The cost to circumcise Africa.Male circumcision employed as a prophylactic surgical intervention for HIV transmission reduction has been publicized in the media following recent results from observational trials conducted in Africa. Yet in all of the discussions concerning circumcision as a public health initiative, including a cost analysis performed on circumcision as a prophylactic for reducing HIV transmission in Africa, none estimates the endeavor's scope or cost. Given the scale of the economics involved in and the number of competing strategies available for addressing the HIV epidemic, funding and cost effectiveness are vital concerns in the field. This raises the question of which treatments and methodologies to fund, or not. In this study, we use circumcision costs, census, and demographic data available from government agencies and other published sources to estimate the cost to circumcise all HIV-negative African adult males, including costs of complications. We compare that cost to another androcentric penile alteration: using condoms (including their purchase and distribution costs). Our findings suggest that behavior change programs are more efficient and cost effective than surgical procedures. Providing free condoms is estimated to be significantly less costly, more effective in comparison to circumcising, and at least 95 times more cost effective at stopping the spread of HIV in sub-Saharan Africa. In addition, condom usage provides protection for women as well as men. This is significant in an area where almost 61% of adults living with AIDS are women.
Keywords: circumcision, HIV/AIDS, condoms, health costs, Africa
The recent announcement by UNAIDS regarding revised figures for the prevalence and incidence of HIV on a global scale has promoted discussion among various stakeholders on the best possible resource commitments and methods to expand access to both prevention and treatment for HIV/AIDS (UNAIDS, 2007a). With an established sense of urgency over the past decades, researchers have looked to identify novel and efficacious methods to reduce transmission. These range from established and proven methods such as education, programs focusing on high-risk individuals, and barrier-based methods, such as condom usage (Weller, 2003), to novel and still controversial (Dowsett, 2007) methods, such as male circumcision.
Sub-Saharan Africa remains the region with the highest number of infected persons and prevalence rate. Some observational studies (Weiss, 2000; Auvert, 2005) have claimed that circumcision offers protection to the male partner in the reduction of female-to-male transmission of sexually transmitted infections. Circumcision as a possible method to protect against HIV has generated much publicity (Myers, 2007). However, it is important from policy, practice, and efficacy viewpoints to analyze the cost effectiveness of circumcision versus that of a proven benchmark such as condom usage in combination with social intervention practices.
The magnitude of the problem of AIDS in sub-Saharan Africa is illustrated by Table 1, compiled from UNAIDS (2007a) and United States census data (Velkoff, 2007). The prevalence of HIV in sub-Saharan Africa among adults has been reduced to 5.0% [95% CI: 4.6-5.5%], down from 6.1% [95% CI: 5.4-6.8%] in 2006 (UNAIDS, 2007a). Given the magnitude of this problem, multiple approaches to reduce incidence and impact can and must be tried, but with a view to maximizing resource application effectiveness and efficiency.
Male circumcision, a medically and ethically controversial procedure (Dowsett, 2007), has been proposed as a mode of reduction of HIV transmission from women to men (Williams, 2006). A study examining the cost-effectiveness of this surgical procedure concluded that, in the environment in which one of the studies supporting circumcision was conducted, the procedure was a cost-effective approach (Kahn, 2006). However, questions remain about the risk conclusions arrived at in that study (Kalichman, 2007). Based on the available clinical evidence, circumcision cannot be shown to prevent transmission to women, transmission of HIV among men who have sex with men, transmission via intravenous drug use, or pre-, peri-, or post-natal transmission (Dowsett, 2007; Millett, 2007). The conclusions need to be extended to a continental scale in Africa, rather than a localized provincial level. Cost estimates, risk, and prevalence factors need to be updated in light of more representative data.
Even if all African adult males were circumcised today (and its real-world effectiveness to stop HIV proves to be as high as purported in recent controlled studies), over the next 10-year period it would reduce the number of HIV cases in sub-Saharan Africa by only eight percent, with a one percent reduction of deaths (Williams, 2006).
In this evaluation, we applied the methods developed by Kahn (2006) and extended them from a local application in a single province (Gauteng, South Africa) to a sub-Saharan environment. In Kahn's study, cost-effectiveness was modeled for 1,000 circumcisions performed within a general adult male population. Intervention costs included performing circumcision and treatment of adverse events. HIV prevalence was estimated from published estimates and incidence among susceptible subjects assuming a steady-state epidemic. Effectiveness was defined as the number of HIV infections averted.
We estimated costs per procedure for male circumcision and updated these using complication rates reported in an environment where higher rates of circumcision prevail and trained medical practitioners (doctors and nurses) and traditional providers performed the procedure (Okeke, 2006). Calculations and comparisons used circumcision costs, census, and demographic data available from government agencies and published sources to estimate the cost to circumcise all HIV-negative, African adult males, including subsequent complication costs, using available data. We compared that cost to another androcentric method of penile alteration: using condoms (including their purchase and distribution costs). To facilitate consistent comparison, we gratefully adopted a number of the procedure ranges and assumptions given in Kahn (2006). However, the complication rate cited for short-term mildly adverse events, short-term adverse events, and long-term events required updating.
Circumcision complication costs should be factored into a realistic cost-effectiveness estimate since they are a real cost of such a plan. In developed countries, complication rates may be from 20.2% and as high as 55%, as one longitudinal study concluded (Patel, 1966). However, the rate varies greatly due to differences in what conditions observers choose to identify as complications and over what time frame they observed the patient. Complications are more frequent in rural clinics or when performed by traditional circumcision practitioners. A study of mass circumcision practices with more comprehensive criteria for identifying complications revealed significantly increased complication rates with mass circumcisions: 15.7% from traditional practitioners, versus 3.8% for medically-trained professionals (Ozdemir, 1997).
We reviewed the literature and found the situation described by Okeke (2006) to be more representative of possible outcomes. While Okeke reports on neonatal circumcisions, a higher rate of complication than one in which all circumcisions are performed by trained medical practitioners is more reflective of potential outcomes and risks. A high rate of circumcision (87%) was reported, with 9% of procedures being performed by traditional practitioners and the remainder by doctors or nurses. Okeke reported on complications such as redundant foreskin, excessive loss of foreskin, skin bridges and amputation of the glans penis. A case of severe hemorrhage requiring transfusion was also reported. Interestingly, Okeke reported no statistically significant (p = 0.051) difference in complication rates between nurses, doctors, and traditional practitioners. This complication rate does not include several more complications listed by Van Howe (2004) in his study estimating complications and costs in a developed environment. Table 2 summarizes our cost inputs and findings.
Our updated analysis suggests that the cost per procedure could increase up to 31% if higher rates of complications are encountered. This first-order complication risk-weighted cost per circumcision procedure of $73.07 may be more representative of conditions that prevail if such a program is instituted continentally. Such an increase in complication rates must be taken into account in evaluations of cost-effectiveness. It must be noted that this estimate is more conservative than the 425% lifetime incremental cost per patient calculated by Van Howe.
Prevalence measures include everyone living with an infectious disease and present a delayed representation of the epidemic by aggregating the new incidences over many years. By contrast, incidence measures the number of new infections, usually over the previous year. Following Kahn (2006), again for the purposes of consistency, for our study effectiveness was defined as the number of HIV infections prevented per cohort of 1,000 men over a period of 20 years. Effectiveness was calculated as the product of the number of HIV-susceptible persons, the HIV-incidence rate, the protective effect of circumcision (adjusted for risk compensation), the projection period (in years), and an epidemic multiplier. Mathematically, this may be represented as:
Effectiveness = Cohort Population x (1 - HIV Prevalence) x Incidence Rate x Net Protective Effect x Projection Period x Epidemic Multiplier, where Net Protective Effect = [1 - (1 - Protective Effect) x (i + Risk Compensation)].
Estimates of incidence do need to take into account the quality and reliability of data available, especially on a sub-Saharan scale. Kahn et al. (2006) solved this problem by estimating incidence from a steady-state model of prevalence and performing sensitivity analysis on prevalence levels.
A key factor driving conclusions of cost effectiveness was the high rate of HIV prevalence (25.6%) among adult men in the Gauteng province studied (Dorrington, 2002). A critical re-evaluation of this parameter is necessary in light of recent UNAIDS data (2007a). While prevalence rates vary widely, a mean prevalence rate of 5% published for the continent was used in this study as a representative value (Table 1, above). Incidence rates may be estimated in two different ways. Using the steady state model yields an incidence rate of 1.05 per 100 person-years (0.0105). Estimating incidence from the UNAIDS data, accounting for the effect of the significant fraction of the population already infected, yielded a significantly lower incidence rate of 0.0023. Both cases were modeled on the illustration in Table 3. While derivative, the use of the Kahn model's ab initio assumptions, corrected for UNAIDS 2007 projections, strengthened a conclusion of Kahn's own study, where cost-effectiveness of the circumcision procedure varied significantly with prevalence and incidence rate assumptions.
Further sensitivity analysis was performed by conservatively varying the protective effect of MC, which is estimated to lie between 0.34 and 0.77 by supporting studies (Auvert, 2005). We did not include the case of statistically insignificant protective effects reported by Millett (2007). Using incidence derived from Kahn's (2006) steady state model and a prevalence rate of 5% (UNAIDS, 2007a), we estimated between 52.4 and 213.2 cases of HIV prevented. Using actual UNAIDS data for incidence yielded a predictably lower effectiveness range of 11.6 and 47.3 cases prevented. Figure 1 summarizes the significant effects that updating these variables has on the cost effectiveness of applying circumcision as a therapeutic prevention modality, based on the variables above. The results suggest that significant re-thinking of the conclusions is necessary when applying the localized findings of the Kahn study on a continental scale.
Let us consider, however, the cost-viability of circumcision versus that of promoting consistent condom usage. In the 44 countries that comprise sub-Saharan Africa, there are 194 million adult males (CIA Factbook, 2007). Of those, 9.1 million men are HIV positive (UNAIDS, 2007a) and about 62% have already been circumcised for cultural reasons (NIAID, 2006). Therefore, approximately 69.9 million uncircumcised, HIV-negative men remain.
Using the $73.07 risk-adjusted circumcision procedure cost calculated by our analysis yielded an estimate of $5.1 billion for performing the procedure for all men in the sub-Saharan region. In order to evaluate the relative impact of this cost, we note the following: $5.1 billion is 27 times the $188 million in 2006 UNAIDS Core Contributions budgeted to fight HIV worldwide (UNAIDS, 2007b). Yet, this estimate has omitted consideration of several additional costs, such as those required to standardize safety and to educate professional and traditional circumcision practitioners. In the sub-Saharan region, a project of this scale would certainly involve public funding, as opposed to private funding for procedures.
[FIGURE 1 OMITTED]
In addition, maintaining the program by circumcising all HIV-negative males as they turn age 15 would involve additional expenses of about $763 million per year. Rural African clinics have insufficient supplies and training to perform aseptic circumcisions. Clinics are currently overburdened in providing circumcision. South Africa and Lesotho report circumcision waiting lists of 6-8 months, and Zambian clinics are falling behind demand. Mass circumcision adoption, including community publicity, would increase demand. Of clinics surveyed, two-thirds did not have a selection of forceps for performing circumcisions. Seventy percent did not have the suturing needles or curved scissors required for adult circumcisions. Perhaps most important, only one-third of surveyed clinics had equipment to keep instruments sterile (Mattson, 2004). Other factors likely to hamper circumcision implementation would include financial concerns, lack of skilled operators, and cultural traditions (Pinock, 2007).
Given the seriousness of the HIV/AIDS crisis and the economic situation in Africa, it is critical to determine the simultaneously least expensive and most effective combination of preventive measures. Since condoms and circumcision both represent male-focused interventions, one permanent and one not, we need to compare these two methods of prevention to illustrate the relative cost of any mass circumcision program.
Regardless of prevalence, a comparison of the effectiveness of circumcision to proven methods, such as use of condoms, is conclusive. Condom usage picks up where the effectiveness range of circumcision leaves off, with a proven effectiveness equivalent to that of preventing pregnancies. Consistent use of condoms results in 80% reduction in HIV incidence (Weller, 2003; de Vincenzi, 1994). Latex condoms, when used consistently and correctly, are highly effective in preventing the sexual transmission of HIV, the virus that causes AIDS and other sexual transmitted infections (Feldblum, 2003; Center for Disease Control, 2002).
Condoms for free distribution cost $0.03 each (Ngwa, 2006). African males require an average of 84 condoms per year (Myer, 2002). For the undiscounted cost of one circumcision (with an associated limited protective effect), a man can receive a 29-year supply of condoms and protect himself and his partner 87-100% of the time (Mattson, 2004; Weller, 2003).
Free condoms need only be provided to poverty-level African men since people above the poverty line can afford to purchase them. World Bank estimates place approximately 46% of sub-Saharan Africans below the poverty line. As a result, a free-condom program for all adult males would cost $224 million per year. This is in stark contrast to the $763 million annually needed to circumcise sub-Saharan African children under age 15 alone.
Missing from the observational studies (Auvert, 2005) supporting circumcision was the "number needed to treat," which is conventionally provided as an effectiveness ratio for treatments. We calculated a cumulative circumcision number needed to treat using data from three studies and found it to be 80. Using data from a meta-analysis (Davis, 1999), where condoms were shown to be 87% effective at stopping HIV when used consistently, we calculated their number needed to treat to be 1,568.
Eighty circumcisions, at a total cost of $5,845.00, would have to be performed to prevent one HIV infection, while the use of 1,568 condoms would prevent the same infection at a cost of $47.00, making condom use 95 times more cost-effective than circumcision. It is worth noting that these two alternatives are not wholly comparable. Providing condoms protects at any level of implementation. The correct use of 10 million condoms provides ten times the effect of one million condoms, and there is no foreseeable negative repercussion of partial implementation. However, the same is not true for circumcision.
The primary arguments against the use of condoms appear to be inconsistency in usage, either due to inability of a sexual partner to insist on usage (Odutolu, 2005) or religio-cultural resistance (Gyimah, 2006; Garner, 2000). However, these arguments may fail to take into account two factors. First, as shown by Talbott (2007), when conducted properly, cross-country regression data do not support the theory that male circumcision is the key to slowing the AIDS epidemic. Rather, it is the number of infected prostitutes in a country that is more significant and robust in explaining HIV prevalence levels across countries. Secondly, as evidenced by the examples of Uganda and Thailand, condom use is central to the prevention of STDs, including HIV among sexually active populations. Social marketing campaigns can play a vital role in promoting usage, together with the increased awareness about AIDS (Finger, 1998; Najjumba, 2003).
Circumcision can hardly be compared with a "vaccine" against HIV, and its high cost makes it a questionable preventative. Behavior change programs, not surgery, are more efficient and cost-effective until such time that an even better method is developed and approved for use. Providing free condoms is estimated to be significantly less costly, more effective in comparison to circumcising, and at least 95 times more cost-effective at stopping the spread of HIV in sub-Saharan Africa. In addition, condom usage provides protection for women as well as men. This is significant in an area where almost 61% of adults living with AIDS are women (UNAIDS, 2007a). Before circumcision programs are created and funds are raised, their costs should be compared to other AIDS prevention programs so that a rational decision-making process, to spare as many lives as possible, is employed.
In order to evolve the best possible allocation of funds to benefit sub-Saharan Africa, fellow investigators should direct their work towards more clearly establishing the increased risk of male-to-female sexually transmitted infections transmission in the post-operative period following circumcision. In addition, a better understanding of higher risk behavior by men following circumcision could improve understanding of the prophylactic abilities of the procedure. Following the suggestions of Siegfried (2005) and others, careful consideration of ethical issues involved as well as regional and national differences in culture, religion, and social norms should also be undertaken before implementing such programs.
Auvert, B., Taljaard, D., Lagarde, E., Sobngwi-Tambekou, J., Sitta, R., et al. (2005). Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: The ANRS 1265 Trial. PLoS Medicine, 2, e298, doi: 10.1371/journal.pmed. 0020298.
Center for Disease Control (CDC). (2002). Male latex condoms and sexually transmitted diseases. Atlanta: CDC. Retrieved January 25, 2008, from http://www.cdc.gov/nchstp/od/ latex.htm
Central Intelligence Agency (CIA). (2007). CIA Factbook: Estimated population for 2007. Washington, DC.
de Vincenzi, I. (1994). A longitudinal study of human immunodeficiency virus transmission by heterosexual partners. New England Journal of Medicine, 331(6), 341-346.
Davis, K. R., & Weller, S. C. (1999). The effectiveness of condoms in reducing heterosexual transmission of HIV. Family Planning Perspectives, 31(6), 272-279.
Dorrington, R. E., Bradshaw, D., & Budlender, D. (2002). HIV/AIDS profile of the provinces of South Africa--Indicators for 2002. Capetown: Centre for Actuarial Research, Medical Research Council and the Actuarial Society of South Africa.
Dowsett, G. W., & Couch, M. (2007). Roundtable: Male circumcision and HIV prevention: Is there really enough of the right kind of evidence? Reproductive Health Matters, 15(29), 3344.
Feldblum, P. J., Welsh, M. J., & Steiner, M. J. (2003). Don't overlook condoms for HIV prevention. Sexually Transmitted Infections, 79, 268-269.
Finger, W. R. (1998). Condom use is increasing. Network, 18(3), 20-23.
Garner, R. C. (2000). Safe sects? Dynamic religion and AIDS in South Africa. Journal of Modern African Studies, 38, 41-69. doi:10.1017/S0022278X99003249
Gyimah, S. O, Takyi, B. K., & Addai, I. (2006). Challenges to the reproductive-health needs of African women: On religion and maternal health utilization in Ghana. Social Science & Medicine, 62(12), 2390-2944.
Kahn, J. G., Marseille, E., & Auvert, B. (2006). Cost-effectiveness of male circumcision for HIV prevention in a South African setting. PLoS Medicine, 3(12), e517. doi:10.1371/journal. pmed.0030517
Kalichman, S., Eaton, L., & Pinkerton, S. (2007). Circumcision for HIV prevention: Failure to fully account for behavioral risk compensation. PLoS Medicine, 4, e138. doi:10.1371/journal.pmed.0040138
Mattson, C.L., Muga, R., Poulussen, R., Onyango, T., & Bailey RC. (2004). Feasibility of medical male circumcision in Nyanza Province, Kenya. East African Medical Journal, 81, 230235. Retrieved January 25, 2008, from http://www.unfpa.org/news/news.cfm?ID=837.
Millett, G. A., Ding, H., Lauby, J., Flores, S., Stueve, A., et al. (2007). Circumcision status and HIV infection among Black and Latino men who have sex with men in 3 US cities. Journal of Acquired Immune Deficiency Syndrome, 46(5), 643-50.
Myer, L., Mathews, C., & Little, F. (2002). Condom use and sexual behaviours among individuals procuring free male condoms in South Africa. Sexually Transmitted Diseases, 29, 239241.
Myers, A., & Myers, J. (2007). Male circumcision--the new hope? South African Medical Journal, 97, 338-341.
Najjumba, M. I., Ntozi, J., Ahimbisibwe, E. F., Odwee, J., & Ayiga, N. (2003). Risk perception and condom use in Uganda. African Population Studies/Etude de la Population Africaine, 18(1), 67-80.
NIAID-National Institute of Allergy and Infectious Diseases. (2006). Q&A, NIAID-sponsored adult male circumcision trials in Kenya and Uganda. Washington: United States Department of Health and Human Services. Item 11. Retrieved January 25, 2008, from http://www3.niaid.nih.gov/news/QA/AMC12_QA.htm.
Ngwa, G. (2006). Condoms are a key to reducing HIV transmission in Zimbabwe. Geneva: United Nations Population Fund. Retrieved January 25, 2008, from http://www3.niaid.nih.gov/news/QA/AMC12_QA.htm
Odutolu, O. (2005). Convergence of behaviour change models for AIDS risk reduction in subSaharan Africa. International Journal of Health Planning and Management, 20(3), 239-252.
Okeke, L. I., Asinobi, A. A., & Ikuerowo, O. S. (2006). Epidemiology of complications of male circumcision in Ibadan, Nigeria. BMC Urology, 6, 21.
Ozdemir, E. (1997). Significantly increased complication risks with mass circumcisions. British Journal of Urology, 80, 136-139.
Patel, H. (1966). The problem of routine infant circumcision. Canadian Medical Association Journal, 95, 576-581.
Pinock, S. (2007). Workforce biggest barrier to roll-out of male circumcision. Lancet, 370(9602), 1817-1818.
Siegfried, N. (2005). Does male circumcision prevent HIV infection? PLoS Medicine, 2(11), e393.
Talbott, J. R. (2007). Size matters: The number of prostitutes and the global HIV/AIDS pandemic. PLoS ONE, 2(6), e543. doi:10.1371/journal.pone.0000543
UNAIDS. (2007a). AIDS Epidemic Update: December 2007. Joint United Nations Programme on HIV/AIDS (UNAIDS) and World Health Organization.
UNAIDS. (2007b). Core Contributions. Washington: UNAIDS. Retrieved January 25, 2008, from http://data.unaids.org/pub/BaseDocument/2007/core_2006_en.pdf?preview=true
Van Howe, R. S. (2004). A cost-utility analysis of neonatal circumcision. Medical Decision Making, 24(6), 584-601.
Velkoff, V. A., & Kowal, P. R. (2007). U.S. Census Bureau, Current Population Reports, P95/071, Population Aging in Sub-Saharan Africa: Demographic Dimensions 2006. Washington: U.S. Government Printing Office.
Weller, S., & Davis, K. (2003). Condom effectiveness in reducing heterosexual HIV transmission (Cochrane Review). In: The Cochrane Library, Issue 4. Chichester, UK: John Wiley & Sons, Ltd.
Weiss, H. A., Quigley, M. A., & Hayes, R. J. (2000). Male circumcision and risk of HIV infection in sub-Saharan Africa: A systematic review and meta-analysis. Journal of Acquired Immune Deficiency, Syndrome, 14, 2361-2370.
Williams, B. G., Lloyd-Smith, J. O., Gouws, E., Hankins, C., & Getz, W. M. (2006). The potential impact of male circumcision on HIV in sub-Saharan Africa. PLoS Medicine, 3(7), e262. doi: 10.1371/journal.pined.0030262
RYAN G. MCALLISTER (a) JOHN W. TRAVIS (b) DAN BOLLINGER (c) CLAIRE RUTISER (c) VEERARAGHAVAN SUNDAR (c)
(a) Georgetown University
(b) Wellness Associates
(c) Independent Researchers
Ryan G. McAllister, Georgetown University; John W. Travis, Wellness Associates; Dan Bollinger, Independent Researcher; Claire Rutiser and Veeraraghavan Sundar, Tott Inc.
Correspondence concerning this article should be addressed to Ryan G. McAllister. Georgetown University, 506 Reiss Science Building, 7th and O St. NW, Washington, DC 20057. Electronic mail: email@example.com
Table 1 Comparison of 2007 HIV/AIDS Statistics for Sub-Saharan Africa and the World Population Persons living New HIV with HIV infections /AIDS Sub-Saharan Africa 752.8 22.5 1.7 World 6,525 33.2 2.5 Adult Deaths prevalence due to AIDS Sub-Saharan Africa 5.0 1.6 World 0.8 2.1 Note. Absolute figures in millions. Table 2 Cost Inputs and Complication Factors in Male Circumcision * Input Kahn, 2006 This Work Ref. (a,b,c) Cost per circumcision 54.72 54.72 a Number of circumcisions performed 1,000 1,000 a Frequency of short-term adverse events (outpatient) 0.037 0.143 b Cost per short-term mild adverse event (outpatient) 13 13 a Frequency of short-term adverse events (inpatient) 0.0013 0.049 b Cost per short-term adverse event (inpatient) 334 334 a Frequency of long-term adverse events 0.0093 0.0093 b Cost per long-term adverse event 13 13 a Pooled costs per 1000 circumcisions: Cost of circumcisions 54,720 54,720 c Short-term mild adverse events (outpatient) 481 1,859 c Short-term adverse events (inpatient) 434 16,366 c Long-term adverse event 121 121 c Net cost 55,756 73,066 c * All cost values are given in US$ (a) Kahn, Marseille and Auvert (2006) (b) Okeke, Asinobi and Ikuerowo (2006) (c) = calculated Table 3 Risk and Incidence Weighted Effectiveness of Circumcision as a Procedure Location Gauteng Province, Sub-Saharan South Africa Africa steady-state model HIV Prevalence 0.256 (a) 0.05 (b) 1- Prevalence 0.744 0.95 Cohort Size, circ. 1,000 1,000 Incidence Rate 0.0380 ([dagger]) 0.0023 * Protective Effect 0.6 0.6 Risk Compensation 0.25 0.25 Net Protective Effect 0.5 0.5 Projected Period 20 20 Epidemic Multiplier 1.5 1.5 Effectiveness 424.08 33.16 Location Sub-Saharan Source Africa UNAIDS data model HIV Prevalence 0.05 (b) 1- Prevalence 0.95 Cohort Size, circ. 1,000 Assumption Incidence Rate 0.0105 ([dagger]) Calculated Protective Effect 0.6 Kahn, 2006 Risk Compensation 0.25 Kahn, 2006 Net Protective Effect 0.5 Calculation Projected Period 20 Assumption Epidemic Multiplier 1.5 Kahn, 2006 Effectiveness 149.63 Calculated (a) Kahn, Marseille and Auvert (2006) (b) UNAIDS (2007a) ([dagger]) Calculated using steady-state model (Kahn, 2006). * Estimated from UNAIDS data (2007).