size of the VMMC than in an area with relatively low demand because of, for population within the target area, mix of communication channels example, cultural norms that previously have not included MC, or to be used, number of languages required to reach the majority of where there is high pre-existing circumcision prevalence. In short, the intended audience, readiness of men in a specific community any estimate on the cost of demand creation in the scale-up of to accept VMMC services (especially regarding the procedure VMMC will be imprecise by definition. Yet the compelling need being performed in a clinical setting), availability of in-country for cost estimates to assist in planning and budgeting justifies production facilities for broadcast or print materials, and cost of further pursuit of this question. media time for broadcast of radio and television spots during peak In sum, although the concept of a ‘‘standard package’’ for hours. VMMC demand creation is desirable to planners, it is not realistic Different regions of a country may have contrasting require- in practice. One can assemble a list of activities that serve as a ments for creating sufficient demand for VMMC services. In widely used ‘‘menu’’ of the most common communication Tanzania, for example, current MC prevalence based on self- activities. Yet the allocation of costs to these different items will reporting varies by setting, with urban areas estimated at 88% vary depending on a number of factors cited above. In fact, prevalence and rural areas at closer to 60% [17]. MC prevalence tailoring demand creation to the specific context increases its also varies by region, reaching a high of nearly 100% in Lindi and effectiveness and makes it more cost-efficient. a low of 21% in Shinyanga. As a result of this wide variation, USAID and UNAIDS currently are working with the government Illustrative Case Study with Cost Data on Demand of Tanzania and with Jhpiego to evaluate potential demand creation resource requirements for significant scale-up of VMMC Creation for VMMC in Tanzania services in the eight regions of the country prioritized by the To date, there has been very little systematic costing of demand Tanzanian Ministry of Health. This exercise involves evaluating creation activities. Although programs keep financial records of the scale-up required for VMMC services in each region, expenses for demand creation activities as part of their overall identifying the most effective communication strategies for those VMMC projects or programs, few have extracted these data for regions, and costing the demand creation strategy. purposes of analyzing the distribution across different channels. The cost of demand creation varies both by the comprehen- One exception is the Jhpiego program in Tanzania. In this section, siveness of the communication program and the extent of we present a description of the context for the VMMC program adherence to standards of best practice for demand creation. and the demand creation activities conducted in two regions— Programs that use multiple and reinforcing channels have stronger Iringa and Njombe—during an 8-wk period from June to August effects than single channel programs [18], and those that run for a 2011. longer duration show greater effects than shorter ones [19]. The Demand creation in the Iringa Region—Dondosha norms of best practice dictate the use of formative research and mkonosweta! In February of 2011, authorities in the Iringa pretesting of materials before production. Costing studies often Region of Tanzania, with support from the USAID-funded exclude research costs and focus instead on programmatic costs, Maternal and Child Health Integrated Program (MCHIP), but formative research and pretesting are essential parts of the through Jhpiego—an affiliate of Johns Hopkins University— design and implementation of sound communication programs, began preparations for a June/July 2011 VMMC campaign regardless of the objective. Comprehensive communication designed to serve 20,000 clients in 8 wk at 21 sites. Concerned that programming goes well beyond messages directed toward the the easy-to-reach clients had been served during the previous primary audience of adolescent and adult men. For example, campaign and normal service delivery, the team endeavored to Table S2 presents an overview of a demand creation strategy for design a multi-channel communications initiative to launch and Nyanza Province, Kenya, produced by the USAID-funded sustain the campaign. Focus group discussions were held to better program C-Change (managed by the Academy for Educational understand male and female opinions about VMMC, garner their Development) in 2010. A comprehensive program functions at levels of knowledge about the purpose of VMMC, and gauge their multiple levels, reaching multiple audiences in an effort to achieve knowledge of service locations. This formative assessment was used multiple outcomes. to design a communications campaign using the local slang term Perhaps most challenging, public health officials at all levels lack for the foreskin. ‘‘Dondosha mkonosweta! Kitendo rahisi, sasa ni bure!’’ hard data on the cost of ‘‘convincing one individual to adopt one (‘‘Take off your sweater sleeve! Easy to do, now free!’’) was tested public health behavior,’’ and this is certainly the case for and approved with the target audience and became the campaign ‘‘triggering the decision’’ to seek MC services. As of 2006, there slogan. Materials designed with the new slogan included print were only 35 studies in the published literature on the cost- media, mass media, and text messaging. effectiveness of behavior change communication for health Print media. Print media included the following: billboards interventions in low- and middle-income countries [20]. Not targeting men and women placed in key markets and road stops; surprisingly, the reported costs differed markedly by type of health posters promoting the campaign and service sites (paper and behavior. In terms of demand creation for VMMC, to our plastic—for longer durability in the wet climate) (see Text S2); knowledge no studies have been published to date on attempts to flyers advertising the campaign and service sites; pop-up banners evaluate the effectiveness of VMMC communication, let alone its to use at outreach activities, the launch event, and for the exteriors
64 Costing Demand Creation for MMC Scale-Up of service sites; and informational brochures for men, women, and channels. The final column in Table S3 presents a more detailed adolescents and their guardians. percentage breakdown by specific channel of communication. The Mass media. Radio is the primary mass media in the Iringa largest demand creation expenses (40% of the total) were for the Region. Therefore the campaign produced three radio spots (one production of pamphlets, flyers, posters, and other supporting targeting men, one targeting women, and one with general print materials for clients/parents. Following in second place was information) and aired them in heavy rotation (at least eight times drama and street theater (27%). Production cost and rental space per day) on a regional radio station and community stations in for billboards represented 24% of the total. Other categories of facility catchment areas. Broadcasts of these spots began one week demand creation expense included the production and broadcast before the campaign launch and ran until 2 wk before the end of of materials for radio (5%), cell phone messaging (2%), honoraria/ the campaign. Additionally, radio talk shows featuring regional per diem payments for mobilizers and peer educators (1%), and officials aired during the campaign period. training for mobilizers and peer educators (1%). Text messaging. Print and mass media materials were used In presenting data on this program, we must warn against any to promote a text messaging service, as described in the results interpretation that ‘‘one size fits all’’ in terms of demand creation. from the key informant interviews above. If one considers even a handful of the more active VMMC Community. The team again relied on HIV prevention programs, this diversity becomes readily apparent. Orange Farm partners (also funded by USAID) to promote VMMC through in South Africa represents one VMMC site with a community- their scheduled interpersonal communication activities. These wide rollout and a gradually expanding radius of influence; partners received print materials and training from MCHIP on VMMC programs with different models are evolving in other VMMC messages. The campaign teams conducted outreach visits parts of South Africa. The Kenya program has concentrated to community leaders and groups, primary and secondary schools, largely on a single province, Nyanza, but with high intensity of and churches in the facility catchment areas. An experiential activity. Tanzania has focused its programmatic efforts on several media agency also promoted the campaign in facility communities provinces. Swaziland is unique, with an accelerated scale-up to through road shows, community events, and football games. An reach a high percentage of males in a traditionally non- ‘‘emergency team’’ was designed to be quickly deployed to sites circumcising society. Thus, these data should be considered as experiencing ebbs in client flow in order to create more demand. descriptive of one program only: the demand creation efforts Media campaign results. Although no formal evaluation implemented by MCHIP in Tanzania. has been completed of this demand creation activity, the June/July 2011 VMMC campaign in the Iringa Region served 31,046 Discussion clients, more than 150% of the intended number. The campaign reached more than 150,000 people during experiential media The key informant interviews conducted in this study revealed a activities alone. The mix of print and mass media, coupled with wide range of views on the elements of a standard package for community-based activities, helped bring clients in large numbers VMMC demand creation, as reflected by the diversity of answers to the service delivery sites. on resource allocation across channels. A first step toward Jhpiego tracked the costs for demand creation activities in this estimating future costs would be to collect actual cost data to campaign. The data in Table S3 indicate that the expenses for learn how countries have allocated their demand creation demand creation across channels came to US$266,950. As shown resources to date. Clearly, data should be collected from all in Figure 1, 70% of the funding supported mass and small media, organizations involved in VMMC demand creation in a given 29% was used for community mobilization, and 1% went to other country to learn how countries allocate their demand creation
Figure 1. Percentage of budget for demand creation costs for Jhpiego VMMC campaigns in Iringa and Njombe Regions, Tanzania. doi:10.1371/journal.pone.0027562.g001
65 Costing Demand Creation for MMC Scale-Up resources between organizations that would likely play comple- prioritize for VMMC programs. (3) Access the demand creation mentary roles. strategy for each country, working with local communication Second, the primary channels for effective demand creation for experts to develop such a strategy if one has not been developed VMMC include targeted radio and interpersonal communication and adapting elements from other programs with well-defined or outreach, accompanied by print materials tailored for both men strategies (e.g., the Nyanza Province VMMC communication and women. Television, billboards, and other media play minor guide [21]). (4) Assess the current rate of VMMC scale-up, the supporting roles in most VMMC scale-up settings. Depending on level of coverage already reached, and the coverage objective, so the costs charged by local phone companies, text messaging can as to estimate the additional work needed in a given country. (5) represent a low-cost, promising strategy that deserves further Collect data on the levels of expenditures for VMMC demand exploration. creation from relevant organizations for two consecutive years (see Third, these common elements do not lend themselves to a Text S3 for sample format). (6) Define a 5-y demand creation standardized package for demand creation that is applicable across program to reach 80% of uncircumcised males, with specific countries. Rather, each country should determine the communi- outputs by region for each year, taking into consideration the cation strategies and elements for successful demand creation, urban/rural distribution of men to be reached, the degree of from which the costing estimates will derive. To complicate the difficulty of reaching different populations, the number of local matter further, different elements may be used in different parts of languages into which materials would need to be translated, and the program (e.g., billboards in urban but not rural areas). The related factors. (7) Budget the expenses in these plans; compare mantra that communication programs must be tailored to the local existing levels of expenditure on demand creation to those needs of a specific country or regional context holds for VMMC. elaborated in the plan and adjust as necessary in light of current Fourth, costing demand creation presents a greater methodo- levels of expenditure. logical challenge than costing service delivery, in part because This approach represents a significant data collection challenge, there is no straight line ‘‘dose–response’’ relationship (i.e., the even for countries that have cost data available on VMMC delivery of a given amount of communication does not yield a activities. For those that have yet to begin the start-up in a major predictable uptake in VMMC services). For example, it is possible way, it may be necessary to use the costs of a ‘‘peer country’’ to to estimate the service delivery costs to provide 1,000 VMMC impute cost estimates. procedures in a given location with fair accuracy. However, one Three adjustments to such cost estimates are required. First, it is cannot predict with the same accuracy the increase in service essential to budget the cost of designing a systematic strategy for uptake that will be generated by a given level of demand creation demand creation, rather than simply estimating production and activity. distribution costs. Most countries with strong demand creation Fifth, it is unclear how demand creation costs may need to vary efforts have benefited from external technical assistance for over time. One could argue that demand creation for VMMC program design, and such costs must be figured into the services would require a large initial investment to create calculations. Second, it is important to introduce an adjustment awareness, which then would be followed by smaller investments factor for start-up, on the assumption that a country’s program once a critical mass of knowledge has been created. Alternatively, must achieve some momentum before demand creation activities one could argue that, in the early stages of scale-up, there may be begin to increase service utilization. It stands to reason that certain latent demand for VMMC and that only over time, once this portions of the population will be more likely to participate in the demand has been met, would there be a need for demand creation VMMC program from the outset and less likely to need among those who need to be convinced of the value of VMMC. encouragement from demand creation outreach. Third, it is This study has several limitations. First, the number of important to estimate the added cost of communication programs informants was small, hence the word ‘‘exploring’’ in the title of intended for ‘‘harder-to-reach’’ populations, such as men over this article. Second, the key informant interviews did not include a 25 y of age. Operational research on the costs of effective demand wider group of participants such as government officials working creation among these men in comparison to those 15–25 y of age in VMMC programs in the prioritized countries. These individ- could inform the field usefully. uals are central to the strategic direction of demand generation In estimating costs, assumptions should be made explicit. programming and implementation in country. However, at the Formative research and pretesting of materials should always form time of this data collection, they did not manage the budgets for part of essential demand creation costs in a well-designed demand creation activities, making it difficult for them to answer communication program. Although monitoring and evaluation questions regarding the cost of different elements of the campaign. of demand creation activities inform iterative adjustments to create For this reason, the researchers selected persons known to manage and sustain strong programs, these costs should be tabulated as large United States President’s Emergency Plan for AIDS Relief– part of the overall monitoring and evaluation budget, rather than funded VMMC communication budgets. In future studies of this in the communication budget. Further, estimated costs for demand nature, it will be important to increase the number of key creation exclude costs related to client counseling. Although informant interviews, include a larger number of the 13 countries, counseling contributes to confirming an individual decision to and ensure the participation of a wider range of participants, move forward with VMMC, it can be argued that by the time including government officials responsible for the VMMC scale- clients are in the clinic to receive that counseling, the demand up. already has been created. Based on these findings and caveats, our proposed approach We base this article on the premise that demand creation is requires country-specific analysis, with the data for each country essential for the scale-up of VMMC in eastern and southern aggregated to reach a total across the 13 countries working to scale Africa. This assumption stems from the fact that almost every up VMMC. The steps in the proposed approach are as follows. (1) major public health intervention designed to bring about behavior Base the costing exercise on existing estimates of the number of change (e.g., condom promotion to prevent HIV, use of bed nets adolescent and adult males who would need to be circumcised to to reduce malaria, adoption of contraception to space or limit achieve the saturation level in each country (see Box 1). (2) Identify births) has both supply and demand components. There is a the provinces or populations with low levels of MC prevalence to growing literature on the effectiveness of communication cam-
66 Costing Demand Creation for MMC Scale-Up
the promotion of VMMC, these estimates could be rendered Box 1. Recommendations by C-Change (Academy for useless rather than informing decision making on program Educational Development; 2010) on the design of a planning. demand creation program for Nyanza Province, Kenya. Despite these challenges, we believe that the 13 priority At the individual level: Continue provider support and countries in eastern and southern Africa can arrive at a reasonable client education, with a focus on promoting VMMC within estimate of the costs of demand creation for their VMMC services the context of broader HIV prevention. by following the steps outlined in this paper. Specific outcomes for this level of programming include: N Circumcised men practice HIV prevention. Supporting Information N Uncircumcised men go for VMMC, get tested, heal safely, Table S1 Types of communication channels that have champion VMMC, and then practice HIV prevention. been used to promote VMMC in eastern and southern N Women encourage their partners to go for VMMC and Africa. are not exposed to added risk by the newly circumcised (DOCX) males. Table S2 Estimated number of additional VMMCs At the family, peer, and community level: Mobilize needed to reach saturation levels in 13 eastern and the community to demand VMMC and incorporate VMMC southern African countries. See also [22]. within broader healthy social norms and attitudes relating (DOCX) to HIV prevention and gender. Specific outcomes for this level of programming include: Table S3 Demand creation costs for the Jhpiego campaign in Iringa and Njombe Regions, June–August N Families and friends of potential VMMC clients are 2011. informed and supportive of VMMC. (DOCX) N Service providers communicate effectively about VMMC. Text S1 Discussion guide for the key informant inter- N Peer educators mobilize effectively around VMMC. views on demand creation for male circumcision. (DOC) At the environmental level: Enhance political support for VMMC, engage key institutions (education, business) Text S2 Sample poster from print media portion of and constituency-based networks to support VMMC Iringa Region VMMC campaign. mobilization efforts, and improve VMMC media coverage. (TIFF) Specific outcomes for this level of programming include: Text S3 Instructions for providing costs data for male N Role models support VMMC. circumcision demand creation. N The media report accurately about VMMC. (DOC) N Elders of the largest ethnic group in Nyanza Province, the Luo, are supportive of VMMC. Acknowledgments N VMMC is promoted by Kenyan churches. The work undertaken for this article was supported by the Research to N VMMC is supported at Kenyan workplaces. Prevention project, funded by USAID, Project SEARCH, Task Order No.2, under Contract No. GHH-I-00-07-00032-00, beginning September 30, 2008, and by the President’s Emergency Plan for AIDS Relief. paigns in different areas of public health (e.g., meta-analysis Research to Prevention is led by the Johns Hopkins Center for Global Health and managed by the Johns Hopkins Bloomberg School of Public performed by Snyder [19]). To date, we have no rigorous Health Center for Communication Programs. Tulane University School of evaluation of demand creation for VMMC. Yet as programs Public Health and Tropical Medicine serves as a subcontractor on advance, we anticipate that evaluation data will be forthcoming on Research to Prevention. We acknowledge Sekai Chideya for her the effectiveness of VMMC demand creation interventions on contribution to the data analysis, writing, and review of this paper. increasing VMMC uptake. Finally, costing demand creation for VMMC assumes political Author Contributions will on the part of governments in the 13 countries to mobilize their own domestic resources and/or make use of available funds Conceived and designed the methods: JB EN CH HM. Collected the data: from international development partners to support VMMC scale- JB SM SF EN HM. Analyzed the data: JB EN SF SM HM CH. Contributed reagents/materials/analysis tools: JB EN SF SM HM CH. up, including effective demand creation programs. Where this is Wrote the paper: JB EN SF SM HM CH. not the case and, even more markedly, where there is opposition to
References 1. Auvert B, Taljaard D, Lagarde E, Sobngwi-Tambekou J, Sitta R, et al. (2005) programme implications. WHO/UNAIDS technical consultation. Male cir- Randomized, controlled intervention trial of male circumcision for reduction of cumcision and HIV prevention research implications for policy and program- HIV infection risk: The ANRS 1265 Trial. PLoS Med 2: e298. doi:10.1371/ ming. Geneva: World Health Organization. journal.pmed.0020298. 5. United States Agency for International Development Health Policy Initiative 2. Bailey RC, Moses S, Parker CB, Agot K, Maclean I, et al. (2007) Male (2009) The potential cost and impact of expanding male circumcision in 14 circumcision for HIV prevention in young men in Kisumu, Kenya: A African countries. Washington (District of Columbia): Futures Group. randomised controlled trial. Lancet 369: 643–656. 6. Njeuhmeli E, Reed J, Bollinger L, Forsythe S, Castor D, et al. (2010) Cost and 3. Gray RH, Kigozi G, Serwadda D, Makumbi F, Watya S, et al. (2007) Male impacts of expanding male circumcision services in Eastern and Southern Africa circumcision for HIV prevention in men in Rakai, Uganda: A randomised trial. [abstract]. XVIII International AIDS Conference; Vienna, Austria; 18–23 July Lancet 369: 657–666. 2010. 4. World Health Organization, Joint United Nations Programme on HIV/AIDS 7. Fieno JV (2008) Costing adult male circumcision in high HIV prevalence, low (2007) New data on male circumcision and HIV prevention: policy and circumcision rate countries. AIDS Care 20: 515–520.
67 Costing Demand Creation for MMC Scale-Up
8. Auvert B, Marseille E, Korenromp EL, Lloyd-Smith J, Sitta R, et al. (2008) 15. Westercamp N, Bailey RC (2007) Acceptability of male circumcision for Estimating the resources needed and savings anticipated from roll-out of adult prevention of HIV/AIDS in sub-Saharan Africa: a review. AIDS Behav 11: male circumcision in sub-Saharan Africa. PLoS ONE 3: e2679. doi:10.1371/ 341–355. journal.pone.0002679. 16. Westercamp M, Bailey RC, Bukusi EA, Montandon M, Kwena Z, et al. (2010) 9. Uthman OA, Popoola TA, Uthman MM, Aremu O (2010) Economic Male circumcision in the general population of Kisumu, Kenya: beliefs about evaluations of adult male circumcision for prevention of heterosexual acquisition protection, risk behaviors, HIV, and STIs. PLoS ONE 5: e15552. doi:10.1371/ of HIV in men in sub-Saharan Africa: a systematic review. PLoS ONE 5: e9628. journal.pone.0015552. doi:10.1371/journal.pone.0009628. 17. Tanzania Commission for AIDS, Zanzibar AIDS Commission, National Bureau 10. UNAIDS/WHO/SACEMA Expert Group on Modelling the Impact and Cost of Statistics, Office of the Chief Government Statistician, Macro International of Male Circumcision for HIV Prevention (2009) Male circumcision for HIV (2008) Tanzania HIV/AIDS and malaria indicator survey 2007–08. Dar es prevention in high HIV prevalence settings: what can mathematical modelling Salaam (Tanzania): Tanzania Commission for AIDS. contribute to informed decision making? PLoS Med 6: e1000109. doi:10.1371/ 18. Salem RM, Bernstein J, Sullivan TM, Lande R (2008) Communication for journal.pmed.1000109. better health. Population reports, series J, no. 56 Baltimore: INFO Project, Johns 11. Kebaabetswe P, Lockman S, Mogwe S, Mandevu R, Thior I, et al. (2003) Male Hopkins Bloomberg School of Public Health. circumcision: an acceptable strategy for HIV prevention in Botswana. Sex 19. Snyder L (2007) Health communication campaigns and their impact on behavior. J Nutr Educ Behav 39: S32–S40. Transm Infect 79: 214–219. 20. Hutchinson P, Wheeler J (2006) The cost-effectiveness of health communication 12. Lagarde E, Dirk T, Puren A, Reathe RT, Bertran A (2003) Acceptability of male programs: what do we know? J Health Commun 11: 7–45. circumcision as a tool for preventing HIV infection in a highly infected 21. Kenya Ministry of Public Health and Sanitation (2010) Voluntary medical male community in South Africa. AIDS 17: 89–95. circumcision (VMMC) communication guide for Nyanza Province. Available: 13. Mattson CL, Bailey RC, Muga R, Poulussen R, Onyango T (2005) Acceptability http://www.c-hubonline.org/7815/17/NYANZA_COMMUNICATION_ of male circumcision and predictors of circumcision preference among men and GUIDE.pdf. Accessed 31 October 2011. women in Nyanza Province, Kenya. AIDS Care 17: 182–194. 22. Njeuhmeli E, Forsythe S, Reed J, Opuni M, Bollinger L, et al. (2011) Voluntary 14. Scott BE, Weiss HA, Viljoen JI (2005) The acceptability of male circumcision as medical male circumcision: modeling the impact and cost of expanding male an HIV intervention among a rural Zulu population, KwaZulu-Natal, South circumcision for HIV prevention in eastern and southern Africa. PLoS Med 8: Africa. AIDS Care 17: 304–313. e1001132. doi:10.1371/journal.pmed.1001132.
68 Review Voluntary Medical Male Circumcision: Strategies for Meeting the Human Resource Needs of Scale-Up in Southern and Eastern Africa
Kelly Curran1,2*, Emmanuel Njeuhmeli3, Andrew Mirelman2, Kim Dickson4, Tigistu Adamu1, Peter Cherutich5, Hally Mahler6, Bennett Fimbo7, Thembisile Khumalo Mavuso8, Jennifer Albertini3, Laura Fitzgerald9, Naomi Bock10, Jason Reed10, Delivette Castor3, David Stanton3 1 Jhpiego, Baltimore, Maryland, United States of America, 2 International Health Department, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, United States of America, 3 United States Agency for International Development, Washington, District of Columbia, United States of America, 4 World Health Organization, Geneva, Switzerland, 5 National AIDS and STI Control Programme, Ministry of Health, Nairobi, Kenya, 6 Jhpiego/Tanzania, Dar es Salaam, Tanzania, 7 Ministry of Health, Dar es Salaam, Tanzania, 8 Ministry of Health, Mbabane, Swaziland, 9 Jhpiego/Swaziland, Mbabane, Swaziland, 10 Centers for Disease Control and Prevention, Atlanta, Georgia, United States of America
and counseling (HTC), screening for sexually transmitted infections, Abstract: Voluntary medical male circumcision (VMMC) condom promotion and postoperative care. The United States reduces female-to-male HIV transmission by approximate- President’s Emergency Plan for AIDS Relief (PEPFAR) currently ly 60%; modeling suggests that scaling up VMMC to 80% provides funding and technical support for implementing VMMC of men 15- to 49-years-old within five years would avert services in the 13 priority countries plus the Gambella National over 3.3 million new HIV infections in 14 high priority Regional State in Ethiopia. countries/regions in southern and eastern Africa by 2025 Mathematical modeling studies suggest that to reach 80% and would require 20.33 million circumcisions. However, VMMC coverage in the priority countries within five years would the shortage of health professionals in these countries entail performing 20.33 million circumcisions between 2011 and must be addressed to reach these proposed coverage 2015 (Figure 1); sustaining 80% coverage thereafter (universal levels. To identify human resource approaches that are being used to improve VMMC volume and efficiency, we coverage) would require an additional 8.42 million circumcisions looked at previous literature and conducted a program between 2016 and 2025 [6]. Modeling also suggests that review. We identified surgical efficiencies, non-surgical approximately 3.36 million new HIV infections and 386,000 AIDS efficiencies, task shifting, task sharing, temporary rede- deaths would be averted through 2025 at an estimated total cost of ployment of public sector staff during VMMC campaign US$2,000,000,000; net savings would be US$16,550,000,000 [6]. periods, expansion of the health workforce through recruitment of unemployed, recently retired, newly Citation: Curran K, Njeuhmeli E, Mirelman A, Dickson K, Adamu T, et al. graduating, or on-leave health care workers, and the use (2011) Voluntary Medical Male Circumcision: Strategies for Meeting the Human of volunteer medical staff from other countries as Resource Needs of Scale-Up in Southern and Eastern Africa. PLoS Med 8(11): approaches that address human resource constraints. e1001129. doi:10.1371/journal.pmed.1001129 Case studies from Kenya, Tanzania, and Swaziland Academic Editor: Stephanie L. Sansom, Centers for Disease Control and illustrate several innovative responses to human resource Prevention, United States of America challenges. Although the shortage of skilled personnel Published November 29, 2011 remains a major challenge to the rapid scale-up of VMMC This is an open-access article, free of all copyright, and may be freely reproduced, in the 14 African priority countries/regions, health distributed, transmitted, modified, built upon, or otherwise used by anyone for programs throughout the region may be able to replicate any lawful purpose. The work is made available under the Creative Commons CC0 or adapt these approaches to scale up VMMC for public public domain dedication. health impact. Funding: The Tanzania and Swaziland programs described in this paper and the program review itself were funded by the President’s Emergency Plan for AIDS Relief (PEPFAR) through USAID and implemented by the Maternal and Child Health Integrated Program (MCHIP, award # GHS-A-00-08-00002-000) managed by Jhpiego—an affiliate of Johns Hopkins University. The Kenya program is Introduction primarily funded by the Bill & Melinda Gates Foundation and PEPFAR through USAID and the US Centers for Disease Control and Prevention via various Observational and experimental studies have shown that volunteer implementing partners. Tanzania’s Ministry of Health and Social Welfare Male medical male circumcision (VMMC) reduces HIV acquisition among Circumcision Technical Working Group, the National AIDS Control Program, the Iringa Regional Health Authority, and the health authorities of Iringa Municipal, heterosexual men by approximately 60% [1–4]. In 2007, after the Iringa District, and Mufindi District played important roles in guiding this work. completion of three randomized controlled trials on VMMC, the Competing Interests: The authors have declared that no competing interests World Health Organization (WHO) and the Joint United Nations exist. Programme on HIV/AIDS (UNAIDS) recommended that VMMC Abbreviations: HTC, HIV testing and counseling; HRH, human resources for should be part of comprehensive HIV prevention programming in health; MC, male circumcision; PEPFAR, United States President’s Emergency Plan regions with a generalized HIV epidemic and a relatively low level of for AIDS Relief; UNAIDS, Joint United Nations Programme on HIV/AIDS; VMMC, male circumcision (MC). They also prioritized 13 countries in voluntary medical male circumcision; WHO, World Health Organization southern and eastern Africa (Table 1) for VMMC scale-up [2–5]. In * E-mail: [email protected] addition to surgical intervention, WHO recommended that VMMC Provenance: Submitted as part of a sponsored Collection; externally peer service packages should include HIV and MC education, HIV testing reviewed.
69 Summary Points sub-Saharan Africa has 25% of the world’s disease burden and 3% of the world’s health workforce [7,8]. As a consequence, sub- N Scaling up voluntary medical male circumcision (VMMC) Saharan Africa needs to triple its health workforce in order to meet could avert millions of HIV infections in southern and existing health needs [9]. Moreover, of the 57 countries globally eastern Africa, but shortages of health professionals are with HRH crises, 36 are on the African continent [10]. likely to limit progress. More specifically, the southern and eastern African countries N Potential responses to this human resource challenge prioritized for VMMC scale-up for HIV prevention have low include task shifting, task sharing, temporary redeploy- physician and nurse/midwife densities (Table 1) and very limited ment of public sector staff during VMMC campaigns, numbers of surgical specialists. For example, in 2008, there were expansion of the health workforce through recruitment only 75 surgeons and ten physician anesthetists serving Uganda’s of unemployed, recently retired, newly graduated, or on- population of 27 million [11]. These countries also have heavy leave health care workers, and the use of foreign burdens of tuberculosis and other diseases that stretch their health volunteer medical staff. systems [12] and that directly impact on their limited health N Approaches to solving the human resource challenge workforce [13]. Finally, out-migration to higher-resource settings associated with VMMC scale-up that have already been and economic limitations within the public sector further deplete implemented include the following: moving public HRH in these countries [14,15]. sector clinicians to high-volume VMMC sites during Because of these HRH concerns, health ministries in the priority campaigns (Tanzania), empowering nurses to conduct countries for VMMC scale-up have highlighted the possible VMMC surgery (Kenya), and identifying untapped disruption to other key HIV, primary care, and surgical services as reserves of qualified nurses (Swaziland). major barriers to planning for or implementing the scale-up of N These approaches provide models for other countries to VMMC services. In this review, we describe three approaches to replicate and adapt. optimizing HRH and minimizing disruptions to existing health N Importantly, the effect of VMMC scale-up on other health services that have already been used in the implementation and services is not known and must be investigated in future scale-up of VMMC for public health impact. studies. How Are HRH Limitations Being Addressed? Unfortunately, the number of clinicians, counselors, and To determine the status of HRH in the southern and eastern support staff needed to achieve and sustain universal coverage of African VMMC priority settings and to identify innovations in VMMC is expected to exceed the available human resources for HRH that address resource challenges in VMMC, we examined health (HRH) in many African countries. WHO estimates that policy and program reports from the 13 priority countries identified by WHO plus the Gambella National Regional State in Ethiopia to identify programs that have augmented their health Table 1. Density of physicians and nurses/midwives in VMMC workforce to increase efficiencies. (The Gambella Region was priority countries. included because PEPFAR supports a VMMC program there, where MC prevalence is low and HIV prevalence is three times the national average.) Specifically, we looked for evidence within Physician national and regional programs of the use of staff to improve Density (per Nurse/Midwife 1,000 Density (per 1,000 volume and efficiency, of a demonstrated capacity to scale up Country Individuals) Individuals) Year services, and of the institution of measures for optimizing quality and safety. We also looked for evidence that these HRH Botswana 0.40 2.65 2004 approaches had been implemented through the public sector Ethiopiaa 0.02 0.24 2007 because this is a prerequisite for achieving universal coverage. Kenyaa 0.14 1.18 2002 VMMC programs based in Kenya, Tanzania, and Swaziland Lesotho 0.05 0.62 2003 emerged from this analysis as programs that were using innovative HRH approaches, so we examined these programs more closely as Malawi 0.02 0.28 2008 case studies. Mozambique 0.03 0.31 2006 From the three selected VMMC program case studies, we Namibia 0.30 3.06 2004 identified seven main policy and programmatic approaches being Rwanda 0.02 0.45 2005 applied in southern and eastern Africa to address the HRH needs South Africa 0.77 4.08 2004 of VMMC scale-up while minimizing disruptions to other key Swaziland 0.16 6.30 2004 health services: surgical efficiencies, non-surgical efficiencies, and five types of human resource efficiencies (task shifting, task sharing, Tanzania 0.01 0.24 2006 temporary redeployment of public sector health care workers Uganda 0.12 1.31 2005 during VMMC campaign periods, targeted recruitment of Zambia 0.06 0.71 2006 unemployed, recently retired, newly graduating, and on-leave Zimbabwe 0.16 0.72 2004 health care workers, and recruitment of volunteer health care Meanb 0.16 1.58 — workers). United States 2.56 9.37 2000 Surgical Efficiencies Sources: [32,33]. In our review of VMMC programs, we identified surgical a VMMC programming focuses on the Gambella Region of Ethiopia and the technique, team composition, and use of prepackaged surgical kits Nyanza Province of Kenya, but these data refer to health care densities over the whole country in both cases. as three surgical efficiencies that decrease provider downtime and bMean density of African-country-specific data is from different years. increase the number of VMMCs that can be performed by a doi:10.1371/journal.pmed.1001129.t001 surgical team while maintaining high quality.
70 Figure 1. Number of circumcisions among men aged 15 to 49 years needed to reach 80% coverage in each of 14 priority countries/ regions within five years. For Ethiopia (Gambella Region) and Kenya (Nyanza Province), only one region or province with low circumcision rates and high HIV prevalence is included, as most men in these countries are already circumcised. doi:10.1371/journal.pmed.1001129.g001
All three techniques for adult circumcision—dorsal slit, forceps- removal, hemostasis, and mattress sutures) [17]. This model is staff guided, and sleeve resection—have similar safety profiles and are and resource intensive and is best suited for high-volume sites. included in the WHO/UNAIDS/Jhpiego surgical manual [16]. The Orange Farm team has also pioneered the use of However, the forceps-guided procedure is, on average, two prepackaged, fully disposable VMMC kits containing all of the minutes, 45 seconds, faster than the dorsal slit procedure, and instruments and consumables needed for one procedure. This seven minutes, 40 seconds, faster than sleeve resection [17]. innovation saves pre-procedure preparation time, and reduces Although the forceps-guided procedure can be less esthetically time delays due to missing supplies or equipment. An assessment pleasing initially, after complete wound healing all the methods of minor surgical capacity in Mozambique, including the capacity produce similar esthetic results [16]. Given the number of to perform VMMC, showed that surgical supplies and equipment VMMCs that need to be completed for public health impact, were often out of stock, and that this seriously undermined surgical even relatively small time savings per procedure can have productivity. Problems with water and electricity also impeded significant effects on the number of procedures completed [6,18]. instrument processing in this setting [19]. The price of Conventionally, adult and adolescent VMMC is provided by prepackaged surgical kits, which are particularly useful for high- one doctor working with one assistant in an operating theater or volume sites, mobile outreach services, and other campaign-type procedure room. Using this model, a doctor–nurse surgical team settings, is a function of economies of scale. The pooled can provide eight to ten VMMCs per day (Figure 2A). This procurement of kits has already resulted in a reduction in market conventional surgical model is inherently inefficient because with price from US$23 to US$11 per kit. Prepackaged kits also reduce only one bed in use the doctor has ‘‘downtime’’ between clients. the number of stock-outs and eliminate problems with instrument Time-and-motion studies conducted by the Bophelo Pele clinical processing due to electricity outages. In program settings where team in Orange Farm, South Africa, showed that physician reusable instruments continue to be used, pre-bundled packages downtime can be almost eliminated by expanding the number of can save time [20]. surgical bays (Figure 2B) and delegating many critical surgical tasks (including injection of local anesthesia and placement of Non-Surgical Efficiencies simple interrupted sutures) to nurses (task sharing), thereby A steady demand for VMMC services among males in the target allowing doctors to focus on the highest-level skill steps (foreskin age group is a necessary component for a successful program.
71 and efficiency of VMMC services and can be adapted to the individual circumstances of a given health situation and setting. Task shifting—the delegation of surgical steps to a trained non- physician clinician such as a nurse or clinical officer—can greatly expand the size of the workforce available to provide surgical services. Specifically, VMMC counseling and HTC can be delegated to lay counselors, and clinical officers and nurses, who are in greater supply in the workforce (Table 1), can replace doctors in all the steps in VMMC surgery. In task sharing, although some steps/procedures are delegated to a non-physician clinician such as a nurse, the highest-level skill steps (e.g., hemostasis) remain the physician’s responsibility. Task sharing works well when combined with approaches to improve the layout and client flow of surgical services (e.g., the use of multiple beds). The expansion of the role of nurses in VMMC programs through both task shifting and task sharing has greatly improved program efficiency [22] and has important implications for countries at various stages of VMMC scale-up. Another HRH efficiency is temporary redeployment of public sector staff involved in other aspects of health care to VMMC services during campaigns—redeployment is an approach that is commonly used during immunization campaigns. The use of a campaign model to provide high-volume VMMC services during periods of high demand (cold season, agriculturally slow periods, and school holidays) proved an effective strategy in Tanzania [21]. Targeted recruitment—the training and engagement of health care workers who are not currently active in the workforce such as part-time and unemployed workers, and recently retired or newly graduated nurses—is a good option for VMMC programs seeking long-term personnel. Objective quantification of existing HRH resources can identify some of these untapped HRH capacities, even in countries with documented HRH crises (see the Swaziland case study for more details). In addition, targeted short-term recruitment of on-leave nurses can be useful during campaigns. Finally, the recruitment of volunteer health care workers from other countries during focused VMMC campaigns can temporar- ily increase the size of the health workforce. To date, Swaziland is the only country in the region that has recruited and deployed volunteer health care providers (specifically doctors) for VMMC services. This approach is a feasible and acceptable option in a country where the nursing scope of practice does not allow task shifting.
Figure 2. Conventional and modified surgical approaches to Tanzania Case Study: Non-Surgical Efficiencies Optimize MC provision. (A) The conventional approach to MC provision the Use of a Public Sector Health Workforce Deployed requires 20–30 minutes per procedure and allows only two surgeries per hour (ten surgeries per day) because of the doctors ‘‘downtime’’ during Campaigns between procedures. (B) With the modified surgical approach to MC The Iringa Region has the highest adult HIV prevalence in provision (MOVE), each procedure still takes 20–30 minutes, but task Tanzania (15.7% compared to 5.7% nationwide) [23] and one of sharing reduces the doctor’s time to 6–8 minutes per procedure, which the lowest MC prevalence rates in the country (29%) [23]. The allows eight surgeries per hour (40 surgeries per day). Tanzanian Ministry of Health, with PEPFAR support, began doi:10.1371/journal.pmed.1001129.g002 implementing VMMC in the Iringa Region in June 2010 in a six- week campaign during which 10,352 adult and adolescent men Insufficient demand for supply is inefficient for the providers. were circumcised across five sites [21]. Some HTC was done in the Conversely, demand that exceeds services can create bottlenecks community, most providers came from the public sector, and task that impact program quality [21]. Supply and demand can be better shifting allowed nurses and clinical officers to perform VMMC matched by capturing information in real time on the number of surgeries. Several additional non-surgical efficiencies were also clients, adverse events, and HIV testing uptake. Also, moving HTC implemented to match supply with demand for VMMC services. and some other services (e.g., postoperative review) into the On the demand side, June and July were selected for the community during high-volume periods can improve client flow campaign because this is a slow period on tea plantations, and optimize the use of the health workforce [21]. coincides with school holidays, and is the cool season, which the local population believes is associated with faster wound healing Human Resource Efficiencies [24]. The program ‘‘harnessed’’ this period of naturally higher Implementation of the five types of HRH efficiencies that we demand by the timely expansion of VMMC services while identified in our program review can improve both the volume ensuring adequate client demand through multiple reinforcing
72 messages to the population (e.g., radio advertisements, posters, of VMMC providers. When the VMMC program was launched, the loudspeaker announcements). Once it became clear that demand nursing scope of practice in Kenya did not cover VMMC; only was extremely high, radio advertisements were halted. medical doctors and clinical officers were legally authorized to On the supply side, to prevent a client-preparation bottleneck, provide VMMC surgery. In 2008, a facility readiness assessment HIV counselors from Ministry of Health sites and from PEPFAR- conducted in Nyanza Province found that while doctors and clinical funded HTC programs were given one week’s training in VMMC officers were in short supply at most health facilities, 85% of facilities education and counseling before the campaign and then deployed had sufficient nurses to provide VMMC services [27]. In June 2009, during the campaign to provide VMMC counseling and HTC in the Ministry of Health announced a new policy enabling nurses to tents pitched at the five VMMC sites and in the community. provide VMMC surgical services, and VMMC implementing Community-based counselors referred interested clients to the partners such as the Nyanza Reproductive Health Society began to nearest VMMC clinic using a referral card system. Community- train nurses to provide the entire surgery. Data to date show no based preoperative preparation of clients was also introduced. In differences in adverse events rates in VMMCs performed by nurses addition, rather than being given a specific appointment time, compared to clinical officers or doctors, as long as providers have clients were typically given a date for their surgery. This advance enough practice to reach competency [28]. scheduling approach helped to ensure that clinics received an Following the 2009 policy shift, the Kenya VMMC program adequate—but not overwhelming—number of clients each day. was able to conduct two successful Rapid Results Initiatives. In Another approach taken to match supply and demand was November and December 2009, more than 36,000 VMMCs were effective, motivational management of human resources. During the conducted; in November and December 2010, another 50,000 campaign, VMMC providers, counselors, and support staff typically men and adolescent boys received VMMC [29]. It is highly worked ten-hour days, Monday through Friday, and six-hour days on unlikely that this volume of services could have been achieved if Saturdays. To motivate personnel, the program provided T-shirts the program had had to rely on doctors and clinical officers to and daily meals. Moreover, site visits were conducted during the provide VMMC surgery. However, although task shifting to campaign to remind frontline providers that modeling studies suggest nurses has improved the efficiency of VMMC sites, the removal of that, in Iringa, one HIV infection will be averted by 2025 for every nurses from their normal public sector health facilities to 4.5 VMMCs done in the next five years. These reminders kept participate in VMMC campaigns has almost certainly affected providers motivated and focused on the HIV prevention benefits of the delivery of other health services and programs. VMMC and generated a healthy competition among sites to see which site could help avert the most new HIV infections. Swaziland Case Study: Preparing to Meet the Human Finally, live capture and reporting of service delivery data Resource Requirements of an Accelerated Saturation through a web-based data collection system (and the use of short Initiative message service text messaging to campaign headquarters when Circumcision is not a traditional practice in Swaziland, which Internet access was a challenge) helped inform the campaign has the highest adult HIV prevalence in the world (26%) [6]. In leadership about trends in service delivery numbers, so that 2006–2007, only 8.2% of Swazi men reported being circumcised necessary adjustments to the deployment of personnel (e.g., [30]. In 2009, Swaziland finalized its national VMMC strategy increased deployment of counselors to the highest-volume sites) [31], which outlined a plan for providing VMMC to 144,688 could be made in a timely manner. males within five years. However, after hearing about the success Tanzania’s application of select approaches illustrates how of Kenya’s 2009 Rapid Results Initiative, the Swaziland Male innovations can lead to successful VMMC scale-up programs. Circumcision Task Force decided to reconsider its timeline for However, rapid scale-up may also have its downsides. Tanzania VMMC scale-up. Termed the ‘‘Accelerated Saturation Initiative’’ does not currently have enough data to confidently measure the in English and Soka Uncobe (Circumcise and Conquer) in SiSwati, effect of VMMC scale-up on its existing health services, but the new initiative, which was formally launched on July 15, 2011, because the scale-up uses public health facilities, it is likely to aims to complete VMMC scale-up in less than two years. increase the work burden on the public sector employees and Modeling studies suggest that while a five-year VMMC scale-up stretch the provision of other services at health facilities. would avert an estimated 66,000 new HIV infections in Swaziland by 2025, accelerating the scale-up to one year would avert 88,000 Kenya Case Study: Empowerment of the Nursing new infections over the same time period [6]. Workforce to Provide VMMC Services Increases Coverage Several approaches have been taken to meet the HRH while Maintaining Client Safety requirements of Soka Uncobe. The first step was to establish an In Kenya, the majority of men are already circumcised for Human Resources Subcommittee within the national VMMC task cultural or religious reasons, but in Nyanza Province only 48% of force. This subcommittee worked with the Clinical Subcommittee men are circumcised. Nyanza is the epicenter of the Kenyan HIV to determine a staffing pattern for high-volume, high-efficiency epidemic, with an adult HIV prevalence of 14.9%—more than VMMC teams capable of providing 40 VMMCs per day at multi- double the national average of 7.1%. Moreover, 13.2% of bed sites (Table 2). This staffing pattern was informed by the uncircumcised men in Kenya are HIV-positive, compared to just experiences of other VMMC programs, including the Iringa 3.9% of circumcised men [25]. Region VMMC campaign in Tanzania, but includes an additional Kenya is the first country in southern and eastern Africa to reach counselor in the form of an HIV-positive ‘‘expert client’’ who scale in its VMMC program. Between its formal launch in October provides additional post-test counseling and linkages to care and 2008 and June 2011, the Kenyan VMMC program conducted treatment for prospective VMMC clients who test HIV-positive. 318,000 VMMCs, most of them in Nyanza Province [22]. Two key The Human Resources Subcommittee determined that all non- innovations contributed to this high volume of VMMCs. First, the clinical staff, including counselors, could be recruited and trained national and provincial VMMC task forces set an ambitious target of from within local communities. However, to limit the disruption to 860,000 VMMCs nationally by 2013, including 426,500 in Nyanza other key health services during Soka Uncobe, it decided that the Province. Targets were also established at the district level [26]. majority of doctors would need to be recruited as volunteers from Second, task shifting to nurses was introduced to increase the number outside of Swaziland. Pilot testing of a formal ‘‘VMMC volunteer
73 Table 2. Composition of VMMC clinical teams in Swaziland initiative.
Title Number Role on Team
Site manager 1 Ensure that the team has all supplies and equipment necessary, supervise team members and coordinate with demand-generation partners Doctor 1 Remove foreskin, achieve hemostasis, place horizontal and vertical mattress sutures Anesthesia/suture nurse 1 Apply local anesthesia, place simple interrupted sutures Bedside nurses 4 Prep and drape client, assist during MC surgery, place bandage Recovery room nurse 1 Monitor vital signs, provide postoperative care instructions, check bandage before discharging client Review nurse 1 Conduct two-day and seven-day reviews Counselors 3 Conduct group education, MC counseling, and HTC Runner (non-clinical) 1 Ensure that each surgical bay has the supplies it needs Hygienist 1 Ensure the clinic is clean, assist with waste management (only one needed due to use of fully disposable MC instrument kits and waste management company) Expert client 1 Link clients who test HIV-positive to care and treatment services, model adherence to antiretroviral therapy Receptionist 1 Book clients, start client record form Data capturer 1 Record client data for monitoring and evaluation purposes Booking agent 1 Manage client booking and link client with transport
doi:10.1371/journal.pmed.1001129.t002 doctor program’’ began in April 2010, when four urologists from public sector. At the peak of Soka Uncobe, up to 245 nurses will be the American Urological Association visited Swaziland for a two- needed to staff up to 35 teams, 182 of whom will be in addition to week period to provide VMMC services at three sites. During this nurses already working full-time in the national VMMC program. pilot trip and two subsequent volunteer doctor visits, 17 doctors However, quantification of the existing nursing workforce in from Cameroon, Ethiopia, Ghana, Lesotho, and the United States Swaziland by the Human Resources Subcommittee revealed that performed 2,935 VMMCs. Before deployment, all visiting doctors there are sufficient Swaziland-based nurses to staff Soka Uncobe and have to complete an online learning tool and send documentation that these nurses are mainly ‘‘unemployed but registered’’ and ‘‘on of their qualifications and licensure for review and registration by leave’’ (Table 3). The ‘‘unemployed but registered’’ category the Swaziland Medical and Dental Council. Upon arrival in consists primarily of foreign (mainly Zimbabwean) nurses. Because Swaziland, their history taking, clinical examination, and surgical foreign nurses are not prioritized for civil service positions, many skills are validated through an observed structured clinical of them remain unemployed. The ‘‘on leave’’ category includes examination before they proceed to supervised clinical practice, nurses on facility-approved and officially scheduled vacations. at which time their skills in the forceps-guided technique are It should be noted that the recruitment out-of-workforce nurses standardized by a full-time Soka Uncobe doctor. for Soka Uncobe has some important limitations. Because nurses will Turning to nurses, who are allowed to inject local anesthesia, need to take leave with preapproval from their employer to place simple interrupted sutures and conduct preoperative and participate in VMMC scale-up, any change in the timing or postoperative exams in Swaziland (task sharing), the Human location of the VMMC campaign might mean on-leave nurses will Resources Subcommittee initially assumed it would be impossible not be in the right place at the right time to work in the campaign. to identify enough nurses from within Swaziland to meet the needs Thus, a dependency on on-leave nurses would render the VMMC of VMMC scale-up given the nursing shortages in the Swazi scale-up program less able to change course to respond to demand.
Table 3. Quantification of the nursing workforce in the Kingdom of Swaziland.
Category Number Comments
Unemployed but registered with the 110 Zimbabwean: 57 (51.8%); Swazi: 32 (29.1%); Zambian: 4 (3.64%); Congolese: 4 (3.64%); Swaziland Nursing Council Ugandan: 1 (0.9%); Ghanaian: 1 (0.9%); Nigerian: 1 (0.9%); Unknown: 10 (9.1%) Recently retired (within five years) 12 These nurses all indicated their interest in returning to work to support MC services Newly graduating professional nurses 8 Eight graduating nurses indicated their interest in working on MC full-time Swazi nurses working in the UK 5 These nurses all indicated their interest in returning to Swaziland work to support Soka Uncobe Newly graduating nursing assistants 17 These nursing assistants can staff recovery rooms or serve as bedside nurses (not anesthesia/suture nurses) On-leave from public sector employment Average of 107 per month Most of these nurses are on vacation, not medical or maternity leave Total 259
doi:10.1371/journal.pmed.1001129.t003
74 In addition, targeted recruitment of out-of-workforce nurses will considerable work to be done in all 14 VMMC priority countries/ provide sufficient nurses for Soka Uncobe only if all new graduated regions. To perform the 20.33 million circumcisions between 2011 nurses enter government employment. This is an unrealistic and 2015 that are required to reach 80% coverage, all the priority expectation, as new graduates may pursue other options or not countries need to look at innovations, best practices, and strategies enter the workforce. in the areas of surgical efficiencies, non-surgical efficiencies, and human resource efficiencies. The approaches and case studies Conclusions discussed in this review can be adapted and refined by these other countries, but additional new approaches that address HRH Although the HRH shortages in the 14 VMMC priority challenges of VMMC scale-up specific to each country will need to countries/regions will probably not be resolved by 2015, when the be developed. Finally, we suggest that the seven approaches we suggested target of 80% VMMC coverage is due, the seven have identified for meeting the human resource needs of VMMC approaches that we discuss here (surgical efficiencies, non-surgical scale-up, and others not revealed by our program review, will also efficiencies, and five human resource efficiencies) highlight be relevant for other high-impact health interventions that are innovative solutions to HRH challenges. Moreover, the three case currently facing HRH challenges. studies that we present suggest that it is possible to achieve good results in this early phase of VMMC scale-up in southern and eastern Africa in a short time frame by developing and Acknowledgments implementing new approaches to reduce HRH shortages that The authors would like to recognize our colleagues at Orange Farm, aim to minimize disruptions to other health services. Importantly, particularly Dino Rech and Dirk Taaljard, for developing and document- the case studies demonstrate various options for addressing HRH ing surgical efficiency approaches. We would also like to thank Amy challenges in different policy environments. Herman-Roloff for providing information about the facility readiness It is unlikely that the seven approaches we have identified are assessment in Nyanza Province, Kenya, which led to the task-shifting the only approaches being used to deal with the HRH challenges policy change. Meghan Swor from Jhpiego/Swaziland provided critical associated with VMMC scale-up in southern and eastern Africa, information on the volunteer doctor initiative, and Augustino Hellar from Jhpiego/Tanzania provided key insights into the human resource planning and further work, in the form of a systematic review, may identify process for the 2010 Iringa Region campaign in Tanzania. Finally, we more approaches. In addition, most of the evidence we reviewed would like to thank Rebecca Chase Fowler and Susi Wyss for their came from programmatic and policy reports rather than the peer- assistance in editing this paper and Alice Christensen for her assistance reviewed literature and must therefore be considered with some with second and third drafts. Some of the findings contained in this paper caution. Of most concern, while it can be expected that campaigns were presented as an oral poster at the International AIDS Society that use existing public health facilities and public sector Conference in Rome, Italy, in July 2011. employees will add a burden to health service facilities and manpower, we found no published data on the repercussions of Author Contributions VMMC scale-up on other health programs. VMMC campaigns Conceived and designed the experiments: KC EN. Analyzed the data: KC could have an effect on other health initiatives by diverting public EN AM. Wrote the first draft of the manuscript: KC EN HM TA. focus, health workers, and government energy and spending. This Contributed to the writing of the manuscript: KC EN AM KD TA PC HM is an area for future study that should be tackled before the BF TKM JA LF NB JR DC DS. ICMJE criteria for authorship read and approaches described in the three case studies are widely applied. met: KC EN AM KD TA PC HM BF TKM JA LF NB JR DC DS. Agree While we found evidence for success in dealing with HRH with manuscript results and conclusions: KC EN AM KD TA PC HM BF shortages during VMMC scale-up in three countries, there is still TKM JA LF NB JR DC DS.
References 1. Weiss HA, Quiqley MA, Hayes RJ (2000) Male circumcision and risk of HIV 9. Chen L, Evans T, Anand S, Boufford JI, Brown H, et al. (2004) Human infection in sub-Saharan Africa: A systematic review and meta-analysis. AIDS resources for health: overcoming the crisis. Lancet 364: 1984–1990. 14: 2361–2370. 10. Callaghan M, Ford N, Schneider H (2010) A systematic review of task-shifting 2. Gray RH, Kigozi G, Serwadda D, Makumbi F, Watya S, et al. (2007) Male for HIV treatment and care in Africa. Hum Resour Health 8: 8. circumcision for HIV prevention in men in Rakai, Uganda: A randomised trial. 11. Ozgediz D, Kijjambu S, Galukande M, Dubowitz G, Mabweijano J, et al. (2008) Lancet 369: 657–666. Africa’s neglected surgical workforce crisis. Lancet 371: 627–628. 3. Bailey RC, Moses S, Corette BP, Agot K, Maclean I, et al. (2007) Male 12. Joint United Nations Programme on HIV/AIDS (2010) Global report: circumcision for HIV prevention in young men in Kisumu, Kenya: a UNAIDS report on the global AIDS epidemic. Available: http://www.unaids. randomized controlled trial. Lancet: 369: 643–656. org/globalreport/global_report.htm. Accessed 15 March 2011. 4. Auvert B, Taljaard D, Lagarde E, Sobngwi-Tambekou J, Sitta R, et al. (2005) 13. Tawfik L, Kinoti SN (2003) The impact of HIV/AIDS on health systems and Randomized, controlled intervention trial of male circumcision for reduction of the health workforce in sub-Saharan Africa. Washington (District of Columbia): HIV infection risk: The ANRS 1265 Trial. PLoS Med 2: e298. doi:10.1371/ United States Agency for International Development Bureau for Africa, Office journal.pmed.0020298. of Sustainable Development. Available: usaid.gov/our_work/global_health/ 5. World Health Organization, Joint United Nations Programme on HIV/AIDS pop/news/hcdworkforce.doc. Accessed 15 March 2011. (2007) WHO and UNAIDS announce recommendations from expert meeting 14. Ogilvie L, Mill JE, Astle B, Fanning A, Opare M (2007) The exodus of health on male circumcision for HIV prevention. Available: http://data.unaids.org/ professionals from sub-Saharan Africa: balancing human rights and societal pub/pressrelease/2007/20070328_pr_mc_recommendations_en.pdf. Accessed needs in the twenty-first century. Nurs Inq 14: 114–124. 15 March 2011. 15. Rakoum C (2010) Nursing human resources in Kenya case study. Geneva: 6. Njeuhmeli E, Forsythe S, Reed J, Opuni M, Bollinger L, et al. (2011) Voluntary International Centre for Human Resources in Nursing. Available: http://www. medical male circumcision: modeling the impact and cost of expanding male ichrn.com/publications/casestudies/ICHRN_Kenya_CaseStudy.pdf. Accessed circumcision for HIV prevention in eastern and southern Africa. PLoS Med 8: 15 March 2011. e1001132. doi:10.1371/journal.pmed.1001132. 16. World Health Organization, Joint United Nations Programme on HIV/AIDS, 7. World Health Organization (2006) World health report 2006: working together Jhpiego (2008) Manual for male circumcision under local anaesthesia. Geneva: for health. Geneva: World Health Organization. Available: http://www.who. World Health Organization. Available: http://www.who.int/hiv/pub/malecir int/whr/2006/whr06_en.pdf. Accessed 15 March 2011. cumcision/who_mc_local_anaesthesia.pdf. Accessed 15 March 2011. 8. World Health Organization (2004) The global burden of disease: 2004 update. 17. Lissouba P, Taljaard D, Rech D, Doyle S, Shabangu D, et al. (2010) A model for Geneva: World Health Organization. Available: http://www.who.int/ the roll-out of comprehensive adult male circumcision services in African low- healthinfo/global_burden_disease/GBD_report_2004update_full.pdf. Accessed income settings of high HIV incidence: the ANRS 12126 Bophelo Pele Project. 15 March 2011. PLoS Med 7: e1000309. doi:10.1371/journal.pmed.1000309.
75 18. Dickson K, Farley T (2010) Male circumcision scale-up [abstract]. 17th Programme. Available: http://www.prb.org/pdf09/kaiskenyadatasheet.pdf. Ac- Conference on Retroviruses and Opportunistic Infections; San Francisco, cessed 15 March 2011. California; 16–19 February 2010. Available: http://www.retroconference.org/ 26. National AIDS and STI Control Programme (2010) Progress report on Kenya’s 2010/Abstracts/39598.htm. Accessed 15 March 2011. voluntary medical male circumcision programme, 2008-09: summary. Nairobi: 19. Come JC, Comissa´rio L, Seclen J, Ponnet M, Bossemeyer D, et al. (2009) National AIDS and STI Control Programme. Available: http://www.malecir Avaliac¸a˜o da situac¸a˜o e necessidades na prestac¸a˜o de servic¸os de cirurgia geral cumcision.org/documents/VMMCP_Report.pdf. Accessed 15 March 2011. ba´sica e pequena cirurgia em Moc¸ambique. Maputo (Mozambique): Ministry of 27. Herman-Roloff A, Otieno N, Agot K, Ndinya-Achola J, Bailey RC (2011) Health (MISAU). Acceptability of medical male circumcision among uncircumcised men in Kenya 20. The United States President’s Emergency Plan for AIDS Relief Male one year after the launch of the national male circumcision program. PLoS Circumcision Technical Working Group (2009) PEPFAR Male Circumcision ONE 6: e19814. doi:10.1371/journal.pone.0019814. Partners’ Meeting: commodities and improved coordination of male circumci- 28. World Health Organization (2010) Considerations for implementing models for sion for HIV prevention. Washington (District of Columbia): The United States optimizing the volume and efficiency of male circumcision services. Geneva: President’s Emergency Plan for AIDS Relief. Available: http://www.malecir World Health Organization. Available: http://www.malecircumcision.org/ cumcision.org/publications/documents/PEPFAR_MC_CommoditiesMeeting. programs/documents/mc_MOVE_2010_web.pdf. Accessed 15 March 2011. pdf. Accessed 15 March 2011. 29. Herman-Roloff A, Bailey R, Agot K, Ndinya-Achola J (2010) Medical male 21. Mahler HR, Kileo B, Curran K, Plotkin M, Adamu T, et al. (2011) Voluntary circumcision for HIV prevention in Kenya: a study of service provision and medical male circumcision: matching demand and supply with quality and adverse events [abstract]. XVIII International AIDS Conference; Vienna, efficiency in a high-volume campaign in Iringa Region, Tanzania. PLoS Med 8: Austria; 18–23 July 2010. Available: http://pag.aids2010.org/Abstracts. e1001131. doi:10.1371/journal.pmed.1001131. aspx?AID = 6780. Accessed 15 March 2011. 22. Mwandi Z, Murphy A, Reed J, Chesang K, Njeuhmeli E, et al. (2011) Voluntary 30. Swaziland Central Statistical Office, Measure DHS (2007) Swaziland demo- medical male circumcision: translating research into the rapid expansion of graphic and health survey 2006-2007: preliminary report. Mbabane: Swaziland services in Kenya, 2008–2011. PLoS Med 8: e1001130. doi:10.1371/ Ministry of Health and Social Welfare. Available: http://www.nercha.org.sz/ journal.pmed.1001130. publications/Publications/report/SDHSPrelimReport11.pdf. Accessed 15 23. Tanzania Commission for AIDS, Zanzibar AIDS Commission, National Bureau March 2011. of Statistics, Office of the Chief Government Statistician, Macro International 31. Swaziland Male Circumcision Task Force (2009) Strategy and implementation (2008) Tanzania HIV/AIDS and malaria indicator survey 2007-08. Dar es plan for scaling up safe male circumcision for HIV prevention in Swaziland Salaam: Tanzania Commission for AIDS. Available: http://www.tacaids.go.tz/ 2009/2013. Mbabane: Swaziland Male Circumcision Taskforce. Available: dmdocuments/THMIS%202007-08.pdf. Accessed 15 March 2011. http://www.malecircumcision.org/programs/documents/Strategy_and_Imple 24. Plotkin M, Hawa Mziray H, Ku¨ver J, Prince J, Curran K, et al. (2011) Embe mentation_Plan_for_Scaling_up_Safe_MC_in_Swaziland_web.pdf. Accessed Halijamenywa: the unpeeled mango: a qualitative assessment of views and 15 March 2011. preferences on voluntary medical male circumcision in Iringa Region, Tanzania, 32. World Health Organization (2008) WHO global atlas of the health workforce. MCHIP Technical Report. Available: http://www.mchip.net/sites/default/ Geneva: World Health Organization Department of Human Resources for files/The%20Unpeeled%20Mango%20-%20Final%20Report%202011.pdf. Health. Available: http://www.hrhresourcecenter.org/global_atlas. Accessed 15 Accessed 28 September 2011. March 2011. 25. National AIDS and STI Control Programme (2009) Kenya AIDS indicator 33. Pincock S (2007) Workforce biggest barrier to roll-out of male circumcision. survey: KAIS (2007)—data sheet. Nairobi: National AIDS and STI Control Lancet 370: 1817–1818.
76 Review Voluntary Medical Male Circumcision: Translating Research into the Rapid Expansion of Services in Kenya, 2008–2011
Zebedee Mwandi1*, Anne Murphy2, Jason Reed3, Kipruto Chesang1, Emmanuel Njeuhmeli4, Kawango Agot5, Emma Llewellyn6, Charles Kirui7, Kennedy Serrem8, Isaac Abuya9, Mores Loolpapit10, Regina Mbayaki11, Ndungu Kiriro12, Peter Cherutich13, Nicholas Muraguri13, John Motoku14, Jack Kioko15, Nancy Knight1, Naomi Bock3 1 Division of Global HIV/AIDS, United States Centers for Disease Control and Prevention, Nairobi, Kenya, 2 United States Agency for International Development, Nairobi, Kenya, 3 Division of Global HIV/AIDS, Centers for Disease Control and Prevention, Atlanta, Georgia, United States of America, 4 United States Agency for International Development, Washington, District of Columbia, United States of America, 5 Impact Research and Development Organization, Kisumu, Kenya, 6 Nyanza Reproductive Health Society, Kisumu, Kenya, 7 Kenya Medical Research Institute—Family AIDS and Care Education Services, Nairobi, Kenya, 8 Catholic Medical Missions Board, Nairobi, Kenya, 9 C-Change Communication for Change, Nairobi, Kenya, 10 Family Health International, Nairobi, Kenya, 11 Engender Health (APHIA II Nyanza), Kisumu, Kenya, 12 Population Services International, Nairobi, Kenya, 13 Kenya National AIDS and STD Control Programme, Nairobi, Kenya, 14 Eastern Deanery AIDS Response Program, Nairobi, Kenya, 15 Ministry of Public Health and Sanitation, Kisumu, Kenya
should include medical MC as part of HIV prevention Abstract: Since the World Health Organization and the interventions and that implementation should be prioritized in Joint United Nations Programme on HIV/AIDS recom- areas with low MC and high HIV prevalence rates [5]. WHO and mended implementation of medical male circumcision UNAIDS identified 13 priority countries for scale-up of medical (MC) as part of HIV prevention in areas with low MC and MC. The United States President’s Emergency Plan for AIDS high HIV prevalence rates in 2007, the government of Relief (PEPFAR) is supporting activities to implement medical MC Kenya has developed a strategy to circumcise 80% of in these 13 countries plus the Gambella National Regional State in uncircumcised men within five years. To facilitate the quick translation of research to practice, a national MC Ethiopia (Table 1). task force was formed in 2007, a medical MC policy was Nyanza Province in Kenya is one of the regions in sub-Saharan implemented in early 2008, and Nyanza Province, the Africa prioritized by WHO and UNAIDS for implementation of region with the highest HIV burden and low rates of medical MC (Table 1). Although more than 80% of men in Kenya circumcision, was prioritized for services under the are circumcised [6], MC coverage varies culturally and geograph- direction of a provincial voluntary medical male circum- ically. Nyanza Province, which is largely Luo, has the lowest MC cision (VMMC) task force. The government’s early and coverage (48%) and the highest prevalence of HIV (14.9%) in continuous engagement with community leaders/elders, politicians, youth, and women’s groups has led to the rapid endorsement and acceptance of VMMC. In addition, Citation: Mwandi Z, Murphy A, Reed J, Chesang K, Njeuhmeli E, et al. (2011) Voluntary Medical Male Circumcision: Translating Research into the Rapid several innovative approaches have helped to optimize Expansion of Services in Kenya, 2008–2011. PLoS Med 8(11): e1001130. VMMC scale-up. Since October 2008, the Kenyan VMMC doi:10.1371/journal.pmed.1001130 program has circumcised approximately 290,000 men, Academic Editor: Stephanie L. Sansom, Centers for Disease Control and mainly in Nyanza Province, an accomplishment made Prevention, United States of America possible through a combination of governmental leader- ship, a documented implementation strategy, and the Published November 29, 2011 adoption of appropriate and innovative approaches. This is an open-access article, free of all copyright, and may be freely reproduced, Kenya’s success provides a model for others planning distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 VMMC scale-up programs. public domain dedication. Funding: This work was funded by the President’s Emergency Plan for AIDS Relief (PEPFAR) through US Centers for Disease Control and Prevention and USAID, and implemented by various PEPFAR partners in Kenya. Kenya Ministry of Introduction Public Health and Sanitation, NASCOP and the health authorities of Nyanza province and all District played important roles in guiding this work. The findings In the past two decades, observational studies have provided and conclusions in this paper are those of the author(s) and do not necessarily represent the official position of the US Centers for Disease Control and increasing evidence that male circumcision (MC) has an HIV Prevention. prevention effect [1]. Moreover, three randomized controlled trials Competing Interests: The authors have declared that no competing interests have reported that medical circumcision of men reduces HIV exist. acquisition from infected female partners by approximately 60% Abbreviations: MC, male circumcision; NASCOP, National AIDS and STI Control [2–4]. This evidence led the World Health Organization (WHO) Programme; PEPFAR, United States President’s Emergency Plan for AIDS Relief; and the Joint United Nations Programme on HIV/AIDS RRI, Rapid Results Initiative; UNAIDS, Joint United Nations Programme on HIV/ (UNAIDS) to issue recommendations in 2007 that countries AIDS; VMMC, voluntary medical male circumcision; WHO, World Health Organization * E-mail: [email protected] Review articles synthesize in narrative form the best available evidence on a topic. Provenance: Submitted as part of a sponsored Collection; externally peer reviewed
77 Summary Points countries in sub-Saharan Africa have made less progress towards meeting their MC targets [10] (Table 1). In this case study, we N Kenya’s male circumcision for HIV prevention policy describe the approach that Kenya has taken to translating research prioritizes Nyanza Province, the region with the highest into policy and program, and identify three key factors— HIV burden and low circumcision rates, for scale-up of government leadership, program flexibility, and a documented voluntary medical male circumcision (VMMC) services. implementation strategy—that have facilitated Kenya’s early N Since the policy’s implementation in October 2008, success in the scale-up of medical MC. We also discuss the lessons approximately 290,000 adult males have been circum- learned and the challenges that still need to be overcome before cised in Kenya, most of them in Nyanza Province. Kenya can reach its MC target. N Government leadership and a documented implemen- tation strategy have been key factors in Kenya’s rapid Government Leadership scale-up of VMMC. N Another key factor has been program flexibility: the The Kenyan Ministry of Health and the National AIDS and introduction of innovative approaches, including task STI Control Programme (NASCOP) began providing leadership shifting, short intensive service campaigns, and, most on medical MC for HIV prevention before the conclusion of the recently, diathermy for hemostasis, have all helped the randomized controlled trials mentioned above and before WHO program respond to challenges. issued its recommendations in 2007 [5]. At a consultative meeting N Kenya’s successful approach to VMMC scale-up provides held in Nairobi in September 2006, researchers, policy makers, a model that other countries can adapt to their own donors, and other stakeholders discussed how Kenya should circumstances. respond to the results of the Kisumu and Rakai randomized controlled trials, whether positive or negative. In December 2006, when these studies were stopped early because of MC’s Kenya [6,7]. Given these data and the WHO recommendations, overwhelming efficacy in reducing HIV transmission risk, Kenya’s the government of Kenya, through the Ministry of Health, Director of Medical Services issued a statement calling for the recognizes medical MC as an additional and important strategy establishment of a national MC task force to advise the for the prevention of heterosexually acquired HIV infection in government on how to proceed. The ‘‘National Guidance for men and has developed a national strategy that aims to circumcise Voluntary Male Circumcision in Kenya’’—the first national MC 80% of uncircumcised HIV-negative men aged 15–49 years policy in sub-Saharan Africa—was drafted by this task force, (approximately 860,000 men throughout the country, 426,000 in approved in December 2007, and published in January 2008 [11]. Nyanza Province alone) between 2009 and 2013 [8]. Modeling To complement the work of the Kenya national MC task force, a studies conducted by the United States President’s Emergency Nyanza Province MC task force and district coordinating bodies Plan for AIDS Relief and UNAIDS estimate that a scale-up were also established in early 2007, and assessments of health strategy of 80% medical MC coverage in five years in Nyanza facilities in Nyanza Province were conducted to determine the Province could avert an estimated 45,000 new HIV infections over province’s preparedness to provide VMMC services. Gaps were 15 years [9]. identified and remedied with support from international donors. Kenya has already made good progress towards meeting its MC While work on the national policy document was proceeding, target (Table 1). By contrast, most other MC scale-up priority the Kenyan government took steps to engage the Luo Council of
Table 1. Target number of HIV-negative males needed to be medically circumcised by 2015 to reach 80% coverage, and approximate proportion of those reached with medical MC services through late 2011.
Target Number of 15- to 49-Year-Old, HIV- Approximate Percentage Circumcised since 2007 Country Negative, Uncircumcised Males (Approximate) WHO Recommendations (Rounded)
Botswana 345,000 5 Ethiopa: Gambella National Regional State 40,000 15 Kenya: Nyanza Province 380,000a/426,000b 55a/50b Lesotho 377,000 ,5 Malawi 2,102,000 ,5 Moazmbique 1,059,000 ,5 Namibia 330,000 ,5 Rwanda 1,746,000 ,5 South Africa 4,333,000 ,5 Swaziland 183,000 15 Tanzania 1,373,000 ,5 Uganda 4,250,000 ,5 Zambia 1,949,000 ,5 Zimbabwe 1,913,000 ,5
aEstimate and calculation based upon Decision Makers’ Program Planning Tool [9]. bEstimate and calculation based upon the Kenya national strategy for VMMC [19]. doi:10.1371/journal.pmed.1001130.t001
78 Elders in Nyanza Province in the scale-up of medical MC. To gain from the launch of the national VMMC policy, Kenya permitted the support of these protectors of Luo culture for medical MC clinical officers to perform VMMC in addition to medical officers, scale-up, the government needed to explain to them why medical thus providing an expanded pool of health care providers who MC would be recommended for HIV prevention and how medical could perform VMMC. However, health facilities were under- MC was biologically protective against the HIV virus. In addition, staffed with clinical officers and medical officers, and in 2009, the the government needed to improve its understanding of the director of medical services, on the recommendation of the council’s potential concerns. Repeated discussions satisfied the Luo Kenyan national MC task force, supported further task shifting to Council of Elders that MC for HIV prevention would be voluntary allow nurses to perform VMMC, thereby ensuring that there and provided for medical and not cultural reasons. As a result, the would be sufficient human resources available to achieve the goals term ‘‘voluntary medical male circumcision’’ (VMMC) was of the VMMC program. Task shifting of VMMC to nurses likely officially adopted in Kenya, instead of just ‘‘male circumcision.’’ also helps to preserve the pool of clinical and medical officers able The day before the official launch of VMMC services in to meet the other demands of the Kenyan health care system, but October 2008, three community-based stakeholders’ meetings research is needed to confirm this possibility. were held with cultural leaders, government ministers (including the prime minister of Kenya, Raila Odinga, and the minister of Clinical Techniques health), local politicians, youth, religious and women’s groups, and Because trained personnel are in short supply—even with task health professionals. In addition, several members of parliament shifting—making the most efficient use of health care providers’ and cabinet ministers publicly disclosed that they were circum- time is important to maximize productivity. The WHO guidance cised, as a show of support for Kenya’s VMMC programs for HIV ‘‘Considerations for Implementing Models for Optimizing the prevention. Volume and Efficiency of Male Circumcision Services for HIV Prevention’’ summarizes options for improving service productiv- Program Flexibility ity, without compromising safety and quality [13], including the time-saving advantages of diathermy for hemostasis. Subsequent Service Delivery Approaches to the release and wide adoption of the service delivery models Service delivery providers have implemented VMMC services in proposed by WHO, the Kenyan VMMC program has begun Nyanza Province using a variety of site and staffing models. incorporating the use of diathermy to improve program efficiency Permanent sites in larger health care facilities staffed with existing and productivity. local health care personnel, outreach services that temporarily deploy health care teams to smaller health facilities, and mobile services that temporarily deploy health care teams to non-health-care facilities Implementation Strategy such as churches, schools, or tents have all been used. These sites are The Kenyan government operationalized its national policy in both public and private, as are staff, and staffing models have a written implementation strategy that is part of the Kenya frequently included government and non-government staff working National AIDS Strategic Plan [8]. The implementation strategy side by side. Drawing upon a variety of public and private site and provides guidance on the quality of service delivery, delivery of staffing models, instead of restricting services to a particular model, services with full human rights’ considerations, correct dissemi- has produced a program that is flexible and adaptable. nation of information within the context of broader HIV Twelve months after the VMMC program began, only 46,000 prevention interventions, demand creation, and monitoring and boys and men had been circumcised, and officials recognized that evaluation. It employs a three-phased approach, consisting of they would not reach the program’s target. Consequently, short-, medium-, and long-term objectives, and outlines annual NASCOP and the Ministry of Health introduced a Rapid Results targets and milestones for each region, through 2013. To date, Initiative (RRI) approach as an additional service delivery model. implementation and service delivery have been concentrated in The RRI is a public health service delivery strategy that focuses on Nyanza Province. short-term, results-oriented activities designed to reach a high From the outset, as part of its implementation strategy, the number of individuals quickly; RRIs have been previously program has focused on strategies for social mobilization, undertaken in Kenya for HIV testing and counseling campaigns advocacy, and health communication. To achieve and sustain and immunization campaigns [12]. social mobilization, journalists have been educated about the The first VMMC RRI, which ran in November and December science underlying the use of VMMC for HIV prevention to 2009 to coincide with school holidays, aimed to circumcise 30,000 ensure accurate reporting of the national strategy and program adolescent and adult males in 11 districts in Nyanza Province in 30 and to help create a positive public perception of VMMC. working days [12]. Tents were used extensively during the Although the initial national-level advocacy yielded tremendous campaign to provide services in areas where more permanent gains, program partners have also engaged in intense consultation infrastructure was lacking. As with all VMMC services for HIV with gatekeepers at the community level. Finally, program prevention, in addition to providing the MC surgical procedure, partners have sensitized health workers to the role of VMMC the RRI offered each client a comprehensive HIV prevention within the context of HIV prevention programming. package that included HIV testing and counseling, screening and treatment for sexually transmitted infections, and promotion and Achievements provision of condoms. A second VMMC RRI, which ran in November and December 2010, aimed to circumcise 46,000 men. Since October 2008, the Kenyan VMMC program has circumcised approximately 290,000 men, mainly in Nyanza Human Resources Province (Figures 1 and 2), and more than 700 providers of In many countries in sub-Saharan Africa, only physicians various cadres have been trained to provide VMMC services. (medical officers) may perform medical MC. In task shifting, a Although the 2009 and 2010 RRIs, which completed about trained individual from a less educated health care cadre (a non- 36,000 and 50,000 VMMCs, respectively (personal communica- physician) is permitted to perform a medical procedure. Right tion, A. Ochieng, NASCOP), boosted the overall number of men
79 Figure 1. Cumulative circumcisions done in Kenya, 2008–2011. doi:10.1371/journal.pmed.1001130.g001 circumcised in Kenya, monthly performances outside these The quality of service delivery has also increased over the life of periods have increased from as low as 3,000 VMMCs in the first the project. For example, uptake of HIV testing among VMMC ten months (October 2008–July 2009) to an average of about clients at Nyanza Reproductive Health Society—one the largest 6,000 VMMCs in recent months (May 2010–June 2011). providers of VMMC services in Kenya—has increased since the Improvements in service efficiency, dedication of full-time space beginning of the program, from 31% in 2008–2009 to more than and staff, increased demand for services, and greater availability of 83% presently, largely because of a shift to a provider-initiated outreach/mobile services have all likely contributed to higher HIV testing approach from opt-in HIV testing (personal overall service numbers outside the RRIs. communication, A. Ochieng, NASCOP).
Figure 2. Monthly circumcisions done in Kenya, 2008–2011. doi:10.1371/journal.pmed.1001130.g002
80 In addition, a routine clinical record and reporting system has campaigns, and by urging men to involve their partners in the been adopted by all service providers, with a standard set of intra- decision-making process. A comprehensive communication cam- operative and postoperative adverse event definitions, based on paign has been implemented recently that addresses demand WHO guidance [14]. Adverse event occurrences, along with other creation and women’s roles in men’s decisions about VMMC [18]. service statistics, are now aggregated and reported through health Kenya also needs to determine the best way to expand VMMC management and information systems to the Ministry of Health for services to other regions, communities, and cultures, and to decide review by the national and provincial MC task forces. Overall, whether to integrate youth and neonatal medical MC into existing moderate and severe adverse event rates have remained at or below health services or to make them standalone services like the 3% since 2009 (personal communication, A. Ochieng, NASCOP). adolescent/adult VMMC program. Specifically, the sustainability Finally, the proportion of men aged 15 years or older of medical MC through the implementation of neonatal undergoing VMMC has increased over time from approximately circumcision, which is not a common practice in Kenya, must 55% in the 2009 VMMC RRI to 84% in the 2010 RRI (personal be explored. communication, A. Ochieng, NASCOP), an encouraging result, given that preferential targeting of VMMC to males who are now Conclusion or soon will be sexually active is needed to accelerate the prevention impact of VMMC programs. The Kenyan VMMC experience has shown that with strong leadership from the government (the Kenyan government has Lessons Learned assumed visible ownership of the VMMC program throughout its development and implementation and has focused stakeholders’ The experiences of Kenya’s VMMC program suggest that early attention on the number of HIV infections likely to be averted engagement of traditional leaders from non-circumcising commu- through VMMC scale-up), and with the enthusiastic participation nities can benefit national policy and implementation strategy of stakeholders, it is possible to initiate and expand VMMC in a development processes. They also suggest that flexible expansion short period of time. The Kenyan VMMC program—one of the of task shifting to allow nurses to perform medical MC can lessen first successful early translations of MC health research into human resource constraints without compromising safety [15]. implementation—provides a model that may help guide other Moreover, they indicate that retooling the implementation strategy countries in the region that are experiencing a slower scale-up of to include mobile and outreach services and RRIs can effectively their VMMC programs. In particular, the engagement of increase the uptake of VMMC. Finally, anecdotal best practice traditional/community leaders, the establishment of national and from the field suggests that using diathermy for hemostasis can local leadership bodies, and Kenya’s willingness to consider improve efficiency [13]; service providers are now beginning to multiple approaches to deal with implementation challenges hold train health care workers on the use of diathermy in an effort to important lessons for other countries. Finally, the experience of further increase productivity. Kenya’s VMMC program emphasizes the importance of having a comprehensive, timed, and actionable implementation strategy to Challenges which full-time staff from both the national government and Despite the general success of the Kenyan VMMC program, agencies of foreign governments are dedicated. several challenges remain. First, more must be done to overcome barriers among older men to go for VMMC services. These Acknowledgments barriers include hesitations about taking time off work after The authors wish to thank Dr. Athansius Ochieng from NASCOP and surgery and particular concerns about abstaining from sex during George Otieno from Nyanza Reproductive Health Society for their helpful wound healing among married men [16]. Presently, the national and timely contributions. and provincial MC task forces are piloting new approaches to recruit older men to services, such as utilizing older circumcised Author Contributions clients as community mobilizers and providing incentives for these community mobilizers when older men present to VMMC Conceived and designed the experiments: ZM AM JR KC EN KA EL CK facilities for information. KS IA ML RM NK PC NM JM JK NK NB. Analyzed the data: ZM AM As VMMC is considered primarily a man’s issue, involving JR KC EN KA EL CK KS IA ML RM NK PC NM JM JK NK NB. Wrote the first draft of the manuscript: ZM EN NB. Contributed to the women in VMMC programs can be a challenge. However, studies writing of the manuscript: ZM AM JR KC EN KA EL CK KS IA ML RM have shown that women play a large role in men’s decision to be NK PC NM JM JK NK NB. ICMJE criteria for authorship read and met: circumcised [17]. Kenya’s VMMC program has made efforts to ZM AM JR KC EN KA EL CK KS IA ML RM NK PC NM JM JK NK reach out to women by encouraging HIV testing and counseling of NB. Agree with manuscript results and conclusions: ZM AM JR KC EN couples, by targeting women with gender-focused communication KA EL CK KS IA ML RM NK PC NM JM JK NK NB.
References 1. Weiss HA, Quigley MA, Hayes RJ (2000) Male circumcision and risk of HIV 5. World Health Organization, Joint United Nations Programme on HIV/AIDS infection in sub-Saharan Africa: A systematic review and meta-analysis. AIDS (2007) New data on male circumcision and HIV prevention: policy and 14: 2361–2370. programme implications. Available: http://libdoc.who.int/publications/2007/ 2. Auvert B, Taljaard D, Lagarde E, Sobngwi-Tambekou J, Sitta R, et al. (2005) 9789241595988_eng.pdf. Accessed 22 September 2011. Randomized, controlled intervention trial of male circumcision for reduction of 6. Kenya Ministry of Health National AIDS and STI Control Programme (2009 HIV infection risk: the ANRS 1265 Trial. PLoS Med 2: e298. doi:10.1371/ September) Kenya AIDS indicator survey 2007: final report. Nairobi: National journal.pmed.0020298. AIDS and STI Control Programme, Available: http://www.nacc.or.ke/ 3. Bailey RC, Moses S, Parker CB, Agot K, Maclean I, et al. (2007) Male nacc%20downloads/official_kais_report_2009.pdf. Accessed 22 September circumcision for HIV prevention in young men in Kisumu, Kenya: a 2011. randomised controlled trial. Lancet 369: 643–656. 7. Kenya National Bureau of Statistics, ICF Macro (2010) Kenya demographic and 4. Gray RH, Kigozi G, Serwadda D, Makumbi F, Watya S, et al. (2007) Male health survey 2008–09. Calverton (Maryland): ICF Macro, Available: http:// circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial. www.measuredhs.com/pubs/pdf/FR229/FR229.pdf. Accessed 22 September Lancet 369: 657–666. 2011.
81 8. Kenya National AIDS Control Council (2009) Kenya national AIDS strategic 14. World Health Organization, Joint United Nations Programme on HIV/AIDS, plan 2009/10–2012/13—delivering on universal access to services. Available: Jhpiego (2009) Manual for male circumcision under local anaesthesia, version http://www.nacc.or.ke/nacc%20downloads/knasp_iii.pdf. Accessed 22 Sep- 2.5. Geneva: World Health Organization, Available: http://www.who.int/hiv/ tember 2011. pub/malecircumcision/who_mc_local_anaesthesia.pdf. Accessed 27 September 9. Unites State Agency for International Development Health Policy Initiative, 2011. Joint United Nations Programme on HIV/AIDS (2009) Decision makers’ 15. Moguche J, Pavin M, Odingo G, Were R, Beatty S, et al. (2011) Improving male program planning tool—calculating the costs and impacts of a male circumcision circumcision coverage through tasking shifting to non-physician clinicians program. Washington (District of Columbia): Futures Group, Health Policy [abstract]. International AIDS Society Conference; Rome, Italy, July 2011. Initiative, Task Order 1. Available: http://www.malecircumcision.org/about/ Available: http://pag.ias2011.org/EPosterHandler.axd?aid = 3707. Accessed 22 documents/DMPPT_manual_110909.pdf. Accessed 22 September 2011. September 2011. 10. World Health Organization, Joint United Nations Programme on HIV/AIDS 16. Herman-Roloff A, Otieno N, Agot K, Ndinya-Achola J, Bailey RC (2011) (2011) Progress in scale-up of male circumcision for HIV prevention in eastern Acceptability of medical male circumcision among uncircumcised men in Kenya and southern Africa: focus on service delivery—2011. Geneva: World Health one year after the launch of the national male circumcision program. PLoS Organization. ONE 6: e19814. doi:10.1371/journal.pone.0019814. 11. Kenya Ministry of Health National AIDS and STD Control Programme (2008) 17. Obure A, Nyambedha E, Oindo B (2011) Interpersonal influences in the scale- National guidance for voluntary male circumcision in Kenya. Available: http:// up of male circumcision services in a traditionally non-circumcising community www.malecircumcision.org/programs/documents/KenyaMCguidance.pdf. Ac- in rural western Kenya. Glob J Community Psychol Pract 1: 1–11. cessed 22 September 2011. 18. Kenya Ministry of Public Health and Sanitation (2010) Voluntary medical male 12. Kenya Ministry of Public Health and Sanitation, National AIDS and STI Control Programme (2011) Voluntary medical male circumcision for HIV circumcision (VMMC) communication guide for Nyanza Province. Available: prevention: report of the first Rapid Results Initiative, conducted in November/ http://www.c-hubonline.org/7815/17/NYANZA_COMMUNICATION_ December 2009. Nairobi: National AIDS and STI Control Programme. GUIDE.pdf. Accessed 22 September 2011. 13. World Health Organization (2010) Considerations for implementing models 19. Kenya Ministry of Public Health and Sanitation, National AIDS and STD for optimizing the volume and efficiency of male circumcision services for Control Programme (2009) Kenya national strategy for voluntary medical male HIV prevention. Available: http://www.malecircumcision.org/programs/ circumcision. documents/mc_MOVE_2010_web.pdf. Accessed 22 September 2011.
82 Review Voluntary Medical Male Circumcision: Matching Demand and Supply with Quality and Efficiency in a High-Volume Campaign in Iringa Region, Tanzania
Hally R. Mahler1*., Baldwin Kileo1., Kelly Curran2., Marya Plotkin1., Tigistu Adamu2., Augustino Hellar1, Sifuni Koshuma3, Simeon Nyabenda1, Michael Machaku1, Mainza Lukobo-Durrell2, Delivette Castor4, Emmanuel Njeuhmeli4, Bennett Fimbo5 1 Jhpiego/Tanzania, Dar es Salaam, Tanzania, 2 Jhpiego, Baltimore, Maryland, United States of America, 3 Ministry of Health, Iringa Region, Tanzania, 4 Office of HIV and AIDS, United States Agency for International Development, Washington, District of Columbia, United States of America, 5 Ministry of Health and Social Welfare, Dar es Salaam, Tanzania
surgical procedure from surgeon to nurses and/or clinical Abstract: The government of Tanzania has adopted officers, ‘‘task sharing’’ less complex steps of the circumcision voluntary medical male circumcision (VMMC) as an procedure to lower credentialed but highly trained health care important component of its national HIV prevention cadres, and the use of forceps-guided surgical techniques, alcohol strategy and is scaling up VMMC in eight regions gel scrubs between surgical cases, and electrocautery for nationwide, with the goal of reaching 2.8 million hemostasis. uncircumcised men by 2015. In a 2010 campaign lasting In 2009, the government of Tanzania adopted WHO’s recom- six weeks, five health facilities in Tanzania’s Iringa Region mendation to scale up VMMC [4]. Tanzania has adult HIV and performed 10,352 VMMCs, which exceeded the cam- paign’s target by 72%, with an adverse event (AE) rate of male circumcision prevalences of 5.7% and 67%, respectively [7]. 1%. HIV testing was almost universal during the Regional variations in both male circumcision and HIV campaign. Through the adoption of approaches designed prevalence exist that correlate inversely. Religious, educational, to improve clinical efficiency—including the use of the ethnic, and cultural factors and differences in the proportion of forceps-guided surgical method, the use of multiple beds people living in urban and rural settings may explain some of the in an assembly line by surgical teams, and task shifting regional differences in male circumcision coverage. Notably, a and task sharing—the campaign matched the supply of national situation assessment on VMMC reported that 93% of VMMC services with demand. Community mobilization respondents in traditionally non-circumcising areas of Tanzania and bringing client preparation tasks (such as counseling, would take their sons to be circumcised if the services were testing, and client scheduling) out of the facility and into the community helped to generate demand. This case Citation: Mahler HR, Kileo B, Curran K, Plotkin M, Adamu T, et al. (2011) Voluntary study suggests that a campaign approach can be used to Medical Male Circumcision: Matching Demand and Supply with Quality and provide high-volume quality VMMC services without Efficiency in a High-Volume Campaign in Iringa Region, Tanzania. PLoS Med 8(11): compromising client safety, and provides a model for e1001131. doi:10.1371/journal.pmed.1001131 matching supply and demand for VMMC services in other Academic Editor: Stephanie L. Sansom, Centers for Disease Control and settings. Prevention, United States of America Published November 29, 2011 This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for Introduction any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication. Several randomized controlled trials have demonstrated Funding: This work was funded by the President’s Emergency Plan for AIDS the safety and efficacy of voluntary medical male circumcision Relief (PEPFAR) through the United State Agency for International Development (VMMC) in HIV transmission prevention among heterosexual (USAID) and implemented by the Maternal and Child Health Integrated Program men [1–3]. Consequently, in 2007, the World Health Organi- (MCHIP, award # GHS-A-00-08-00002-000) managed by Jhpiego—an affiliate of zation (WHO) and the Joint United Nations Programme on Johns Hopkins University. Tanzania’s Ministry of Health and Social Welfare Male Circumcision Technical Working Group, the National AIDS Control Program, the HIV/AIDS (UNAIDS) recommended that countries with a high Iringa Regional Health Authority, and the health authorities of Iringa Municipal, prevalence of HIV and a low prevalence of male circumcision Iringa District, and Mufindi District played important roles in guiding this work. scale up VMMC within their comprehensive HIV prevention Technical staff from USAID was involved in study design, decision to publish, and preparation of the manuscript. programming [4]. WHO currently recommends a minimum package for VMMC services that includes group and individual Competing Interests: The authors have declared that no competing interests exist. education on VMMC, HIV risk reduction, and other male sexual Abbreviations: AE, adverse event; MCHIP, Maternal and Child Health Integrated and reproductive health issues; condom promotion and provision; Program; PEPFAR, United States President’s Emergency Plan for AIDS Relief; STI, HIV testing and screening; treatment of sexually transmitted sexually transmitted infection; UNAIDS, Joint United Nations Programme on HIV/ infections (STIs) and physiological abnormalities; and provision AIDS; USAID, United States Agency for International Development; VMMC, of VMMC under local anesthesia with postoperative observation voluntary medical male circumcision; WHO, World Health Organization and two to three postoperative visits (two, seven, and 42 days * E-mail: [email protected] postoperatively) [5]. WHO guidance on optimizing the volume . These authors contributed equally to this work. and efficiency of VMMC services [6] recommends the use Provenance: Submitted as part of a sponsored Collection; externally peer of multiple surgical beds per surgical team, ‘‘task shifting’’ the reviewed.
83 Summary Points Description of the Campaign N The government of Tanzania has adopted voluntary The Iringa Region VMMC campaign was conducted in three medical male circumcision (VMMC) as an important districts of the region between June 21 and July 31, 2010, to component of its HIV prevention strategy and aims to coincide with school leave, the end of the harvest season, and reach 2.8 million uncircumcised men within the next Iringa’s cool season (previous formative assessment revealed strong three years. community preference for VMMC during the cool season [12]). N In June and July 2010, a six-week VMMC campaign in VMMC services were provided Monday to Friday from 8:00 to Tanzania’s Iringa Region performed 10,352 circumci- 17:00, and Saturdays from 8:00 to 13:00, and were coordinated by sions. committees and teams acting at regional, district, and site levels. N Strategies adopted by the campaign to generate At the regional level, a committee of key stakeholders, which demand included the widespread dissemination of was formed three months before the campaign and included messages focused on the provision of free VMMC by regional health and administrative authorities, campaign manag- specially trained health care providers and on the HIV ers, and monitoring and evaluation experts, provided campaign prevention benefits of VMMC. oversight. The committee supervised VMMC facilities, provided N Clinical efficiency was improved through, for example, quality assurance of services, and oversaw the dispensing of the use of multiple beds in an assembly line, and the regular supplies of HIV test kits, condoms, and other consumable efficient use of staff time through task shifting and task commodities such as sutures, gloves, lidocaine, and antibiotics sharing. provided by MCHIP. MCHIP also provided additional surgical N The experiences of this campaign suggest that high- beds and instruments, and other infrastructure items to the volume VMMC can be performed without compromising facilities providing VMMC services during the campaign. client safety, and provide a model for matching supply At the district level, demand creation subcommittees were and demand for VMMC services elsewhere. composed of district officials, health facility staff, and international and community-based organizations. These subcommittees, which were tasked with recruiting clients and sensitizing community available, and identified the major deterrents to seeking VMMC political and administrative leaders, met several times before and as financial costs and having no family or cultural history of during the VMMC campaign. circumcision [8]. Finally, each campaign site had a management team com- In 2010, the government of Tanzania set a goal of 80% posed of the health facility’s medical officer in charge, who was VMMC coverage in its draft proposal ‘‘National Strategy for responsible for overall service delivery and quality at the site, and a Scaling Up Male Circumcision for HIV Prevention’’ [9]. To site campaign manager, who was responsible for daily reporting on achieve this goal, approximately four million circumcisions must and oversight of the VMMC service. Every evening during the be completed in the next five years, 2.8 million of them during campaign, site managers participated in a debriefing session with the initial three-year implementation stage. The Tanzanian campaign headquarters, during which problems with campaign strategy for VMMC scale-up prioritizes eight regions of relatively implementation, shortages of commodities, issues related to client high HIV and low male circumcision prevalence, with men aged demand, and AEs were addressed. 10–24 and 25–34 years as the primary and secondary priority These three levels of organization worked together to ensure the groups, respectively. The strategy permits task shifting and efficiency, quality, and safety of the campaign. In particular, as specifies that VMMC services should be free of charge within the described below, they followed the guidance provided by WHO public sector. for using human resources efficiently, for maximizing the Iringa Region, a largely rural region with a population of 1.9 throughput of clients, for detecting AEs, and for creating a million, has the highest adult HIV prevalence in Tanzania minimum VMMC package [5,6]. Demand creation innovations (15.7%) and relatively low circumcision coverage (29%) [7]. As were added to this package by MCHIP to match supply with part of the national strategy, the five-year target for VMMC in demand. Iringa Region is 264,990 circumcisions [9]; modeling estimates suggest that one HIV infection will be averted for every 4.5 Human Resources VMMCs performed in this region [10]. Iringa Region was one of The regional campaign committee developed a human the first Tanzanian regions to adopt VMMC as a part of its HIV resources plan that considered the WHO guidance for improving prevention program [11]. The Iringa regional health authorities, efficiencies, the number of surgical bays available, the expected in collaboration with the United States President’s Emergency client load, and available counseling and clinical staff. In this Plan for AIDS Relief (PEPFAR) and the Maternal and Child assembly-line service model, every four beds at a site required one Health Integrated Program (MCHIP), a project funded by the circumcising surgeon, four bed nurses, one or two equipment and United States Agency for International Development (USAID) commodities runners, an equipment cleaner, and two HIV and managed by Jhpiego (an affiliate of Johns Hopkins counselors. Regardless of the number of beds, each site also University), launched the region’s VMMC program in September required a receptionist, an autoclave operator, a janitor, and a 2009. By April 2010, five health facilities were providing routine, data manager. To reduce burnout, the clinical staff on the VMMC closely monitored VMMC services that covered three of the team, who were dedicated fully to VMMC throughout the region’s eight districts. campaign, rotated between roles (surgeon, bed nurse, and In June 2010, the Iringa Region VMMC program undertook its counselor). In addition, the most productive ‘‘surgeons’’ rotated first VMMC campaign. In this case study, we describe the among sites to encourage a spirit of collegiality. Other motivators campaign’s approach to service delivery and the factors that for providers included T-shirts, daily meals, and overtime pay influenced its quality, efficiency, and safety, and we discuss how equivalent to US$7.00 per day. The Ministry of Health and the experiences of this campaign might serve as a model for future Social Welfare paid staff salaries, with the exception of six VMMC campaigns in the Iringa Region and elsewhere. counselors working for local non-governmental organizations,
84 who were funded through other PEPFAR-supported programs. by peer educators and outreach workers through activities such as Finally, to ensure efficient and high-quality service provision, all traditional theater, small group sessions, one-to-one peer education, the nurses (who constituted 80% of the campaign’s VMMC and speeches to community groups. These community workers providers, including ‘‘surgeons’’), clinical officers, and physicians received no additional compensation or incentive for adding the involved in the campaign were trained using the WHO/ topic of VMMC to their usual activities. Brochures and other print UNAIDS/Jhpiego male circumcision manual [5], and were pro- materials were distributed that targeted specific audiences including vided with comprehensive on-site mentoring by a proficient adolescents and their guardians, female partners of potential VMMC physician. VMMC clients, and men aged 18 years and above. In addition, three radio advertisements promoting the campaign ran eight times Demand Creation per day on regional radio stations, and regional officials appeared Several approaches were taken to create demand for VMMC on local chat and health-related programs in the weeks leading up to during the campaign. Through the district-based demand creation the campaign. Finally, during the campaign, the regional health subcommittees, community-based organizations received a one-day authorities arranged for announcements about the availability of training session designed to enable the promotion of the campaign services to be made across facility catchment areas. Based on the
Figure 1. The layout of Ngome Health Centre. Before the campaign, this space was empty, having been built but not yet configured to function as a reproductive health facility. The Iringa Region team adapted the space for efficient VMMC service delivery by expanding the number of surgical bays (eight beds to accommodate two surgical teams), providing a large space for decontamination, increasing the number of individual counseling areas (including a tent to accommodate additional counselors), and including a separate postoperative area. doi:10.1371/journal.pmed.1001131.g001
85 national situation assessment, in which cost was identified as a Quality and Safety barrier to services [8], the demand creation messages emphasized As stipulated in WHO guidance, HIV testing was offered to all that VMMC services were now free and would be performed by clients on an opt-out basis during individual counseling. Consent specially trained health care providers. Additional messages focused for HIV testing was verbal (the standard of care in Tanzania); a on the HIV prevention benefits of VMMC. Demand creation guardian’s consent was required for clients under the age of 18. activities were scaled back by the fourth week of the campaign VMMC clients who tested HIV-positive received circumcision if because of overwhelming demand, and midway through the they were eligible for surgery (eligibility was based upon their campaign all sites began scheduling clients 1–2 weeks in advance, overall health and the absence of physiological abnormalities of except for clients traveling long distances, who were given same-day the penis or STIs that would otherwise preclude them) and gave services. consent (during the campaign, all clients provided written consent for surgery), but were counseled about the lack of HIV prevention Increasing Client Volume benefits for HIV-infected men and the increased risk of All of the efficiency considerations recommended by WHO transmitting HIV to sexual partners during postoperative healing. were adopted for increasing client volume, with the exception of All HIV-positive clients were referred to HIV care and treatment electrocautery (which was not endorsed for VMMC services in services, which were available at each circumcising site; clients Tanzania) and the use of disposable surgical instruments kits. with STIs were referred to STI treatment services at the same Specifically, time-saving surgical techniques, such as the forceps- health facility and counseled to return for circumcision after their guided method of circumcision, were used, and the number of treatment was complete; and clients with physiological abnormal- surgical kits was doubled at sites without an autoclave to minimize ities were referred to the district or regional urologist. the interruption of services caused by transfer of kits to autoclave- Infection prevention quality standards were applied at each of equipped sites for processing. To reduce client congestion during the campaign facilities, and the definitions for AEs in the WHO/ initial and follow-up visits and increase efficiency, the campaign UNAIDS/Jhpiego ‘‘Manual for Male Circumcision under Local provided additional trained counselors to prepare a large number Anaesthesia’’ were used to monitor AEs during the campaign [5] of clients for the surgery; added tents and other temporary During counseling, the VMMC team emphasized the importance structures at VMMC sites (Figure 1) to create additional space for of clients returning for their postoperative reviews, AEs were counseling, postoperative follow-up, and data entry; scheduled described to clients to encourage reporting, and each client clients up to two weeks in advance of surgery; paid special received details of a 24-hour emergency phone number staffed by attention to maintaining the motivation of participating health VMMC team clinicians. All AEs were recorded using the standard care workers; and developed campaign-specific data monitoring WHO form [5] and were discussed during the site managers’ tools. nightly debriefs. The regional committee received weekly briefings
Figure 2. Site-specific upward trends in VMMC service delivery during the campaign. The substantial increase in VMMC delivery at Ngome Health Centre (uppermost line) was due to the addition of more surgical bays mid-campaign. IRH, Iringa Regional Hospital. doi:10.1371/journal.pmed.1001131.g002
86 Table 1. VMMCs performed during the campaign, by site.
Number of VMMC Clients from Outside Number of Surgical Total a Site District Characteristics of Site Beds Teams VMMCs the Catchment Area N Percent
Iringa Regional Hospital Iringa Municipal Urban; regional referral hospital 4 1 1,784 1,114 62% Ngome Health Centre Iringa Municipal Urban; collaboration with 8 2 2,781 843 46% Iringa Regional Hospital to serve overflow of clients Lugoda Hospital Mufindi Rural; services aimed primarily 4 1 1,847 6 0.3% at tea plantation workers Mafinga District Hospital Mufindi Peri-urban; district referral 51b 1,896 1,874 33% hospital Tosamaganga Hospital Iringa Rural Rural; large, well-established, 4 1 2,044 1,394 68% and well-utilized facility Total — — 25 6 10,352 5,231 50%
aCatchment area defined as within 15 km of the facility. bWith extra nurse. doi:10.1371/journal.pmed.1001131.t001
on the occurrence of severe AEs and looked for trends to intensive monitoring and feedback. This system allowed each site determine whether there might be cause for concern at any one management team to receive nightly reports of clients prepared, site. clients circumcised, AEs, and other key client data. The Johns Throughout the campaign, client-level data on service delivery Hopkins University Institutional Review Board and the Tanzanian and AEs were entered daily by data clerks within each health Ministry of Health and Social Welfare approved the use of these facility team into an electronic web-based data system designed for data for this case study.
Figure 3. Postoperative return rates two and seven days after surgery. doi:10.1371/journal.pmed.1001131.g003
87 Figure 4. Age distribution of VMMC clients. doi:10.1371/journal.pmed.1001131.g004
Achievements of the Campaign which dedicated a large space to the VMMC campaign, achieved 2,781 circumcisions with eight beds and two surgical teams During the six weeks of the VMMC campaign, six circumcising (Table 1). Staff retention was high throughout the campaign. teams of 16 individuals, plus site managers, drivers, and data clerks Overall, 50% of clients came from outside facility catchment areas (140 participants in total) circumcised 10,352 adolescent and adult (defined as further than 15 km from the VMMC site), which males, 1.72 times the campaign’s target of 6,000 men. The average suggests that clients were willing to travel great distances for free number of clients served per week increased as the campaign and safe services. However, there was substantial variability across progressed (Figure 2), and all the VMMC sites served more than sites: the percentage of clients coming from outside facility 1,700 clients during the campaign period; Ngome Health Centre, catchment areas ranged from 0.3% at Mafinga District Hospital
Table 2. Adverse events during the campaign.
AE Occurrence Severity Total AEs Intra-operative Postoperative Mild/Moderate Severe
Damage to penis 1 11 Excess skin removal 3 33 Excessive bleeding 4 44 Swelling of the penis or scrotum 28 15 13 28 Infection 64 64 64 Total 8 92 79 21 100
doi:10.1371/journal.pmed.1001131.t002
88 to 68% at Tosamaganga Hospital (Table 1). Postoperative return ‘‘long-distance’’ caseload may have been related to client rates ranged from 67% to 80% of clients for the two-day follow-up, perceptions that a referral facility would provide higher quality and from 64% to 77% of clients for the seven-day follow-up services, a possibility that warrants further investigation. The high (Figure 3). Finally, almost 93% of the clients reached during the uptake of HIV testing seen during the campaign is not unusual— campaign were between 10 and 24 years old (Figure 4). Lugoda HIV testing acceptability is generally very high in Tanzania—but Hospital, which is on a tea plantation, had the highest percentage may have been magnified by a low perception of HIV risk among of clients in the 20- to 34-year-old age range (43%), presumably young (pre-sexual) clients, by the decreasing stigma associated with because of the large population of adult tea workers in its an HIV-positive status in the region, or by the knowledge that catchment area. HIV care and treatment services were available at the circumcis- The overall AE rate during the campaign was less than 1% ing sites. (Table 2). Eighty percent of AEs were minor or moderate; only 20% were categorized as severe. The most common intra- Safety operative AE was bleeding; swelling and infection were the most AE rates fell below those of normal service delivery (from just frequent postoperative AEs. All AEs were treated and clients were under 2% to 1%) during the campaign period [13]. This suggests healed or healing by the end of the campaign. that VMMC providers might become more proficient during high- HIV testing uptake was virtually universal (99%), and the volume campaigns, although increased oversight and supervision, overall HIV prevalence was less than 1% for the campaign clients, training by highly qualified VMMC mentors, and quality although it increased with age (Table 3). Of the 74 individuals who improvement exercises undertaken during the campaign period tested HIV-positive, all went on to be circumcised. Twenty-two may also have helped reduce AE rates. STI cases were identified during the campaign; all clients with STIs were referred for treatment and subsequently circumcised. Challenges Lessons Learned This case study highlights several major challenges for future high-volume VMMC campaigns. For example, it suggests that Efficiency ways will need to be found to improve the participation of older The data collected during the Iringa Region campaign indicate male clients. Only 24% of clients served during the Iringa Region that, using the efficiency model adopted by the campaign, a four- campaign were older than 20 years, and previous modeling has bed/one surgeon facility can circumcise up to 60 clients and an shown that for the greatest immediate public health impact, eight-bed/two surgeon facility can achieve 120 circumcisions per VMMC should cover the sexually active population [10]. Other day over a six-week period, and that the efficiency of VMMC service challenges for future campaigns that are revealed by this case study provision can increase over time. Importantly, the experiences include the possibility that demand may be higher than gained during this campaign indicate that, by transferring some anticipated, that there may be insufficient sites providing VMMC VMMC providers from larger sites to smaller sites, it is possible to services in rural regions, and that the procurement of commodities provide a high-volume service at small sites without greatly that are not readily available may require considerable lead times. impacting on the provision of normal health services provided at Furthermore, the case study illustrates how, in many developing these sites. The data presented in this case study also suggest that countries, surgical instruments, disposable commodities, and supply and demand for VMMC can be matched by focusing on pharmaceuticals may need to be imported from abroad, and community-driven demand, and by ensuring efficient site-level client reusable surgical instruments may have to be transported long flow by adding counselors as needed and by expanding the space distances over difficult terrain to and from facilities with available for VMMC with tents, careful scheduling, detailed logistics, autoclaves. Finally, it draws attention to how infrastructure issues, and the adoption of surgical efficiencies. electricity outages, and the geographic terrain may pose additional challenges in many developing countries. Quality The long distances traveled by many clients to receive services during the campaign (some clients stated that they traveled as far Conclusions as 100 km) was unanticipated and suggests that a well-motivated The Iringa Region experience shows that VMMC service population will travel long distances to VMMC sites during delivery can be provided to large numbers of men efficiently campaigns. However, the willingness to travel long distances without compromising quality of service and client safety through may also reflect a need for anonymity or client perceptions of a campaign mode of service delivery implemented almost service quality. For example, Iringa Regional Hospital’s high exclusively in the public sector. Although there are considerable challenges associated with implementing such campaigns, they are Table 3. HIV testing, by age group. not insurmountable, as this case study illustrates. Moreover, with contextualization, we suggest that similar campaigns could be replicated in other settings in east and southern Africa where Result Age VMMC for HIV prevention has been prioritized. Notably, since the completion of the Iringa Region campaign, 10–14 y 15–19 y 20–24 y 25–34 y $35 y expansion of VMMC services to the remaining five districts of the Negative 2,199 5,616 1,754 565 134 Iringa Region has become a priority. In December 2010, a three- Positive 10 10 9 31 14 week campaign that coincided with the school holidays resulted in Not tested 05500 nearly 3,000 clients being circumcised. By April 2011, all districts Total 2,209 5,631 1,768 596 148 in the Iringa Region were offering VMMC and, between June 20 and August 13, 2011, another eight-week campaign served 31,046 doi:10.1371/journal.pmed.1001131.t003 VMMC clients across the Iringa Region.
89 Acknowledgments oral poster presentation at the 6th International AIDS Society Conference on HIV Pathogenesis, Treatment and Prevention in July 2011. The authors would like to acknowledge the following people who played important roles in the implementation of the campaign and synthesis of Author Contributions campaign lessons learned: Ezekial Mpuya, Paul Luvanda, Menrad Dimoso, Caroline Kiame, Leonard Ndeki, Sekasua Mndeme, Rajabu Muhombo- Conceived and designed the experiments: HM BK KC MP TA AH EN lage, Benny Lugoe, Flora Hezwa, and the health providers of Iringa BF. Analyzed the data: HM BK MP MM MLD BF. Wrote the first draft of Region. We also wish to thank the Tanzania National AIDS Control the manuscript: HM. Contributed to the writing of the manuscript: HM Programme and USAID/Tanzania. BK KC MP TA AH SK SN MM MLD DC EN BF. ICMJE criteria for The campaign data were presented as a poster at the 18th Conference authorship read and met: HM BK KC MP TA AH SK SN MM MLD DC on Retroviruses and Opportunistic Infections in February 2011 and in an EN BF. Agree with manuscript results and conclusions: HM BK KC MP TA AH SK SN MM MLD DC EN BF.
References 1. Auvert B, Taljaard D, Lagarde E, Sobngwi-Tambekou J, Sitta R, et al. (2005) tacaids.go.tz/dmdocuments/THMIS%202007-08.pdf. Accessed 18 September Randomized, controlled intervention trial of male circumcision for reduction of 2011. HIV infection risk: the ANRS 1265 Trial. PLoS Med 2: e298. doi:10.1371/ 8. Tanzania National Institute for Medical Research, Tanzania Ministry of Health journal.pmed.0020298. and Social Welfare (2009) Situation analysis for male circumcision in Tanzania: 2. Bailey RC, Moses S, Parker CB, Agot K, Maclean I, et al. (2007) Male final report. Dar es Salaam (Tanzania): National Institute for Medical Research. circumcision for HIV prevention in young men in Kisumu, Kenya: a Available: http://malecircumcision.org/programs/documents/TanzaniaMale- randomised controlled trial. Lancet 369: 643–656. CircumcisionSituationAnalysis_September_09.pdf. Accessed 1 November 2011. 3. Gray RH, Kigozi G, Serwadda D, Makumbi F, Watya S, et al. (2007) Male 9. National AIDS Control Programme (2010) National strategy for scaling up male circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial. circumcision for HIV prevention: enhancing men’s role in HIV prevention. Dar Lancet 369: 657–666. es Salaam (Tanzania): National AIDS Control Programme. 4. World Health Organization, Joint United Nations Programme on HIV/AIDS 10. Njeuhmeli E, Forsythe S, Reed J, Opuni M, Bollinger L, et al. (2011) Voluntary (2007) New data on male circumcision and HIV prevention: policy and medical male circumcision: modeling the impact and cost of expanding male programme implications. Available: http://libdoc.who.int/publications/2007/ circumcision for HIV prevention in eastern and southern Africa. PLoS Med 8: 9789241595988_eng.pdf. Accessed 18 September 2011. e1001132. doi:10.1371/journal.pmed.1001132. 5. World Health Organization, Joint United Nations Programme on HIV/AIDS, 11. Tanzania Prime Minister’s Office, Regional Commissioner’s Office (2010) Iringa Jhpiego (2009) Manual for male circumcision under local anaesthesia, version regional action plan for the prevention of new HIV infections (July 2010–June 2.5. Geneva: World Health Organization. Available: http://www.who.int/hiv/ 2012). Iringa: Tanzania Prime Minister’s Office. pub/malecircumcision/who_mc_local_anaesthesia.pdf. Accessed 18 September 12. Plotkin M, Ku¨ver J, Curran K, Mziray H, Prince J, et al. (2011) Embe 2011. Halijamenywa: the unpeeled mango. A qualitative assessment of views 6. World Health Organization (2010) Considerations for implementing models for and preferences concerning voluntary medical male circumcision in Iringa optimizing the volume and efficiency of male circumcision services for HIV Region, Tanzania. Available: http://www.mchip.net/sites/default/files/ prevention. Geneva: World Health Organization. The%20Unpeeled%20Mango%20-%20Final%20Report%202011.pdf. Ac- 7. Tanzania Commission for AIDS, Zanzibar AIDS Commission, National Bureau cessed 25 October 2011. of Statistics, Office of the Chief Government Statistician, Macro International 13. Maternal and Child Health Integrated Program/Tanzania (2010) Male (2008) Tanzania HIV/AIDS and malaria indicator survey: 2007–08. Dar es circumcision service delivery data from Iringa region, November 2009 to May Salaam (Tanzania): Tanzania Commission for AIDS. Available: http://www. 2010. Dar es Salaam (Tanzania): Jhpiego.
90