(introductory text...)

More than 150 million women become pregnant in developing countries each year and an estimated 500,000 of these women die from pregnancy-related causes. Well over one-fourth of all deaths to women of reproductive age in many developing countries are pregnancy related. The five main causes of maternal deaths are hemorrhage, unsafe abortion, hypertensive disorders, sepsis, and obstructed labor.

The death toll is greatest in Sub-Saharan Africa and South Asia, where maternal mortality ratios (maternal deaths per 100,000 live births) may be as much as 200 times higher than those in industrial countries. This is the widest disparity in human development indicators yet reported (table 1.1, box 1.1). The profound difference in maternal health and mortality between developing and industrial countries is expressed even more starkly by comparing lifetime risk: one in every 21 women in Africa dies of complications of pregnancy, delivery, or abortion, compared with only one in every 10,000 in Northern Europe (Rochat 1987).

Table 1.1 Selected Measures of Maternal ant Perinatal Mortality by Region and Subregion

Region/ subregion	Maternal mortality ratio, 2988 (per100,000 live births)	Total fertility rate, 1991	Lifetime risk of maternal death	Perinatal mortality rate, 1983 (per 1,000 live births)
World	370	3.4	1 in 67	57
Industrial countries	26	1.9	1 in 1,687
Developing countries	420	3.9	1 in 51
Africa	630	6.1	1 in 22	81
North	360	5.0	1 in 47	62
East	680	6.8	1 in 18	81
Middle	710	6.0	1 in 20	80
West	760	6.4	1 in 18	94
South	270	4.6	1 in 68	77
Asia	380	3.9	1 in 57	59
East	120	2.2	1 in 316	20
Southeast	340	3.4	1 in 72	52
South	570	4.4	1 in 34	87
West	280	4.9	1 in 61	55
South America	220	3.3	1 in 115
North America	12	2.6	1 in 2,671	13
Europe	23	1.7	1 in 2,132	14
Oceania	600	2.6	1 in 54	13
Commonwealth of Independent States	45	2.3	1 in 805	28

a. Maternal mortality ratio and perinatal mortality rate are explained in box 1.1.

b. The total fertility rate is the number of children a woman would bear if she lived to the end of her childbearing
years and bore children at each age in accordance with prevailing age-specific fertility rates.

c. Lifetime risk=1-(1-MMR)1.2(TFR) where the MMR is expressed as a decimal - for example, 0.2 - and the total fertility rate is adjusted by 1.2 to account for pregnancies not ending in births (Herz and Measham 1987).

Source: For maternal mortality ratio, WHO (1991b); total fertility rate, Haub, Kent, and Yanagishita (1991); and perinatal mortality rate, WHO (1989a).

Maternal mortality is not the only adverse outcome of pregnancy. Because of miscarriage, induced abortion, and other factors, well over 40 percent of the pregnancies in developing countries result in complications, illnesses, or permanent disability for the mother or child (WHO 1992a). More than 7 million newborn deaths are believed to result from maternal health problems and their mismanagement.

Poor maternal health hurts women's productivity, their families' welfare, and socioeconomic development. So safe motherhood should be a critical part of any broader strategy to expand female education and employment and to improve health, nutrition, and gender equality - and thus alleviate poverty.

Given the magnitude of these problems and the interventions available, why hasn't more been done? Unfortunately, these problems are silent. They remain, to a large extent, uncounted and unreported. The death of a pregnant women is often attributed to fate. Changing this prevailing attitude to a belief that "pregnancy is special" is the major challenge for safe motherhood programs (Winnard 1991). This paper focuses on clarifying policy and program alternatives and identifying cost-effective health-related program interventions that are likely to reduce maternal morbidity and mortality expeditiously.

Box 1.1 Maternal and Perinatal Mortality - Definitions of Ratios and Rates

Maternal mortality is defined as "death of women during pregnancy or within forty-two days of termination of pregnancy, irrespective of the duration and the site of pregnancy, from any cause related to or aggravated by the pregnancy or its management, but not from accidental or incidental causes."

Women dying from causes related to pregnancy and childbirth

Maternal mortality ratio = 100,000 live births

The ratio measures the risk women face of dying once pregnant - the obstetric risk. The maternal mortality ratio ranges from 25 to 2,000 in studies from developing countries and averages 10 in industrial countries.

Women dying from causes related to pregnancy and childbirth

Maternal mortality rate = 100,000 women age 15-49

The rate reflects the maternal mortality ratio and the fertility rate (births per 1,000 women of reproductive age); it is influenced by the likelihood of becoming pregnant and by the obstetric risk.

Perinatal mortality includes all newborn infants - stillborn or live - with a birthweight of at least 1,000 grams, who die before day 7 (168 hours).

Fetal deaths and early neonatal deaths
Perinatal mortality ratio = x 1,000

Live births

The Perinatal mortality ratio ranges from 40 to 60 per 1,000 live births in most developing countries, but it is between 6 and 10 in industrial countries.

Fetal deaths and early neonatal deaths

Perinatal mortality rate = x 1,000

Total births

Source: WHO 1992d.

(introductory text...)

Health problems associated with pregnancy affect the health and quality of life of the mother, but also of the newborn, the family, and the wider community.

Effects on the mother

About 80 percent of maternal deaths in developing countries are direct obstetric deaths - they result "from obstetric complications of the pregnant state (pregnancy, labor, and puerperium), from intervention, omissions, incorrect treatment, or from a chain of events resulting from any of the above" (WHO 1977). The remaining maternal deaths are from indirect causes aggravated by pregnancy or its management, such as malaria, viral hepatitis, diabetes, anemia, or rheumatic heart disease. Of the direct obstetric deaths, hemorrhage contributes about 25 percent, unsafe abortion at least 13 percent.

Hypertensive disorders (eclampsia) about 12 percent, infection (sepsis) about 15 percent, and obstructed labor and other direct causes about 8 percent each (figure 1.1 and appendix 1) (WHO 1992a).

Direct obstetric complications or health problems exacerbated by pregnancy can also harm the mother's health without killing her. Acute complications - many from the causes associated with maternal death - affect an estimated 50 million or more women in developing countries yearly, with more than 20 million cases serious enough to warrant referral-level care (WHO 1992a).

One of the worst chronic consequences of childbirth is obstetric fistula, a common consequence of untreated obstructed labor. Women suffering from obstetric fistula continuously leak urine and sometimes feces. They often become social outcasts. Obstetric fistula is particularly common when the first pregnancy occurs soon after puberty (Cottingham and Royston 1991).

An uncounted number of women suffer pregnancy-related disabilities long after delivery. Between 9 and 25 percent of women under forty-five years of age suffer uterine prolapse in Colombia, Pakistan, the Philippines, and Syria (Omran and Standley 1976, 1981). Because prevalence increases as a result of childbearing and its frequency, prolapse is likely to affect younger women in countries where marriage and childbearing begin early and fertility is high. In Egypt, for example, a recent community-based study in a rural area found that more than half of all women suffered from uterine prolapse (Zurayk 1991).

Figure 1.1 Medical Causes of Maternal Deaths in Developing Countries

Poorly timed unwanted pregnancies carry high risks of morbidity and mortality, as well as social and economic costs, particularly to the adolescent. In a study in Nigeria, girls age 15 and under had a maternal mortality ratio seven times that of women age 20 to 24, and in Jamaica, women over 40 were five times more likely to die during pregnancy than women age 20 to 24 (Royston 1989). Many unwanted pregnancies end in unsafe abortion. For example, when contraception and abortion were illegal in Romania in 1988, the maternal mortality ratio was reported to be 159 deaths (per 100,000 live births), 86 percent of them caused by complications from unsafe abortion. After legalization in 1989, the frequency of abortion persisted because of a scarcity of contraceptive information and supplies, but the mortality ratio fell by 50 percent in 1990 (Hord and others 1991).

Countless women suffer severe chronic illnesses that can be exacerbated by pregnancy and the mother's weakened immune system. Anecdotal information indicates that levels of these illnesses are extremely high. Malaria is more prevalent in pregnant women than in nonpregnant women and is most common in the first pregnancy. Resistance to malaria developed during childhood begins to diminish in the pregnant woman at about the fourteenth week. Viral hepatitis is far more prevalent among pregnant than nonpregnant women in developing countries and 3.5 times as likely to prove fatal (WHO 1991b). Severe viral hepatitis can lead to premature labor, liver failure, or severe hemorrhage.

In addition, an increasing number of pregnant women are testing positive for the human immunodeficiency virus (HIV), which is a precursor to acquired immune deficiency syndrome (AIDS). In Sub-Saharan Africa, 3 million women have been infected with the AIDS virus (WHO 1992c). In Lusaka, Zambia, for example, in the early 1990s, nearly one-quarter of the pregnant women attending maternity clinics were infected - a jump from 8 percent in 1985 (U.S. Bureau of the Census 1992).

From 60 to 70 percent of pregnant women in developing countries are estimated to be anemic (Sloan and Jordan 1992). An infant may secure adequate iron stores at the expense of the mother's reserves. The resulting anemia in the mother may impede her ability to resist infection or survive hemorrhage, increasing the likelihood of death in childbirth by a factor of four (Chi, Agoestina, and Harbin 1981; Llewellyn-Jones 1965).

Effects on the fetus or newborn

Pregnancy involves a dyad - the mother and the fetus. When the mother suffers, the fetus or newborn is vulnerable. Of the 13 million deaths each year in children under 5 years old in the developing world, 3 million occur in the first week after delivery. In addition, there are some 4 million stillbirths or late fetal deaths each year (WHO 1989b). These 7 million perinatal deaths are associated with maternal complications, poor management techniques during labor and delivery, and the woman's general health and nutritional status before and during pregnancy.

If a woman dies, the effect on her fetus or newborn is devastating. The overwhelming majority of pregnancies that end in a maternal death also result in fetal or perinatal death. Among infants who survive the death of the mother, fewer than 10 percent live beyond their first birthday (Koenig and others 1988; Chen and others 1974).

Even if a woman does not die, the effects of a complicated birth are also staggering. Antepartum hemorrhage, eclampsia, and other complications excluding abortion - are associated with at least 1.5 million perinatal deaths each year in developing countries, plus considerable suffering and poor growth and development for those infants who survive. Poor management during labor and delivery is associated with an additional 1.5 million perinatal deaths and many more developmental impairments among children (MotherCare 1991). For example, birth asphyxia, a lack of oxygen before birth or in the first minutes of life, kills or causes brain damage - most notably cerebral palsy - to more than 2 million children each year (CAMHADD 1990).

At least 3 million additional perinatal deaths and an unknown number of infant deaths are associated with women's health problems during pregnancy (WHO 1989b). Countless more infants who survive suffer consequences from their mother's ill health. For example, women suffering from malaria in Sub-Saharan Africa give birth to an estimated 3 million severely underweight babies (USAID 1991).

Similarly, women with poor nutritional status (short stature, low pre-pregnancy weight, inadequate weight gain during pregnancy, and anemia), reproductive tract infections, or other infections during pregnancy are more likely to deliver a low-birth-weight infant. Of the estimated 25 million low-birth-weight babies born each year worldwide, 24 million are in developing countries (WHO and UNICEF 1992). The perinatal mortality rate for low-birth-weight babies is five to thirty times higher than for fetuses or infants of normal weight. Low-birth-weight infants who survive may have serious neurological problems and hearing and visual defects and may be subject to slow development throughout life. Low-birth-weight girls are of special note: they are less likely than boys to catch up, because they are fed less, marry early, carry a heavy workload, and spend a considerable portion of their lifespan in pregnancy and lactation. Persistent low nutritional status and high energy expenditure predispose such girls to bear low-birth-weight babies themselves, passing the problem on to the next generation (Garcia and Lofti 1991).

A woman with HIV has a 25 to 40 percent chance of passing the infection on to her fetus in the womb or at birth. Using a 25 percent transmission rate, the World Health Organization (WHO) estimates that 4 million infants of HIV-infected mothers will have been born by the end of 1992 in Sub-Saharan Africa, and nearly a million are expected to be infected at birth (Chin 1990). The progression of HIV to AIDS in children is less well-documented than in adults, but according to WHO, 25 percent of the children born with HIV will be diagnosed with AIDS in the first year and 80 percent by the fourth year. As death follows about a year after diagnosis of the disease in children in Africa, 80 percent of those infected at birth - 800,000 children - are not likely to survive to their fifth birthday (Chin 1990).

Effects on the household and community

The death of a woman of reproductive age can bring economic hardship to a family in poverty. At least one-fourth of male-headed households rely on female earnings for more than half of total income (Agarwal and others 1990). The situation is worse when a woman who dies is the head of a household. Women are estimated to be the sole breadwinners in one-fourth to onethird of the world's households. In India, for example, women head one-third of all families below the poverty line (World Bank 1991).

The family loses not only the woman's contribution to household income, but also her contribution to household maintenance. Women in Africa produce most of the food necessary for a household; women in Bangladesh raise vegetables or small animals, despite their seclusion. It is the woman who cooks for the family, fetches water from a tubewell or pond, cleans the house, disburses food, and cares for the children, the sick, and the elderly at home. It is estimated that if women's unpaid household labor were enumerated, the gross national product of most developing countries would increase by about one-third (Sivard 1985). And if the work performed by women were given monetary value, the death of a woman of reproductive age would translate into a substantial financial loss.

Barriers to maternal health care

Despite the clearly demonstrated need for family planning and maternal health services, women often lack access to relevant information, trained providers and supplies, emergency transport, and other essential services. Furthermore, cultural attitudes and practices may impede women's use of services that are available. Decisions about whether to seek care are generally not the woman's alone, but are often made by the husband or mother-in-law (Thaddeus and Maine 1990; Huque and Koblinsky 1991).

Most pregnant women in the developing world receive insufficient or no prenatal care and deliver without help from appropriately trained health care providers. Only about half of the married women of reproductive age in the developing world practice contraception. In some countries in Africa, family planning and maternity care coverage is less than 10 percent (WHO 1992e). Even in countries with relatively well-developed health systems, preventable maternal morbidity and mortality persist. A study of four institutions in Mexico City classified 85 percent of the maternal deaths examined as potentially preventable; clinical or surgical misjudgement was blamed for more than eight out of ten of the preventable deaths (Bobadilla 1992).

Lessons learned

The significant reduction of maternal mortality - in an industrial or developing country - requires an active effort, one that can be managed with limited resources. Past macroeconomic growth in industrial countries did not, by itself, reduce maternal mortality. Increased availability of midwives trained to assist in home deliveries and the introduction of aseptic techniques led to a dramatic decline in maternal mortality in Sweden in the eighteenth and nineteenth centuries (Högberg and Wall 1986, Högberg, personal communication 1992). In the 1930s, maternal mortality fell in England, the Netherlands, and other European countries when antibiotics for infections, blood transfusions for hemorrhage, and improved surgical techniques for cesarean section became available (Loudon 1991). Modern family planning and safe abortion services later reduced maternal mortality further.

Socioeconomic improvements are no substitute for family planning and quality health care during pregnancy, delivery, and postpartum. Even in an industrial country, the morbidity and mortality associated with pregnancy and childbirth are high if obstetric care is not used. In the United States, in 1983, well-nourished, well-educated, and relatively affluent women of a fundamentalist religious sect had a maternal mortality ratio of 872 - compared with a national level of 8 - per 100,000 live births because they did not believe in obstetric or other modern medical care (Kaunitz and others 1984).

If a developing nation allots scarce resources appropriately, relatively low maternal mortality rates can be achieved. Appropriate services need to be easily accessible to all pregnant women, or pregnant women must be able to move closer to them when necessary.

Concentrating medical attention only on women identified as at risk has not proved effective. This is because the criteria for risk are often broadly defined and are not closely linked with adverse outcome. Recent studies have found that most of the women who are identified as having "risk factors" do not actually develop life-threatening complications and that a majority of pregnancy-related deaths result from unpredicted complications, that is, among women not identified as "at risk." Thus, risk screening is useful only when based on demonstrated risk factors and combined with monitoring, referral, and prompt treatment to deal with complications as they develop, whether predicted or not (see box 2.2 in chapter 2).

In a study of non-hospital birth centers in the United States, about one of thirteen "low-risk" women - who had an average of eleven prenatal visits - developed a serious complication (Rooks and others 1989). In Zaire, a study to predict complications during pregnancy found the best predictor was a history of problems in previous pregnancies. These women were nine times more likely to suffer obstructed labor. Still, more than two-thirds of the women with obstructed labor had been identified as low-risk (Maine 1991).

In Guatemala, referral following prediction by traditional birth attendants of complications based on demographic risk categories - for example, age or the number of previous deliveries - would overwhelm the service delivery system. Instead, a project in the rural highlands bases referrals on the detection of complications of pregnancy or delivery. There are no certified midwives at health centers in Guatemala who can assist with prenatal screening (Schieber 1991).

In a rural subdistrict of Bangladesh, maternal mortality has declined substantially in the past ten years because of new approaches to family planning and maternity care. An effective community-based family planning project has raised contraceptive prevalence to above 50 percent in the study area - compared with 23 percent in the control area - and reduced the maternal mortality rate (maternal deaths per 100,000 women of reproductive age) by about one-third. Family planning succeeded in decreasing total pregnancies, and thus the number of pregnancy-related deaths, but did not change the risk of death faced by women, once pregnant (Fauveau 1991).

In addition, an effective community-based maternity care project - added to the family planning program - reduced the risk of dying once pregnant (the maternal mortality ratio) by twothirds. The combination of basic family planning and maternal care cut the maternal mortality rate by more than one-half. The maternity care project posted trained midwives in the community to assist women with births in their homes, if requested. The midwives provided prenatal care, carried supplies to stabilize or treat women with complications, and had access to transport and referral services for cases they could not manage (Fauveau 1991).

Other projects have contributed knowledge about effective - and ineffective - interventions (appendix 2). In Indonesia, training traditional birth attendants in the absence of skilled backup support did not decrease women's risk of dying once pregnant (Alisjahbana 1991). In the Gambia, a similar approach helped reduce excessively high maternal mortality ratios, but only to a still relatively high level. There, trained traditional birth attendants provided monthly prenatal care visits to pregnant women. Each woman was also examined by a physician and treated for any illness identified. While this care resulted in a reduction of the maternal mortality ratio from 2,230 to 1,052 deaths per 100,000 live births, it remained excessively high because women who developed complications during late pregnancy or delivery could not obtain the care they needed in time (Greenwood 1991).

In Ethiopia and Nigeria, maternal mortality was reduced through prenatal screening of demonstrated risk factors and identification of danger signs (see box 2.2) - provided by certified nurse-midwives working with traditional birth attendants at the health-center or community level. The screening identified women with poor obstetric histories, and very young and very short women giving birth for the first time, discerned medical problems or complications, and referred the women before the onset of labor (Poovan, Kifle, and Kwast 1990, Brennan 1991). In the program in Ethiopia, maternity waiting homes near a rural referral hospital or health center are successfully used by women who live far away. The community constructs and maintains the homes. Thus, men have contributed financing and labor and are thus more likely to allow their wives to use the facilities (Poovan, Kifle, and Kwast 1990).

In Zaire, women's lives have been saved by delegating essential obstetric care - cesarean sections - to nurses who were readily accessible to women during birth (White, Thorpe, and Maine 1987). In Kenyatta National Hospital, Nairobi, substantial health care resources were being used to manage incomplete abortions. After introduction of the manual vacuum aspiration technique, clients and providers have benefited from shorter hospital stays, better results, and costs that have been reduced by 23 to 66 percent. Given these substantial savings, a Ministry of Health task force developed a plan to extend manual vacuum aspiration training and service delivery to all Kenyan hospitals (IPAS 1991).

In Zaria, Nigeria, the incidence of women diagnosed at the district hospital with obstetric fistulae, a common consequence of prolonged labor, was alarmingly high. Prolonged labor also contributes to maternal mortality. A radio campaign was developed to alert women to the dangers of a labor lasting more than twenty-four hours. In the several years following the campaign, the incidence of obstetric fistulae decreased significantly at the hospital serving the area of the campaign. There was no corresponding decrease at hospitals outside the reach of the campaign (Harrison 1986).

In Lahore, Pakistan, a nongovernmental organization, the Maternal and Child Welfare Association of Pakistan, established a program to provide family planning and maternal and child health services to urban slum dwellers out of reach of public services. Over three years, the association increased the contraceptive prevalence rate by 50 percent and reduced infant and maternal mortality rates, relying entirely on indigenous organizational and financial support (MCWAP 1992).

In Tunisia, the Faculty of Medicine in Sousse has institutionalized postpartum care in the family planning and maternal and child health program. The doctor or midwife who attends the birth schedules as postpartum visit for the mother and infant forty days after the birth. (In Tunisia, this marks the end of the traditional period of seclusion of the mother and her newborn.) As a result, the proportion of women and infants who receive postpartum care as well as family planning and other appropriate health care has increased substantially.

Lessons from ongoing demonstration projects in developing countries confirm the need to ensure women's access to:

· Information and education to promote health practices for safe motherhood.
· Family planning services and appropriate abortion management (appropriate treatment for complications of unsafe abortion and safe services for pregnancy termination, where legal).
· Skilled assistance during pregnancy, delivery, and the postpartum period.
· Referral services and transport for complications and emergencies.

Although all safe motherhood programs in the developing world need these priority elements to reduce maternal mortality, the exact strategy developed by a country will depend on its requirements and resources. The model presented in the next chapter illustrates the events leading to maternal deaths and the factors that should be considered in developing strategies to save lives.