Estimating the obstetric costs of female genital mutilation in six African countries
David Bishai,a Yung-Ting Bonnenfant,a Manal Darwish,b Taghreed Adam,c Heli Bathija,c Elise Johansenc & Dale Huntingtonc for the FGM Cost Study Group of the World Health Organization
a Johns Hopkins Bloomberg School of Public Health, 615 N Wolfe Street, Baltimore, MD, 21205, United States of America.
b Assiut University, Assiut, Egypt.
c World Health Organization, Geneva, Switzerland.
Correspondence to David Bishai (e-mail: email@example.com).
(Submitted: 09 March 2009 – Revised version received: 03 July 2009 – Accepted: 21 September 2009 – Published online: 20 January 2010.)
Bulletin of the World Health Organization 2010;88:281-288. doi: 10.2471/BLT.09.064808
The term “female genital mutilation” (FGM) denotes any procedure involving partial or total removal of the external female genitalia, as well as injury to the female genital organs for non-medical reasons.1 The World Health Organization (WHO) estimates that between 100 and 140 million girls and women worldwide are presently living with FGM,1 and every year about three million girls are at risk.1,2 FGM is a fundamental violation of human rights. It is not only a severe form of discrimination against women, but also a violation of the rights of girls, on whom it is most commonly performed.1 FGM violates the right to health and to freedom from torture or cruel, inhuman or degrading treatment and, in some cases, even the right to life.1 As a result, support for the abandonment of FGM can be found in numerous international and regional human rights treaties and consensus documents.1
FGM carries serious health consequences both for the girl or woman who undergoes the procedure and for her offspring. The procedure can lead to direct medical complications. In a study of women in Nigeria who had FGM, the most common of these were severe pain and bleeding.3 Infection also poses an immediate risk.1,4,5 Long-term health effects include psychological1,6 and psychosexual1,7 trauma, infertility,1,8 susceptibility to bacterial vaginosis and genital herpes,1,9 and obstetric complications, including perinatal death.1,10–12
This study focuses specifically on the medical costs associated with the obstetric complications of FGM. Prior studies have shown higher rates of Caesarean section, postpartum haemorrhage, prolonged hospitalization and perinatal death among women who have suffered FGM.10 Yet these obstetric complications account for only a small portion of the overall health impact of FGM in a population, and their financial costs are merely one among the many costs associated with FGM. However, when the financial burden that FGM imposes on the health system is measured, it becomes obvious that caring for women who have undergone this procedure imposes an even greater economic burden and that the cost of efforts to prevent FGM can be wholly or partially offset by the savings generated when complications are prevented. The objective of this paper is to build on prior estimates of the obstetric risks linked to FGM so as to estimate how much FGM-related obstetric complications cost the health-care system and society.
A large WHO study quantifying the relative risk of obstetric complications among African women by type of FGM made it possible to carry out this cost study.10 In the WHO study, which was conducted from November 2001 to March 2003 in 28 obstetric centres in Burkina Faso, Ghana, Kenya, Nigeria, Senegal and the Sudan, women and their neonates were prospectively followed for adverse outcomes from admission before labour or in early labour until discharge. The study was limited to 28 393 women who had a singleton delivery. Women’s FGM status was determined by direct examination of the external genitalia and in accordance with WHO’s four-category classification of FGM (Box 1). The relative risks of adverse maternal and infant health outcomes for each type of FGM (with no FGM as the reference category) were also estimated. These outcomes included Caesarean section, haemorrhage (postpartum blood loss ≥ 500 ml), extended maternal hospital stay (≥ 3 days), low birth weight (< 2500 g), resuscitated infant, and inpatient perinatal death. With the exception of episiotomy, our cost analysis was based on the frequency and relative risk of each complication for each type of FGM as reported in the published WHO study.
Box 1. Types of female genital mutilation as defined by the World Health Organization
Excision of the prepuce, with or without total or partial excision of the clitoris.
Excision of the clitoris with partial or total excision of the labia minora.
Total or partial excision of the external genitalia and stitching or narrowing of the vaginal opening (infibulation).
Unclassified, which includes pricking, piercing or incising the clitoris and/or labia; stretching the clitoris and/or labia; cauterizating the clitoris and surrounding tissue; scraping the tissue surrounding the opening of the vagina (angurya cuts) or cutting the vagina (gishiri cuts); introducing corrosive substances or herbs into the vagina to cause bleeding or to tighten or narrow it; and any other procedure that can be included in the definition of FGM noted above.
FGM, female genital mutilation.
The model created six simulated cohorts of 100 000 women for each of the six African countries previously mentioned. In the model, the survival and birth history of each woman from age 15 to 45 years was constructed by applying the relevant fertility13 and mortality14 rates for each country. With each birth, the members of the cohort were programmed to seek formal obstetric care with a likelihood consistent with the most recent observations from Demographic and Health Survey (DHS) data for the study countries (Appendix A, available at: http://sites.google.com/site/fgmappendix/appendix-1).
Each of the six cohorts was followed under four different FGM scenarios: no FGM, all FGM–1, all FGM–2 and all FGM–3. The respective FGM categories were assigned to every woman at age 15 and retained for life. In each run of the model, each woman’s risks of suffering complications depended on her FGM status and on her choice to have or not have a medically attended delivery (Table 1). Since this study was a secondary analysis of anonymous data and did not involve human subjects, it was exempted from review board approval by the Johns Hopkins University internal review board.
Table 1. Prevalence of obstetric outcomes, by type of female genital mutilation, in a modelled cohort of women of reproductive age (15–45 years) in six African countriesa
We calculated the unit costs associated with Caesarean section, maternal haemorrhage, extended maternal hospital stay, infant resuscitation, inpatient perinatal death and episiotomy. (Some suggested that haemorrhage after FGM can be perineal rather than uterine and thus more easy to manage. However, a reanalysis of the WHO data confirmed that most haemorrhage among the women sampled, regardless of FGM type, was caused by an atonic uterus.) For each type of FGM, the relative risk of the first five outcomes was taken from overall risk calculations for all six African countries combined, as given in the aforementioned WHO study. Episiotomy was the only outcome for which the risk was recalculated for each FGM type, since the WHO study presented the relative risk of episiotomy separately for primiparous and multiparous women, rather than for both combined. We estimated the risk of episiotomy by FGM type for all women by reanalysing the WHO data set.
The baseline prevalence of each adverse outcome in women without FGM was modelled with a set of beta distributions, and the log relative risks associated with FGM type were modelled with normal distributions drawn from the WHO study. Outputs included the event counts of each obstetric complication, years of life lost, and differences in discounted costs and discounted life years between a given FGM type and no FGM. The model was run with @RISK version 4.5 (Palisade Corporation, Newfield, NY, USA).
The model adopts the perspective of the health system and assumes that only medically attended deliveries are likely to impose monetary health system costs. The probability of having an attended delivery was based on country-specific DHS estimates of delivery attendance by a physician in a given country. Using DHS data, we tried to determine if having an attended delivery depended on FGM status but found no statistically significant differences by FGM status for any country except Nigeria.
All costs in this analysis are presented in international (purchasing power parity) dollars (I$), which adjust for the cost of living in each country. The costs associated with Caesarean section and haemorrhage were estimated from the Lancet Neonatal Survival Series and the BMJ CHOICE Series.15,16 For all countries, we estimated the cost of a Caesarean section and a blood transfusion for haemorrhage to be I$ 36.40 and I$ 29.79, respectively.15,16 Country-specific costs per bed day were estimated to be I$ 2.98 for Burkina Faso, I$ 3.71 for Ghana, I$ 4.82 for Kenya, I$ 5.38 for Nigeria, I$ 6.73 for Senegal and I$ 2.69 for the Sudan, based on estimates provided by Adam et al.17
Years of life lost
Years of life lost were calculated for women who died from haemorrhage. We took the case fatality rate of 1.7% (95% confidence interval, CI: 1.2–2.5) that was found among women who experienced a blood loss ≥ 500 ml in the WHO study and applied it in our simulation to attended deliveries. We assumed conservatively that the rate for unattended peripartum haemorrhage patients would be three times higher, or 5.1% (95% CI: 3.6–7.5). These numbers are consistent with those in previous studies.18,19 Life expectancy between the ages of 15 and 45 years was estimated based on WHO life table14 data for women in each country.
The model was constructed for women of reproductive age with a 30-year time horizon. Women entered the model at 15 years of age and exited upon turning 46 years old. Costs and life years were discounted by 3% per annum, so that the estimated lifetime savings from investing to prevent FGM today are equivalent to the present value of future savings.
Table 2 shows the estimated years of life lost and obstetric costs associated with every new case of FGM, as well as the expected cost increments and life year decrements if each type of FGM were imposed on a 15-year-old girl (relative to no FGM). The table shows that, on average, a girl of 15 years who undergoes FGM–3 will lose nearly one-fourth of a year of life and impose on the medical system a cost of I$ 5.82 over her lifetime. Clearly, FGM–3 is both detrimental to health and a waste of money. The other types of FGM also reduce survival and lead to monetary losses over each woman’s lifespan, but to a lesser extent.
Table 2. Incidence-based estimates of costs and years of life lost per incident case of female genital mutilation, by type, in a modelled cohort of women of reproductive age (15–45 years) in six African countries
The losses of life and money are different in each country because of differences in fertility rates and in the proportion of medically attended deliveries. Women who have more deliveries are more exposed to the risk of dying and of suffering complications. Countries that have achieved higher rates of medically attended deliveries also incur higher medical costs because of the complications associated with such deliveries. However, the model should not be interpreted to mean that the rate of attended deliveries should be reduced to lower the obstetric costs of FGM, but rather the opposite. Because deaths from complications are more common when births take place at home, reducing attended deliveries would result in higher death rates.
Table 3 shows the estimated annual medical costs for the entire population of women with FGM. The 53 million African women who live in the six countries studied represent a total of I$ 3.7 million in medical costs for the management of the obstetric complications linked to FGM, equivalent to between 0.1% and 1% of total government health spending on women of reproductive age (Fig. 1). Table 4 shows the estimated future loss of life attributable to FGM-related obstetric haemorrhage. The current population of 2.8 million 15-years-old girls in the six African countries studied will lose approximately 130 000 years of life as a result of the FGM procedures that will be performed over the next 12 months.
Table 3. Prevalence-based estimates of costs per prevalent case of female genital mutilation, by type, in a modelled cohort of women of reproductive age in six African countries
Fig. 1. Annual obstetrical costs related to female genital mutilation as a percentage of all government health spending on women aged 15–49 years
Table 4. Future years of life lost as a result of incident female genital mutilation cases for the 15-year-old population in six African countries
Multivariate sensitivity analysis, shown in Appendix A, measures the degree of uncertainty surrounding our central finding that FGM imposes an economic burden on the health system. The proportion of runs in which FGM led to either higher costs, years of life lost or both was 77%, 85% and 93% for FGM–1, –2 and –3, respectively. Thus, we can be fairly confident that every new girl subjected to FGM represents a future stream of preventable obstetric costs and/or a future death from obstetric complications.
FGM violates human rights. It is impossible to say how much society should spend to prevent it, but as shown by the results of our study, any money spent on preventing FGM would be partially offset by savings to the health system. Indeed, if the health system were to spend as much as I$ 5.82 per FGM–3 prevented or I$ 2.50 per FGM–2 prevented, the value of avoided obstetric complications would entirely offset the costs of prevention.
An estimated $23 million or more were spent by bilateral, multilateral and private foundation donors in 2007 on activites surrounding the prevention of FGM.20 However, the cost-effectiveness of such spending per case prevented is unknown. Typically, interventions for the prevention of FGM take the form of community-based programmes, media outreach and advocacy and often target religious leaders and excisors.20
This study did not address the effects of FGM on a woman’s mental health or any of the medical complications stemming from the initial procedure, namely pain, bleeding and infection.1,3–5 These complications are described elsewhere. In Egypt, Elnashar & Abdelhady found that circumcised women had significantly higher rates of psychological problems than women who were not circumcised.21 Others have also found a higher prevalence of posttraumatic stress disorder6 and sexual health problems among circumcised women.1,21 We were unable to include the immediate complications of FGM in our model for lack of quantitative estimates of their frequency.
Our data were limited to a few health centres in each of six countries that participated in an international study, and this could have biased our results. However, we used cluster-specific random effects when possible to produce the best estimates. We may have also over-counted some obstetric costs related to FGM because the cost of some bed days is usually included in the cost of a Caesarean section. It is also possible that some of the estimated costs are paid out of pocket by the families of the women concerned rather than by the health system.
Our analysis may have underestimated the total costs of FGM to the health system because it did not include the treatment of post-delivery complications in women who delivered at home. Costs to the health system are likely to increase in future years as access to health care increases in the poor and rural areas of these countries and a greater percentage of women deliver in health facilities. The estimated costs of treating obstetric complications within the health system may be inaccurate as well, since some costs are extrapolations and thus not country-specific and even those that are were based on assumptions that may not be valid for all parts of the country.
Despite the large size of the WHO study sample, the CIs for many of the parameters of interest remain wide. Additional epidemiological research could narrow the CIs around the known parameters. A further limitation of our study, which applies to virtually all cost studies built on models, is the use of multiple data sources to support the estimates.
Civil society should give the highest priority to addressing the human rights violation inherent in FGM. However, measures to address the problem require financing. Efforts to combat FGM have been traditionally underfunded, but as shown by the results of this study, African health ministries that invest in curbing the practice of FGM are likely to recover a large portion of the investment by saving money from prevented obstetric complications. Societies would also benefit from other reduced costs not measured here, including the costs of treating FGM-related psychological and sexual health problems.
The FGM Cost Study Group is composed of Taghreed Adam, Heli Bathija, Dale Huntington and Elise Johansen from the World Health Organization, Geneva, Switzerland; David Bishai and Yung-Ting Bonnenfant from the Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA; and Manal Darwish from the Assiut University, Assiut, Egypt. We thank Paul van Look and other reviewers at WHO for their helpful comments on the draft manuscript.
The funding source for this study was WHO, and WHO scientists were involved in the writing and decision to submit the paper for publication.
- Eliminating female genital mutilation: an interagency statement. Geneva: World Health Organization; 2008.
- Yoder PS, Abderrahim N, Zhuzhuni A. Female genital cutting in the demographic and health surveys: a critical and comparative analysis. Calverton, MD: Macro International, Inc.; 2004.
- Dare FO, Oboro VO, Fadiora SO, Orji EO, Sule-Odu AO, Olabode TO. Female genital mutilation: an analysis of 522 cases in South-Western Nigeria. J Obstet Gynaecol 2004; 24: 281-3 doi: 10.1080/01443610410001660850 pmid: 15203627.
- Almroth L, Bedri H, El MS, Satti A, Idris T, Hashim MS, et al., et al. Urogenital complications among girls with genital mutilation: a hospital-based study in Khartoum. Afr J Reprod Health 2005; 9: 118-24 pmid: 16485592.
- Chalmers B, Hashi KO. 432 Somali women's birth experiences in Canada after earlier female genital mutilation. Birth 2000; 27: 227-34 doi: 10.1046/j.1523-536x.2000.00227.x pmid: 11251507.
- Behrendt A, Moritz S. Posttraumatic stress disorder and memory problems after female genital mutilation. Am J Psychiatry 2005; 162: 1000-2 doi: 10.1176/appi.ajp.162.5.1000 pmid: 15863806.
- el-Defrawi MH, Lotfy G, Dandash KF, Refaat AH, Eyada M. Female genital mutilation and its psychosexual impact. J Sex Marital Ther 2001; 27: 465-73 doi: 10.1080/713846810 pmid: 11554208.
- Almroth L, Elmusharaf S, El HN, Obeid A, El Sheikh MA, Elfadil SM, et al., et al. Primary infertility after genital mutilation in girlhood in Sudan: a case-control study. Lancet 2005; 366: 385-91 doi: 10.1016/S0140-6736(05)67023-7 pmid: 16054938.
- Morison L, Scherf C, Ekpo G, Paine K, West B, Coleman R, et al., et al. The long-term reproductive health consequences of female genital cutting in rural Gambia: a community-based survey. Trop Med Int Health 2001; 6: 643-53 doi: 10.1046/j.1365-3156.2001.00749.x pmid: 11555430.
- WHO study group on female genital mutilation and obstetric outcome. Female genital mutilation and obstetric outcome: WHO collaborative prospective study in six African countries. Lancet 2006; 367: 1835-41 doi: 10.1016/S0140-6736(06)68805-3 pmid: 16753486.
- Jones H, Diop N, Askew I, Kabore I. Female genital cutting practices in Burkina Faso and Mali and their negative health outcomes. Stud Fam Plann 1999; 30: 219-30 doi: 10.1111/j.1728-4465.1999.00219.x pmid: 10546313.
- Larsen U, Okonofua FE. Female circumcision and obstetric complications. Int J Gynaecol Obstet 2002; 77: 255-65 doi: 10.1016/S0020-7292(02)00028-0 pmid: 12065141.
- Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat. World population prospects: The 2006 revision and world urbanization prospects. New York; UN Population Division; 2007. Available from: http://esa.un.org/unpp [accessed 3 October 2007].
- Life tables for WHO Member States 2005. Geneva: World Health Organization; 2006. Available from: http://www.who.int/whosis/database/life_tables/life_tables.cfm [accessed 3 October 2007].
- Darmstadt GL, Bhutta ZA, Cousens S, Adam T, Walker N. Evidence-based, cost-effective interventions: how many newborn babies can we save? Lancet 2005; 365: 977-88 doi: 10.1016/S0140-6736(05)71088-6 pmid: 15767001.
- Adam T, Lim SS, Mehta S, Bhutta ZA, Fogstad H, Mathai M, et al., et al. Cost effectiveness analysis of strategies for maternal and neonatal health in developing countries. BMJ 2005; 331: 1107- doi: 10.1136/bmj.331.7525.1107 pmid: 16282407.
- Adam T, Evans DB, Murray CJ. Econometric estimation of country-specific hospital costs. Cost Eff Resour Alloc 2003; 1: 3- doi: 10.1186/1478-7547-1-3 pmid: 12773218.
- Martey JO, Djan JO, Twum S, Browne EN, Opoku SA. Maternal mortality due to hemorrhage in Ghana. Int J Gynaecol Obstet 1993; 42: 237-41 doi: 10.1016/0020-7292(93)90217-K pmid: 7901078.
- Etuk SJ, Asuquo EE. Maternal mortality following post-partum haemorrhage in Calabar a 6-year review. West Afr J Med 1997; 16: 165-9 pmid: 9329285.
- Gibbs S. Female genital mutilation/cutting: what are donors doing? Washington, DC: American Public Health Association; 2008. Available from: http://apha.confex.com/apha/136am/techprogram/paper_185716.htm [accessed 18 January 2010].
- Elnashar A, Abdelhady R. The impact of female genital cutting on health of newly married women. Int J Gynaecol Obstet 2007; 97: 238-44 doi: 10.1016/j.ijgo.2007.03.008 pmid: 17445819.