12 mars, 2024

Critical appraisal of the Cost-Effectiveness of Male Circumcision for HIV Prevention in a South African Setting

Male circumcision has been diffusely promoted on the basis of its beneficial effects on male's health.
However, these praises do not have a solid scientific basis.
The study carried out in 2005, named: First Trial of Male Circumcision against HIV, was promoted as essential proof of the effectiveness of circumcision and yet a critical reading of the method reveals anomalies which cause significant bias in the analysis.

The first Randomized Controlled Trial (RCT) about male circumcision (MC) in HIV infection took place in Orange Farm: a semi urban area in South Africa (1). 
3,274 young, heterosexual uncircumcised men joined the study (2), they were divided in two groups: intervention and control groups. 
Men belonging to the intervention group had a circumcision and were commanded not to perform sexual intercourses for six weeks after the operation, while the control group haven’t had any operation nor restriction in the beginning of the study and were supposed to have MC at the end of it (1). 
Both groups were asked to have periodic clinical visits during which they would be examinated and are asked to provide information about their sexual lives. They would also have counselling sessions about STIs with a focus on HIV prevention. 
All participants were encouraged to have voluntary HIV testing in the dedicated centers. 
During the study, the number of new infections in the control group was higher than in the intervention group (49 versus 20) (2). 
Following these results, the study has been interrupted under the advice of the Data and Safety Monitoring Board and the researchers proposed circumcision to all participants (2).
These results have been considered in favor of the use of circumcision as an effective preventive measure against HIV infection amongst men.
In the continuity of this study, an article (3) was published by PLoS Medicine in 2006, where the authors used available data from the Orange Farm RCT and some observational studies, combined with epidemiological inputs to evaluate the cost-effectiveness of MC in order to advocate for a wider adoption (3).
The study targets the government of South Africa and takes a public health perspective for a period of 20 years (3).  The client population was made of 1000 male adults older than 18 years (3).
We will start by summarizing the Cost-effectiveness analysis of MC for HIV prevention in South African setting: we will go through the inputs, outputs and results of the study. Then we will assess this economic evaluation.

The Cost-Effectiveness of Male Circumcision for HIV Prevention in a South African Setting:

Cost inputs:
According to the study, MC should be provided by the health system in South Africa and / or other African countries that would have a high HIV prevalence and a low MC coverage (3). 
Authors also advocate for a large implementation through time (3).
The amount of each resource was derived either from the OF trial or from other studies.
Cost inputs were financial costs that included: Direct health care costs like the cost of a Health program performing MC and providing treatment of its adverse effects and the lifetime cost of HIV treatment based on a study in South Africa with and without antiretroviral treatment (ART) and Recurrent costs like Community Publicity cost (3). 
Direct non health costs haven’t been taken into consideration since MC programs are designed to be grafted on existent health programs using the same staff, without any need for specific training (3). 
Indirect costs endured by population haven’t been taken into consideration since this evaluation doesn’t take a societal approach. In the OF trial, participants received an amount of money (2) that would compensate fees during the trial, but this wasn’t taken into consideration during this evaluation.
The cost inputs were adapted to inflation of cost using the US consumer price index, which measure the change in prices in urban setting in the USA (4). The use of this method in a semi-urban setting in a sub-Saharan context (1) could be questioned.

Epidemiological inputs
These inputs were either derived from the OF trial or estimated by the authors with confidence intervals that would integrate wide-ranging fluctuations or changes (3). The list of these inputs is as follow:
-Proportion of persons susceptible to HIV infection in the general population (all HIV negative persons)
-HIV incidence rate
-Estimated protective benefit of MC for males based on the OF trial results.
-Increased risk due to risk compensation observed in men with MC, tending to have a riskier behavior (including multiple partners)
-Adjustment of HIV infection due to epidemic dynamic or epidemic multiplier (including HIA in women and children due to the protective effect of MC tickling down).

Sensitivity analysis:
In order to assess inputs values, multiple sensitivity analysis have been conducted.
First, multiple one-way sensitivity analysis (3) were performed for fluctuations in cost inputs, protective benefit of MC, risk compensation and epidemic multiplier.
Second, a 3-way sensitivity analysis (3) was conducted for the changes in MC cost, protective benefit of MC and the adjustment for HIV infection due to epidemic dynamic.
Third, a Threshold analysis (3) was performed for the variations of MC cost and risk compensation.
Finally, Multivariate Monte Carlo Simulations (3) have been conducted to have a comprehensive overview about outputs within uncertainty from all inputs.
In total, three scenarios have been selected including (3):
-Variation in epidemic settings
-variations in clients characteristics 
-Coverage of MC or variation of acceptability and uptake.

The outputs: 
For the cost-effectiveness analysis of MC, authors defined effectiveness (3) of MC as the number of HIV infections averted per 1000 newly circumcised men within 20 years after the intervention.
In order to compare effectiveness of MC to the negative effects of adverse events of circumcision, authors proceeded to the calculation of the net DALYs through indirect methods using pre-existing values and assuming a good acceptability of the intervention based on observed and stated preferences (3).

Results of the Cost-effectiveness analysis:
Assuming that there is a common unit of effect, MC was compared to other effective interventions using the cost per HIV infection averted (HIA).
The table 1 summarizes Cost per HIA for the different mentioned interventions.
Table 1: Cost per HIA for the different HIV prevention interventions

According to this cost-effectiveness analysis, MC seems to be part of the recommended interventions aiming HIV prevention in men, with a net DALY reduction; since, the study reports that even with pessimistic assumption about the negative effect of MC adverse effect, the net DALYs was in favor of the provision of MC, taking into consideration the benefit of averted HIV infection. 
The authors of the article took as a fait accompli the effectiveness of male circumcision in combating HIV infection. They carried out an economic study of the application of this intervention, presenting the reduction in the number of new HIV infections as directly due to MC.
Finally, authors seem to be advocating for an independent MC program outside the framework of the HIV control and prevention, designing MC services a ‘Portal’ for men to access sexual and reproductive services (3).

Assessment of the economic evaluation of MC for HIV prevention:
In the face of incurable diseases, Western doctors, especially Americans, often have recourse to surgical operations, including circumcision (5).
The orifical surgical society funded in 1890 in Chicago, used to provide trainings to perform this type of surgery (5). Therapeutic powers are often attributed to circumcision: the review of the orifical surgical society publications finds various indications for circumcision (5): headache, scoliosis in children, joint disease, hydrocephalus (6) ...

To date, there is no biological link that could explain a relationship between reducing the risk of HIV infection for humans and circumcision. Some hypotheses exist but aren’t confirmed (7).
 Data reported from studies about the distribution of HIV infection and the circumcision phenomenon have been widely criticized. These studies neglected other important social data such as the age of the first sexual relation, polygamy, the practice of female genital mutilation but also socio-economic status, education and culture (5).

These studies have also been only of interest in HIV epidemic contexts in underdeveloped countries and few studies report the relationship between MC and HIV infection in developed countries. The USA have the highest rate of circumcised sexually active men and yet has one of the largest HIV epidemics in developed countries (5). In addition, European countries with the highest rates of HIV infections are those with larger communities of migrant populations, including the largest number of circumcised Muslim immigrants (5).

Even if the OF trial was described as statistically robust (8), it is questionable since the study was interrupted for ethical issues and had many dropouts (9). Indeed, many specialists believe that its results must be confirmed by other studies. Indeed, there are criticism about methodology and ethics (10). OF trial does also applies “western standards in a non-western context” (9) which is also largely debatable.
In addition, the acceptability of circumcision seems to have been very positively estimated in this cost-effectiveness analysis. Literature finds that in Sub-Saharan Africa, there are many different tribes, some of which hold very strong cultural emotions against circumcision (11).
Furthermore, circumcision as per se, imposes an ethical issue since the removal of healthy normal erectile tissue is arguable (12, 13). 

We can conclude that the estimated protective benefit of MC for males isn’t so clearly confirmed.
When it comes to the design of the cost-effectiveness analysis, some facts are to be mentioned:
- There are some points to be highlighted about sampling since the study doesn’t take into consideration men already circumcised, that are reported to constitute 35% of the male population (3). This economic evaluation only considerate young men aged more than 18 years, knowing that sexual life starts at a younger age.
- The definition of the main outcome is also questionable since it supposes that all averted HIV infections are directly due to MC, taking out all the effect of social determinants for HIV infection (3).
When it comes to the results of the analysis:
-The epidemic multiplier is based on the assumption that the protective effect of MC would tickle down to women and children, knowing that data suggests little or no protection for the partners of men with MC (14), it isn’t clear how the authors estimate that in 20 years there would be 2/3rd HIA in male and 1/3rd in women.
-MC is presented as a pretty effective measure that cannot be provided alone and needs to be added to the ongoing interventions against HIV. There are no competing alternatives since MC cannot have an added value by its own as it’s shown in OF trial where circumcised men that miss their medical check up (where they have counseling) do have the same risk of HIV infection as the uncircumcised ones (2).
-The authors position themselves for MC and against education in the sub-Saharan context on the basis of economic efficiency; which can be perceived as a colonial and racist approach.
-Another economic evaluation about MC published in 2005, shows that in order to have one HIA, there is a need to circumcise 23 148 children which would cost then 9.6 million dollars (5). This evaluation defines the outcome differently, showing low effectiveness of MC in HIV prevention when provided in a younger population questioning the long-lasting protective effect mentioned in the different pro-MC studies.

The authors of this cost-effectiveness analysis tried to advocate for a large adoption of MC in sub-Saharan context through demonstrating a high cost-effectiveness of this intervention in preventing HIV infections.
The economic analysis was well designed but the studied question wasn’t well-defined since the program’ effectiveness is still questionable because it was based on studies that are subjects to many critics, that the outcome doesn’t have a direct proven relationship with the main output which is MC and that the effect of other socio-economic determinants haven’t been taken into consideration.
Many organizations around the world are against male circumcision, define it as a genital mutilation and believe that there is a political agenda promoted by the USA to spread genital mutilation in African countries (15).

[1] NCBI. 2005. First Trial of Male Circumcision against HIV. [online] Available at: <https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1262557/> [Accessed 6 January 2022].
[2] Auvert, B., Taljaard, D., Lagarde, E., Sobngwi-Tambekou, J., Sitta, R. and Puren, A., 2005. Randomized, Controlled Intervention Trial of Male Circumcision for Reduction of HIV Infection Risk: The ANRS 1265 Trial. PLoS Medicine, 2(11), p.e298. 
[3] Kahn, J., Marseille, E. and Auvert, B., 2006. Cost-Effectiveness of Male Circumcision for HIV Prevention in a South African Setting. PLoS Medicine, 3(12), p.e517.
[4] U.S. Bureau of Labor Statistics, Consumer Price Index for All Urban Consumers: All Items in U.S. City Average [CPIAUCSL], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/CPIAUCSL, January 6, 2022.
[5] Aldeeb Abu-Sahlieh, S., 2003. Circoncision masculine circoncision féminine Débat religieux, médical, social et juridique. Paris: L'Harmattan, pp.253-274.
[6] Wallerstein, E., 1980. Circumcision an American health fallacy. New York: Springer, pp.38-39.
[7] Who.int. 2022. Dossier d'information sur la circoncision et la prévention du VIH. [online] Available at: <https://www.who.int/hiv/mediacentre/infopack_fr_4.pdf> [Accessed 7 January 2022].
[8] Keymanthri, M., 2022. Ethical Issues related to the Orange Farm Study. [online] Research Gate. Available at: <https://www.researchgate.net/publication/309411525_Ethical_Issues_related_to_the_Orange_Farm_Study> [Accessed 7 January 2022].
[9] (2005) A Landmark Paper in HIV Research? PLoS Med 2(11): e293. https://doi.org/10.1371/journal.pmed.0020293
[10] Siegfried, N., 2005. Does Male Circumcision Prevent HIV Infection?. [online] PloS Medicine. Available at: <https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.0020393> [Accessed 7 January 2022].
[11] Constance Rose, A., 2009. Cultural Determinants of Adoption of HIV/AIDS Prevention Measures and Strategies Among Girls and Women in Western Kenya. [online] Google Books. Available at: <https://books.google.fr/books?id=lzQ-iQCnlR8C&pg=PA67&lpg=PA67&dq=orange+farm+rct&source=bl&ots=IeA7GKHO1Q&sig=ACfU3U316mI01VtAWmmSErDxXYW3jaHl9g&hl=fr&sa=X&ved=2ahUKEwjQo-OL5pz1AhXByIUKHducAcMQ6AF6BAgREAM#v=onepage&q=orange%20farm%20rct&f=false> [Accessed 7 January 2022].
[12] Hutson JM (2004) Circumcision: a surgeon's perspective. J Med Ethics 30:238-240.
[13] Helsten SK (2004) Rationalising circumcision: from tradition to fashion, from public health to individual freedom -- critical notes on cultural persistence of the practice of genital mutilation J Med Ethics 30: 248-253.
[14] Bonner K (2001) Male circumcision as an HIV control strategy: Not a ‘‘natural condom’’. Reprod Health Matters 9: 143–155.
[15] Droit au corps. 2013. Circoncision et VIH/sida : l’ablation du prépuce ne protège pas. [online] Available at: <https://www.droitaucorps.com/circoncision-sida-protection> [Accessed 9 January 2022].


Aucun commentaire: