Medical and social consequences of abortion

During the conference which took place in Mauritius in 1994 on "Unsafe abortion and post-abortion family planning in Africa", it was reported that every day 10,000 women in Africa have an unsafe abortion, and many women die as a result or suffer chronic pain, illness and sterility (IPPF, 1994). The World Health Organisation estimates that in Africa, 99% of abortions practised are illegal and therefore unsafe (World Health Organisation, 1998). Legality and safety are generally linked, and morbidity and mortality levels relating to abortion decrease dramatically in countries where abortion has been legalized (The Alan Guttmacher Institute, 1999). Anderson B.A. and al. (1994) believe the most important determinant of the impact that abortion has on women's health is its legal status. Studies on abortion-related complications and mortality only give a partial picture of the importance of this phenomenon and its consequences. In fact only serious complications which have led to treatment in a health structure are taken into account. For economic, social or geographical reasons, not all women have access to medical structures capable of dealing with these complications. This means that women living in rural areas or underprivileged environments, or young women who are not financially independent etc., will not be able to access such treatment.

Abortion-related maternal morbidity and mortality

Most data on abortion-related complications comes from studies conducted in hospitals, often in large cities. Aside from information on the characteristics of the women involved and the methods used (which we have seen already), these studies provide information on the different complications involved and the potential deaths.

Method-related complications

Abortion-related complications and their severity are closely linked to the type of abortive methods used. The most "unsafe" methods are above all those based on traditional pharmacopoeia, the insertion of objects into the uterus, overdoses of medication, and chemical products. However surgical methods are not without risk either, above all when they are practised by unqualified health workers or conducted in an insufficiently sanitary environment. The use of certain methods does not always produce the desired effect and sometimes results in the failure of the abortion. This has aftereffects on the unborn child, such as malformation and stillbirth, or health problems for the woman. The most frequent complications are genital lesions, chronic pelvic pain, infection, haemorrhaging, hysterectomy, sterility and incontinence problems and even the death of the woman involved (Anderson B.A. and al., 1994). Added to these health problems are all the psychological and social consequences involved for women, consequences that are touched on very little in these studies.

In the Demographic and Health Survey in Gabon, 27% of women reported having had complications after an abortion (Barrère M, 2001). These complications occur most frequently in women aged between 20 and 24, living in rural areas, who are poorly educated and use traditional methods (such as plants  or other methods) or overdoses of pharmaceutical products. In rural areas, women use mainly traditional methods to abort. The same is true for poorly educated women - only 29% of such women have had an abortion by curettage or vacuum aspiration, whereas amongst women with a secondary school education the proportion is 46%.

In a survey in Bamako conducted on women consulting practitioners at health centres, 59% reported having had complications after an abortion, above all haemorrhaging (Konate M.K. and al., 1996). In a survey conducted in Bamako hospital, Diallo Diabate and al. (2000) identified the principal post-abortion complications as haemorrhaging (47%), infectious complications (33%) and medicinal poisoning (4%).

Women admitted to the teaching hospital in Abidjan had severe abortion-related complications, such as signs of peritonitis caused by the use of an intra-uterine method, and neurological disorders linked to the ingestion of herbal teas (Goyaux N.,and al., 1999). In Nigeria, Ikpeze O.C. (2000) reports abortion-related complications such as genital lesions, abscesses and peritonitis, and perforation of the uterus. The same thing is noted by Bazira E.R. (1992), who showed that in Uganda, women who stayed in hospital for at least two days with abortion-related complications, had haemorrhages, infections and genital lesions. In Nigeria, Konje, J. C. and Obisesan, K. A (1991) also mention peritonitis as a major complication of illegal abortions.

These hospital studies generally only show the immediate consequences of illegal abortions, which lead the women involved to seek hospital treatment, but not all consequences are immediately apparent. For Mpangile and al. (1999), abortion-related complications should be measured in the long term and can include gynaecological problems, miscarriages or premature births, and even sterility. In Ethiopia, complications related to induced abortions (essentially using intra-uterine methods) include septicaemia, haemorrhages, infections and genital lesions. In the longer term, however, these illegal abortions can cause chronic pelvic inflammation and sterility (Madebo T and Tsadic GT, 1993). Some such abortions may also cause the death of the woman.

The mortality rate associated with these abortions is analysed as a way of determining programmes of action that could be implemented to reduce the problem.

Abortion-related mortality

Maternal mortality is difficult to measure, and its levels are often based on estimations. Abortion-related mortality levels are even more difficult to determine. Estimation techniques have been established based on data concerning abortion-related complications (Singh S. and Deirdre W., 1991; Singh S. and al., 1997). In Africa, maternal mortality rates are often estimated during Demographic Surveys using indirect techniques, however these estimations do not provide any information on the causes of mortality. Certain specific surveys on maternal mortality and abortion-related complications enable us to pinpoint the extent of the role played by illegal abortions in maternal mortality, in comparison with other causes.

The WHO estimates that 20 million unsafe abortions are carried out every year (which is one abortion for every ten pregnancies), that 95% of these occur in developing countries and around 80,000 women die each year from the ensuing complications (World Health Organisation, 1998). In Africa, the number of unsafe abortions is estimated at five million, with an ensuing mortality rate varying from 24 per 100,000 live births in North Africa to 121 in West Africa and 153 in East Africa. Abortion is a major cause of maternal deaths (7 to 19%). Its magnitude is under-estimated because the act is illegal, and also because some deaths may be classified under other causes such as haemorrhage or infection (see table 1).

The risks associated with abortion vary according to the region and the legal or illegal status of the act. Thus when analysing the mortality risks associated with pregnancy, it emerges that in developing countries this risk is at the level of four to six deaths per 100,000 legal abortions, and between 100 and 1000 per 100,000 illegal abortions (the rate for these two kinds of abortions in the USA is between 1 and 50), whilst the risk of death in developing countries following a caesarian is only between 160 and 220 per 100,000 births (World Health Organisation, 1998). The link between the legalisation of abortion and the maternal mortality rate is obvious. Romania is a good illustration of this, being a country where abortion legislation has constantly alternated between liberalisation and restriction, and where the maternal mortality rate and the part played by abortion in this rate have fluctuated according to these changes (Johnson, B. and al., 1996)

In South Africa, Vandecruys H. I. and al. (2002) highlight a decrease over the previous four years in maternal mortality rates, which have fallen from 268 to 94 per 100,000 live births - this decrease is attributable to a decline in abortion-related complications. Before 1997, when abortion was legalised, 200,000 illegal abortions were carried out every year, leading to 45,000 hospitalizations and at least 400 deaths annually (Bennett T., 1999). A multicentric study conducted in 56 public hospitals in nine provinces showed a high incidence of morbidity and mortality associated with incomplete abortions (Rees, H. and al., 1997).

It is difficult to estimate evolutions in maternal mortality over time, particularly mortality attributable to abortion, but various hospital studies have demonstrated its significance. In 1993, Rogo K.O. described how hospital studies confirmed that unsafe abortions and their subsequent complications were responsible for up to 50% of maternal deaths in Sub-Saharan Africa. These studies also highlighted a low use of contraception in the region and a large number of unwanted pregnancies.

Using epidemiological studies conducted in 41 African countries, Benson, J., and Johnson BR. (1994) show that maternal mortality rates vary from 200 to 600 per 100,000 live births and that 18 to 35% of these deaths are attributable to abortions.

In Uganda at the beginning of the 1990s, Kampikaho A. and Irwig L.M (1991) estimated that in Kamapala's hospitals, the rate of maternal mortality following abortion was 3.58 per 1000 abortions, whilst the rate attributable to other causes was 2.0 per 1000 deliveries. Mirembe F.M. (1994) points out that illegal abortion accounts for one third of maternal deaths, the maternal mortality rate in the country being 550 per 100,000 births. In 2000, Kasolo J. specifies that illegal abortion is still responsible for 20% of maternal deaths. In Uganda, still at Mulago hospital, 35% of maternal deaths are due to abortion (Mbonye A., 2000).

In Zimbabwe in the 1990s, abortion was responsible for 12% of maternal deaths (Mahomed al., 1994). In Lesotho, the maternal mortality rate is estimated at 220 per 100,000 births, and abortion accounts for 25 to 33% of maternal deaths ((Mosaase, 1994). In Mozambique, Songane F.F. and Bergstrom S. (2002) demonstrate that abortion is the third largest cause of maternal mortality (responsible for 18% of deaths) after haemorrhages (30%) and puerperal infections (23%). In Botswana, according to Ramalefo C. and Modisaotsile I.M. (1994), at least 14% of maternal deaths in hospitals are attributable to abortion, and in Tanzania the rate is 10 to 12% (Temba P. and Nkya A., 1994). In Malawi, Chiphangwi J.D. and al. (1992) note that abortion is the third largest cause of maternal mortality (18%) amongst haemorrhages (25%) and labour complications (20%). In Nairobi, in Kenya, 35% of deaths linked to pregnancy are due to unsafe abortions (Anonymous, 1998).

In Zambia, Koster-Oyekan W. (1998) highlights the significance of abortion with regard to maternal deaths (120 per 100,000 births), and says that over half of these deaths occur in young women who are still schoolgirls or students.

In Ethiopia, several studies show that 25 to 35% of maternal deaths are attributable to abortions (Gebreselassie H. and Fetters T., 2002). In 1994, Ali, Y. demonstrates in a study that abortion had contributed to 41% of maternal deaths. In Brazzaville in Congo, where maternal mortality is estimated at 408 deaths per 100,000 live births, Iloki L. H. and al. (1997) cite the following as major direct causes of obstetric deaths: haemorrhages (40.5%), post-partum septicaemia (12.7%), post-abortion septicaemia (18.9%) and eclampsia (11.2%). In Egypt, the maternal mortality due to abortion is fairly high, accounting for between 2.1 and 9.8% of all maternal deaths (Lane Jok and al., 1998).

In West Africa, abortion is an equally serious public health problem. In Benin at the end of the 1980s, Unuigbe J. A. and al. (1990) highlighted a high level of maternal mortality amongst teenagers, mainly due to induced abortions - 72% of these deaths were attributable to illegal abortions. In Guinea, Toure B. and al. (1992) put abortion amongst the principle causes of maternal mortality, along with complications arising from hypertension and post-partum haemorrhages. In Ghana, abortion is responsible for 22% of maternal deaths (Billings, 1997). In Ouagadougou, a study conducted on hospitalized teenagers aged between 13 and 19 shows that 30% of maternal deaths are caused by illegal abortions, and 30% by chronic anaemia (Lankoande, 1999). In Nigeria, Oye-Adeniran B. A. and al. (2002) indicate that 30 to 40% of maternal deaths are attributable to abortions, and in Illorin 90.3 deaths occurred out of 1000 abortions, although this proportion is actually estimated at one death per 1000 to 2000 abortions in developing countries (Anate M. and al., 1995). In Abidjan, 3.6% of women died as a result of abortion-related complications. The majority of these deaths follow the ingestion of herbal teas designed to induce abortion (Goyaux N., and al., 1999). In Benin, Senegal and Cameroon, the case fatality rate of induced abortions is 2.3%, varying between 1.8% in Benin, 2.8% in Senegal and 4.1% in Cameroon (Goyaux N. and al., 2001).

These few elements of data are disparate and do not wholly reflect the issue of abortion in Africa. They do, however, demonstrate the consequences involved for women's health and lives. We are now going to consider the economic and social consequences.

The cost of abortion

Abortion has a social and economic cost for both women and society as a whole (Gebreselassie H. and Fetters T., 2002). The reasons for seeking an abortion are a significant illustration of the risks that women consider they are exposing themselves to in continuing with a pregnancy that is not accepted by their family or by society. Induced abortions also generate significant costs for health systems, which puts a strain on their already insufficient financial and human resources.

The cost of treatment to health systems

The cost of abortion to health systems is essentially measured by the level of treatment of abortion-related complications, since the practice is illegal in most African countries. These costs may be set out in terms of hospital bed occupancy rates, the amount of time spent by medical staff, and the resources allocated to the treatment of such abortions.

Several authors have highlighted this expense for hospitals, particularly for emergency and obstetrical gynaecology departments. Thus in some developing countries, two thirds of gynaecology and obstetrics beds, along with 50% of these departmental budgets, are taken up by abortion-related complications (The Alan Guttmacher Institute, 1999). In their summarizing article on abortion in Sub-Saharan Africa, Benson J. and Johnson BR. (1994) show that abortion-related complications account for 27 to 60% of admissions to gynaecology departments in the country.

In Egypt, Mashalli A.E. (2000) notes that, based on a survey conducted in 1998-99 in a hospital, one admission in eight to the emergency obstetrics department was due to an abortion. Lane J. and al., 1998, highlights the fact that the treatment of such complications consumes around 50% of the annual maternity budget. In South Africa, before abortion was legalized in 1997, several authors highlighted the enormous cost to hospitals represented by the treatment of abortion-related complications (Benson J and al., 1996; Kay B. J., and al., 1997). In Mozambique, where abortion is legal under certain conditions, hospital fees are a barrier to accessing these services (Machungo F. and al., 1997). In Senegal, abortion-related complications cost the health services between CFA Francs 6,014 and CFA Francs 56,275 (around between US$10 and 100), which is a high cost given the standard of living in the country (Camara M. and Cisse L., 1998). In Tanzania, the treatment of abortion-related complications costs an estimated TZS1500 per day (US$1.5), which is seven times the annual Health Ministry budget per person (TZS210 around US$0.2) (Mpangile and al. 1999). In Morocco, there is a whole network of doctors practicing illegal abortions in private clinics or hospitals for between US$95 and US$175, which is an inaccessible sum for many women (Naamane-Guessous, 1993).

The cost for hospitals is not only counted in monetary terms, but also in terms of the workload involved for staff in treating a pathology. Thus in Kenya, Solo J. and al. (1995) demonstrate that in 18 hospitals, staff treating incomplete abortions consists on average of 1.4 specialists, 0.4 doctors and 3.9 nurses.

The treatment of such abortions sometimes necessitates long periods of hospitalization. In Nigeria, the length of a stay in hospital is 10.5 days for an induced abortion and 7.5 days for a spontaneous abortion (Anate M. and al., 1995). In another study conducted in hospitals, Adewole I.F. (1992) shows that most women are hospitalized for between three and seven days (38.3%) or between seven and fourteen days (41%) for complications. The cost of treating the consequences of an abortion is around NGN350, or US$35, which is higher than the national monthly salary. In Kenya, a quarter of hospital gynaecology and obstetrics department staff's time, along with 30% of departmental beds, are devoted to the treatment of these incomplete abortions (Ankomah A. and al., 1997). According to Bradley J. and al. (1993), over half of admissions to gynaecology and obstetrics departments are related to incomplete abortions or associated complications. In Harare hospital's gynaecology department, half of all emergencies are due to abortion-related complications ((Mahomed and al., 1992).

The costs of treating these complications vary greatly according to how the process is managed. Based on the experiences of various African and Asian countries, King T. and Benson J. (1998) demonstrate the advantages of the manual vacuum aspiration method, as compared with dilatation and curettage, in the treatment of abortion-related complications. This method reduces the length hospital stays by at least a third, and the cost is reduced from US$78.81 for dilatation and curettage to US$8.50 for vacuum aspiration. A study in Kenya has shown that if abortion-related complications were treated with vacuum aspiration, women would stay in hospital for between 18 and 24 hours, as opposed to between 41 and 101 hours when curettage is used, which considerably reduces the costs of treatment (Johnson B. R. and al., 1993). Equally in Senegal, a reduction in treatment costs has been noted when complications are treated by vacuum aspiration, this reduction being linked to a decrease in the duration of hospitalization, which fell from 79 to 51 hours (CEFOREP, 1998).

The introduction of manual aspiration methods constitutes a significant improvement in the treatment of abortion-related complications. It is a less expensive procedure, poses fewer risks to women's health, is quicker and involves shorter stays in hospital. Such programmes should become widespread in order to minimise costs to health services.

The socio-economic cost to families

The cost of abortion to women and families is largely dependent on the method used, the type of assistance the woman receives and the place where the abortion takes place. Sometimes there is also the cost of complications to be added. Where abortion is illegal, the least dangerous methods, i.e. medicalized methods, are very expensive. The cost of an abortion is also linked to the term of the pregnancy - thus in Abidjan, the cost of an abortion by curettage increases with the term of the pregnancy (Guillaume A., 1999).

In Gabon, for example, over half of women who have undergone an abortion in a private health establishment (57%), along with 41% of women who have had an abortion by curettage and 42% who were helped by a doctor, paid more than CFA Francs 100,000 (around US$135) for the act. This cost is significant when compared to the level of income in the country (Barrère M,, 2001). By contrast, when a woman uses traditional methods, the abortion is free or affordable, but has grave consequences on her health.

In Kenya, the cost of abortion varies between US$9 and US$145 (Ankomah A. and al., 1997). In 1991 in Egypt the cost was estimated at between US$ 60 and 150 (which is more than one month's average salary) for an illegal abortion carried out in a medical environment, and at less than US$ 1 for traditional methods which entail serious risks to health (Lane, Jok, and al., 1998). In Abidjan, abortion is free, or very cheap, if it is carried out using traditional methods, and costs between CFA Francs 30,000 (US$ 40) and CFA Francs 180,000 (around US$ 240) if conducted in a private clinic or in hospital.

In Nigerian cities, the cost of an abortion for the poorest women, who use herbal methods, is 160 Nairas (around $2), and can reach up to 1,900 Nairas (around US$24) if it is carried out by a doctor in a private clinic. For women from the most privileged classes, the cost is between 260 and 4,000 Nairas (between US$3 and 50) for the same kind of service (Makinwa-Adebusoye P. and al., 1997).

Abortion is expensive for women, particularly for teenagers. Women with limited resources often have problems finding the necessary funds to finance a pregnancy interruption and abort at an advanced gestation term, which exposes them to unsafe procedures (The Alan Guttmacher Institute, 1999).

The difficulties involved in paying for an abortion also raise the issue of the role played by men in the decision to abort, along with the question of assumption of the related costs. This role generally varies. Sometimes it is limited to financial participation or, in other cases, to finding a practitioner who will carry out the abortion. The investment a man makes in the procedure of pregnancy interruption depends on the nature of his relationship with the woman involved.

In addition to the financial costs of abortion borne by women and families, there are also a certain number of other costs. Abortion can cause psychological disorders, sterility and family problems (Camara C. M. and Cisse L., , 1998; Leke, R. J. and Chikamata, D. M., 1994). The medical aftereffects of abortion are sometimes apparent in the short term, and sometimes only in the longer term. Thus women may suffer from chronic pain which is disabling and can disrupt their lives, particularly in their professional activities, leading to a decrease in their productivity. Amongst the disabling aftereffects involved are incontinence and sterility, two problems that can contribute to social exclusion. Genital lesions, post-abortal infections and of course hysterectomy are all sources of sterility. In their study on sterility in Africa, Gerais, A. S. and Rushwan, H. (1992) point out that amongst women who have had problems with blockages in their fallopian tubes, 8% had previously suffered abortion-related complications. Sterility can cause conjugal and family problems for women, particularly for very young women who have never had children. However, these young women often prefer to use abortion rather than contraception, because they fear the supposed risk of sterility that is associated with contraception. Occasionally some young women also seek to end a pregnancy that they initially desired in order to assure themselves of their own fertility, without considering the risks to which they are exposing themselves by aborting (Guillaume A, 1999).

One of the aspects of abortion-related complications that is rarely taken into account are the psychological and social consequences involved, such as rejection by the family, conjugal problems, feelings of guilt and shame and fear of sterility.

In some studies, abortion is presented as a form of violence because of the mortality it generates. In recent analyses of maternal mortality, causes of death have in fact been reinterpreted following changes made to the definitions in the WHO's international statistical classification of diseases and related health problems (10th revision) (ICD-10). The concept of maternal death has evolved, and we now talk about pregnancy-related death, which is defined as the death of a woman during pregnancy or within 42 days after pregnancy has terminated, whatever the cause of death may be. The concept of late maternal death has also emerged - this is defined as the death of a woman as a direct or indirect result of obstetrical causes, which occurs more than 42 days but less than one year after her pregnancy has terminated. These new concepts mean that a wider range of maternal deaths are now taken into account. Thus some deaths which occur as a result of violence, homicide and suicide are now considered to be maternal deaths. This change has lead some authors to consider deaths associated with abortion, which were initially classified as deaths directly resulting from obstetrical causes, as deaths caused by violence. In Mozambique, for example, Granja A.C. and al. (2002) interpret abortion as an act of violence when it is conducted illegally, by unqualified practitioners in inadequately safe conditions. Women also inflict a form of violence on themselves when they induce their own abortions using unsafe methods. This debate is already under way in Africa and deserves to be taken into consideration, since it categorizes abortion not just as a health problem, but also as a social problem.

Illegal induced abortions have serious consequences on the health and lives of women. The data presented here only shows a part of the issue, the weight of the phenomenon being definitely greater for certain groups of women. Abortion remains a significant cause of mortality in countries where access to it is restricted. It is therefore becoming urgent that the political and medical authorities review their legislation with the consequences of illegal abortion in mind.

The treatment of abortion-related complications generates high costs for health systems, putting a strain on their budgets and burdening medical staff with extra workloads. The development of post-abortion care programmes, and in particular the introduction of manual vacuum aspiration methods, is important in order to alleviate these burdens. These methods, which are less traumatic, cheaper and necessitate shorter periods of hospitalization, also have the advantage of reducing the harmful consequences for women's health.

We should also highlight the financial cost of abortion for women. Under current legislation, opting for a safe practice in the best possible medical conditions is far from the reach of many women. There is also a social cost to abortion linked to the complications that may ensue when the practice does not come with all possible medical guarantees. Thus women who suffer serious consequences from abortion, who are generally the most resourceless and/or the youngest women, find themselves physically damaged and even sterile, and in addition to this are exposed to social rejection. Aside from the not inconsiderable financial and social costs of abortion, the long-term consequences need therefore to be pointed out.