Reducing Maternal Mortality in Developing Countries

A note on this page's publication date

The content on this page has not been recently updated. This content is likely to be no longer fully accurate, both with respect to the research it presents and with respect to what it implies about our views and positions.

Published: 2009

  • A woman in a developing country is 97 times more likely to die as a result of pregnancy than a woman in a developed country.
  • We have not identified an intervention whose effectiveness at reducing maternal mortality is strongly supported by the available evidence. We have seen a track record of programs being recommended without strong evidence and recommendations being changed over time in response to lack of progress.
  • We have not found any charity doing demonstrably successful work in this area.

This page reviews what we know of the track record of programs focused on reducing maternal mortality. We have not identified an intervention whose effectiveness at reducing maternal mortality is strongly supported by the available evidence. We have seen credible success stories, but they have hinged on broad, systemic improvements to the provision of health care, which is likely outside the scope of a program run by a charity.

Within this cause, there is a track record of programs being recommended without strong evidence and recommendations being changed over time in response to lack of progress. Therefore, donors should be wary of current recommendations not supported by strong evidence. (Note: Some of the programs discussed below may reduce infant mortality, but here we focus on maternal mortality.)

Table of Contents


The World Health Organization (WHO) estimates that in 2005 over 500,000 women died from pregnancy- and birth-related causes.1 A woman in a developing country is 97 times more likely to die as a result of pregnancy than a woman in a developed country.2 The majority of these deaths occur during and immediately following birth: 25% are caused by severe bleeding, 15% by infection, 12% by eclampsia (a seizure disorder), and 8% by obstructed labor. The remaining deaths are due to unsafe abortion (13%), other direct causes (8%), and indirect causes such as HIV and malaria which may be aggravated by pregnancy.3 The technologies needed to prevent deaths from most of these causes exist. For this reason, the World Health Organization designates such deaths as "avoidable."4 Additionally, World Health Organization claims that the interventions to reduce maternal mortality are cost-effective, but does not present evidence to support this claim.5

Charities working to reduce maternal mortality in developing countries are involved in a wide range of activities including training traditional birth attendants,6 providing skilled care at birth,7 and distributing clean delivery kits.8 Other interventions not detailed in this report include safe abortion and post-abortion care,9 family planning,10 and obstetric care.11

Summary of the evidence

Although the medical interventions needed to prevent pregnancy-related deaths exist, programs to reduce maternal mortality in developing countries have a mixed track record. On a "macro" level, the success of Sri Lanka in dramatically reducing maternal mortality over the past half century is evidence that long-term government commitment to broad, systematic improvement of health services for pregnant women can save lives effectively in a low-income country. On a "micro" level, however, we have not found rigorous evidence for the effectiveness of many seemingly logical interventions.

Since a majority of maternal deaths occur during and soon after delivery, many interventions concentrate on this period. Traditional birth attendants (TBAs) assist many developing-world mothers during birth. Programs have attempted to utilize this existing system by giving short training courses to TBAs. There is little evidence that such programs are effective in reducing maternal mortality, though they may be effective in reducing mortality among newborns (more below). Efforts to increase the number of births attended by skilled attendants also hope to reduce deaths around the time of delivery, but are not associated with strong evidence of effectiveness (more below). Clean delivery kits may help reduce infection during birth, but the evidence available is neither conclusive nor rigorous (more below).

There is some evidence that having fewer medical visits per pregnancy does not increase the risk of death (more below). One promising, yet not thoroughly studied intervention, is the creation of facilitator-led community groups for pregnant women (more below).


The sources for the research on this page were drawn primarily from two online databases: the medical journal database PubMed and the Cochrane Library review database. For each intervention listed below, we used a number of combinations of applicable terms, and we explored the 'related articles' suggested by the database. We also browsed the Cochrane Library's category "Pregnancy and Childbirth." Priority was given to reviews and meta-analyses over studies that examined a single project or experiment. Relevant articles from these databases were used as sources of references to other relevant articles. Additional sources came from the World Health Organization's website, and Google and Google Scholar search engines.

Large-scale successes in reducing maternal mortality

One of 20 case studies in Millions Saved: Proven Successes in Pubic Health is devoted to the reduction of maternal mortality in Sri Lanka. Since 1950, Sri Lanka has reduced maternal deaths "from between 500 and 600 maternal deaths per 100,000 live births in 1950 to 60 per 100,000."12 Levine (2007) attributes this decline to four major factors:13

  • Broad, free access to a strong health system.
  • The professionalization and broad use of midwives.
  • Gathering of health information and use of this information for policy making.
  • Targeted quality improvements to vulnerable groups.

Sri Lanka accomplished its large reduction in maternal mortality while spending a smaller percentage of GDP on health than most countries at its income level.14

Maternal mortality decreased more rapidly than female death rates in general. Also, death rates from specific causes of maternal mortality, such as hypertensive disease and sepsis, fell. This suggests that maternal mortality fell due to factors other than general improvements in health.15

A World Bank study looked at seven countries that have had some success in reducing maternal mortality and concluded that, given current technology, it was unlikely that countries could speed up the process that took Sri Lanka decades.16 The study also named six "factors of success," which largely coincided with the factors listed by Levine in Millions Saved (above) and highlighted the importance of a concerted government effort to address maternal mortality.17

Country-level success stories are useful in showing the feasibility of the goal of reducing maternal mortality. It is not possible with this type of evidence, however, to establish a cause-and-effect relationship between a particular intervention and falling maternal death rates. Interventions were implemented concurrently and there was no control group used to see what would have happened to maternal mortality rates without the interventions.

(More on our interpretation of “macro evidence”.)

Evidence on specific interventions

Training traditional birth attendants

In developing countries, many births are assisted by "traditional birth attendants" (TBAs), who acquire their skills through experience and apprenticeship,18 rather than through the formal training that characterizes "skilled birth attendants" (which include doctors, midwives, and nurses).19 Programs to provide short training courses to TBAs, to teach them how to respond to minor complications and to recognize and refer major complications, were recommended by the World Health Organization in the 1970s through 1990s. The World Health Organization believed that such training courses could reduce maternal mortality rates.20

There's very little strong evidence that training TBAs is an effective program for reducing maternal mortality.21 The available evidence suggests that TBA training increases knowledge among TBAs and may reduce infant mortality, but does not have an demonstrable impact on maternal mortality.22

A 2007 Cochrane Review found only one rigorous study that measured the relationship between training of TBAs and maternal mortality. The study was a large, randomized controlled experiment in Pakistan. It evaluated the effect of a three-day training program for TBAs "in the context of rural homebirth where TBAs, women and families have access to an improved health system."23 The study found reductions in death rates for newborns, but did not find a statistically significant decrease in maternal death rates (though there were fewer maternal deaths in the intervention group than in the control group).24

It is difficult to make conclusions about the effectiveness of TBA training programs in general from this one study. Training programs vary in length, content, clinical practice, and supervision.25 Despite these difficulties, the Cochrane review concluded, "The potential of traditional birth attendant (TBA) training to reduce perinatal mortality is promising when combined with improved health services."26

There are a number of reasons why training TBAs may fail to reduce maternal mortality. These include:27

  • Lack of medical services to which to refer women with major complications.
  • Barriers to learning due to lack of formal education among TBAs.
  • Training that is insufficient to give TBAs the skills to perform life-saving interventions.

The World Health Organization now recommends that countries work toward the goal of having every birth attended by a skilled birth attendant—a doctor, midwife, or nurse who has received formal education in the management of pregnancy and childbirth.28

Skilled birth attendants

The World Health Organization advocates for expanded use of skilled birth attendants to reduce maternal mortality.29 Evidence for the use of skilled attendants primarily relies on non-experimental analysis,30 and it is still unclear whether or by how much it may reduce maternal mortality.31

Two recent studies used non-experimental methods to study the relationship between births attended by skilled attendants and maternal mortality. Neither study found a strong link between the two, though the limitations in the design of these studies makes us approach any conclusion with caution.

The first compared two districts in Burkina Faso, one that received a number of interventions designed to increase use and effectiveness of skilled attendants and another that received a much more limited set of services. The study found no statistically significant difference in maternal mortality rates between the two districts.32 The second used country-level data and found that a compelling case for a relationship between skilled attendance and maternal mortality could not be made.33

Challenges faced by a program to expand the use of skilled attendants include the inadequate supply of midwives and doctors, lack of health facilities to which to refer complicated cases, and reluctance among women to use such services.34

Antenatal care

Antenatal care is comprised of a number of interventions administered to women during pregnancy, including screening tests, immunizations, and treatment for identified complications.35 A 2001 review found that evaluation of the effectiveness of antenatal care in preventing maternal deaths was sparse.36 We have not found any studies conducted since then that directly address the effectiveness of antenatal care.

The World Health Organization (WHO) reviewed studies that compared standard models of antenatal care with models that reduced the number of visits a woman had per pregnancy. They found seven randomized controlled trials, which included over 60,000 women. World Health Organization concluded that fewer visits did not result in higher maternal mortality rates.37 Since a model with fewer visits is less expensive than the standard model,38 cost-effectiveness considerations call for a reduction in the number of visits, especially in resource-poor areas, where funds and staff time may be urgently needed elsewhere.

The World Health Organization has suggested some reasons why antenatal care may fail to improve maternal outcomes. These reasons include:

  • Difficulty in predicting birth complications during pregnancy.39
  • Lack of communication between antenatal and delivery care personnel.40
  • Poor quality of antenatal services.41

While the latter two could be remedied with improved program design, questions remain about the effect of even the best designed antenatal programs on maternal mortality.

The Disease Control Priorities in Developing Countries report notes that experts recommend a number of specific interventions during pregnancy to protect infants. These include screening and antibiotic treatment for syphilis and immunization against tetanus.42 We do not discuss these interventions here, as they are focused on infant (as opposed to maternal) mortality.

Community mobilization

A recent review of community-level interventions to reduce maternal mortality found only one randomized controlled study that did not focus on training of traditional birth attendants (see above) or comparing antenatal care models (see above). The study, conducted in Nepal, examined the effect of "facilitator-led women's groups to improve perinatal care practices."43 It attributed a large, statistically significant reduction of maternal mortality to the intervention, though the author noted some problems with the study design that could raise doubts about the result.44 The authors of the study estimated the cost of the intervention at $3442 per infant life saved (which would be a significantly higher cost per life saved than that of our priority interventions), but did not provide a cost-effectiveness analysis for maternal lives saved.45 This is a promising area of research, but one study in one location is not enough to convince us that such a program should be implemented in other places without high-quality monitoring and evaluation.

Clean delivery kits

According to the World Health Organization, 15% of maternal deaths are due to infection.46 Programs that provide clean delivery kits hope to reduce infections among mothers delivering at home and in health centers, as well as among their infants.47 Kits include such items as soap for washing of hands and vagina, clean razors and cord ties for cutting the umbilical cord, plastic sheets for creating a clean delivery surface, and a pictorial instruction sheet for directing mothers and their attendants on how to use the items in the kit.48

The effect of clean delivery kits on infection rates is not entirely clear. A study in Tanzania found significant reductions in infections among women who used the kits and were taught World Health Organization recommended hygienic procedures, and an even larger reduction among their infants.49 The study was not a randomized controlled trial, so concerns about the validity of the relationship between the kits and infection rates remain.50 Additionally, it is not possible to say whether the kits, the hygiene lessons, or the combination of the two was responsible for the outcome.


  • 1

    "In 2005, there were an estimated 536 000 maternal deaths worldwide." World Health Organization, "Maternal Mortality."

  • 2

    "A woman's lifetime risk of maternal death is 1 in 7300 in developed countries versus 1 in 75 in developing countries." World Health Organization, "Maternal Mortality."

  • 3

    "The four major killers are: severe bleeding (mostly bleeding postpartum), infections (also mostly soon after delivery), hypertensive disorders in pregnancy (eclampsia) and obstructed labour. Complications after unsafe abortion cause 13% of maternal deaths. Globally, about 80% of maternal deaths are due to these causes. Among the indirect causes (20%) of maternal death are diseases that complicate pregnancy or are aggravated by pregnancy, such as malaria, anaemia and HIV." World Health Organization, "Maternal Mortality."

  • 4

    "Most maternal deaths are avoidable, as the health care solutions to prevent or manage the complications are well known." World Health Organization, "Maternal Mortality."

  • 5

    "For women to benefit from those cost-effective interventions they must have antenatal care in pregnancy, in childbirth they must be attended by skilled health providers and they need support in the weeks after the delivery." World Health Organization, "Maternal Mortality."

  • 6

    See, for example, (accessed May 4, 2011). Archived by WebCite® at

  • 7

    See, for example,

  • 8

    See, for example, (accessed August 26, 2009). Archived by WebCite® at

  • 9

    See, for example,

  • 10

    See, for example,

  • 11

    See, for example,… (accessed August 26, 2009). Archived by WebCite® at

  • 12

    Levine 2007, Pg 1.

  • 13

    Levine 2007, Pg 1.

  • 14

    "Sri Lanka has spent less on health - and achieved far more - than most of the countries at similar income levels. In India, for example, the maternal mortality ratio is more than 400 per 100,000 live births, and spending on health constitutes over 5 percent of GNP. In Sri Lanka, the ratio is less than one-quarter of that, and the country spends only 3 percent of GNP on health." Levine 2007, Pg 1.

  • 15

    "One way to answer the question of whether system changes caused declines in maternal deaths is to compare the overall decline in female deaths with deaths due to maternal causes...By 1996, while both maternal and all female deaths declined, maternal causes accounted for only 1.2 percent of all female deaths in the reproductive age range. Another way to understand the cause-effect relationship is to look at the changes in maternal deaths due to individual causes known to be associated with specific health care delivery strategies. So, for example, deaths due to hypertensive disease and sepsis—two causes that are associated throughout the world with lack of access to skilled attendance—declined dramatically during the 1940s." Levine 2007, Pg 6.

  • 16

    "Can current program strategies reduce maternal mortality more quickly than the decades required in the historically successful countries of Malaysia and Sri Lanka? The answer reached after conducting case studies and primary research on safe motherhood programs in seven countries is, no." Koblinsky 2003, Pg 1.

  • 17

    "Six factors of success":

    • “Increased availability of a skilled birth attendant"
    • “Increased availability of health facilities to provide skilled birthing care”
    • “Service costs appropriate for the setting”
    • “Strong policy guidance for delivery care”
    • “A functioning referral system, beginning with providers at the community-level
    • “Accountability for providers' performance”

    Koblinsky 2003, Pgs 15-30.

  • 18

    Sibley et al. 2007, Pg 2.

  • 19

    World Health Organization, "Skilled Birth Attendants."

  • 20

    “Between the 1970s and 1990s, the World Health Organization promoted traditional birth attendant (TBA) training as one strategy to reduce maternal and neonatal mortality.” Sibley et al. 2007, abstract.

  • 21

    "Yet, after more than three decades of experience, the evidence to support TBA training has been limited and conflicting. The impact of TBAs on maternal and neonatal mortality is uncertain, fuelling a continuing debate over the cost-effectiveness of TBA training in relation to the global Safe Motherhood Initiative (Bang 1994; Bang 1999; Bergstrom 2001; Fortney 1997a; Kumar 1998; Levitt 1997; Maine 1992, Maine 1993; Rahman 1982; Sibley 2004a; Starrs 1998; Tinker 1993; UNICEF1997; WHO 1992)." Silbey et al. 2007, Pgs 2-3.

  • 22

    "Main findings include moderate to large, positive effect sizes for MCH [maternal and child healthcare] knowledge, attitudes, behaviour, and advice associated with training, with small effect sizes associated with perinatal health outcomes, such as the effect size for cause-specific neonatal mortality due to birth asphyxia indicate an 11% decrease from the untrained TBA baseline. The data were not sufficient to document an association between training and maternal mortality (Sibley 2002; Sibley 2004c)." Silbey et al. 2007, Pg 3.

  • 23

    Sibley et al. 2007, Pg 19.

  • 24

    “One cluster-randomized trial found significantly lower rates in the intervention group regarding stillbirths (adjusted OR 0.69, 95% confidence interval (CI) 0.57 to 0.83, P 0.001), perinatal death rate (adjusted OR 0.70, 95% CI 0.59 to 0.83, P 0.001) and neonatal death rate (adjusted OR 0.71, 95% CI 0.61 to 0.82, P 0.001). Maternal death rate was lower but not significant (adjusted OR 0.74, 95% CI 0.45 to 1.22, P = 0.24) while referral rates were significantly higher (adjusted OR 1.50, 95% CI 1.18 to 1.90, P 0.001).” Sibley et al. 2007, abstract.

  • 25

    "Training programs differ considerably in the way in which they address these objectives (Fortney 1997a). TBAs are trained by individuals, non-governmental organizations and missions, as well as by local, state and national governments. The training programs range from very basic to quite elaborate and may last from several days to several months. They may, but often do not, include clinical practice at a referral facility. They may, but sometimes do not, include continued contact with trained TBAs through supervision and further education. The content of TBA training also varies but usually includes performance of hygenic deliveries and cord care and use of appropriate techniques for delivery of the placenta to prevent immediate postpartum haemorrhage." Silbey et al. 2007, Pg 3.

  • 26

    Sibley et al. 2007, Pg 19.

  • 27

    Piper 1997, abstract.

  • 28

    World Health Organization, "Skilled Birth Attendants."

  • 29

    "In issuing this statement, WHO, ICM and FIGO are advocating for skilled care during pregnancy, childbirth and the immediate postnatal period." World Health Organization 2004, Pg 1.

    See also the United Nations Population Fund (UNPFA)'s article, "Skilled Attendance at Birth."

  • 30

    "Evidence on the effectiveness of the alternative models at a population level is lacking, and support for skilled attendance at delivery is, thus, based primarily on historical and contemporary ecological analysis (De Brouwere and Van Lerberghe 2001)." Jamison et al. 2006, Pg 512.

  • 31

    "Although skilled attendance at delivery has been widely recommended for pregnancy-related mortality reduction (WHO 2005; Horton 2006), there remain questions about the best means for implementing this strategy, and the magnitude of its impact on pregnancy-related mortality remains uncertain." Hounton et al. 2008, Pg 53.

  • 32

    “We did not find any significant impact of the SCI on pregnancy-related mortality within the reference period of the study. This study highlights the challenges in evaluating progress in safe motherhood even with resources available to carry out a census of about 2.5 million person-years." Hounton et al. 2008, Pg 59.

  • 33

    "There are some populations in which more than a quarter of their deliveries occur without health professionals but the level of maternal mortality is below 250 per 100,000 live births, such as Peru, Tunisia, Egypt and Namibia. Conversely, there are other countries with nearly half of their deliveries with health professionals but maternal mortality remains high – above 500 per 100,000 live births, such as Malawi, Ghana, Bolivia, and Zambia." Graham, Bell, and Bullough 2001, Pg 125.

  • 34

    "The main obstacles to the expansion of care are the dire scarcity of skilled providers and health-system infrastructure, substandard quality of care, and women's reluctance to use maternity care where there are high costs and poorly attuned services." Koblinsky et al. 2006, abstract.

  • 35

    Bergsjø 2001, Pg 41, Table 3.

  • 36

    "As in other areas of maternal health, good-quality evidence is scarce and, just as in many aspects of health care generally, there are interventions in current practice that have not been subjected to rigorous evaluation." Carroli, Rooney, and Villar 2001, abstract.

  • 37

    "There was no clinically differential effect of the reduced number of antenatal visits when the results were pooled for pre-eclampsia (typical odds ratio 0.91 [95% CI 0.66-1.26]), urinary-tract infection (0.93 [0.79-1.10]). postpartum anaemia (1.01), maternal mortality (0.91 [0.55-1.51]), or low birthweight (1.04 [0.93-1.17])." Carroli et al. 2001, abstract.

  • 38

    "The cost of the new model was equal to or less than that of the standard model." Carroli et al. 2001, abstract.

  • 39

    "Most life-threatening complications of pregnancy and childbirth are difficult to predict or prevent." Safe Motherhood Inter-Agency Group 1997, Pg 26.

  • 40

    "It is often separated from delivery care - provided in a different location, by different personnel who have little communication with the providers of delivery care." Safe Motherhood Inter-Agency Group 1997, Pg 26.

  • 41

    "It is frequently of poor quality - provided in a rote fashion, with information gathered and recorded but no analysis conducted and no action taken." Safe Motherhood Inter-Agency Group 1997, Pg 26.

  • 42

    "Many health experts, however, do accept screening and treatment for syphilis and immunization with tetanus toxoid as important prenatal interventions." Jamison et al. 2006, Pg 512.

    "In both industrialized and developing countries, but particularly the latter, the prevention of congenital syphilis by antenatal screening is cost-effective and may be cost-saving. Yet, globally, there are probably >500 000 fetal deaths a year from congenital syphilis, a figure rivalling that from mother-to-child transmission of human immunodeficiency virus (HIV), which receives far greater attention." Schmid 2004, abstract.

    "Available evidence supports the implementation of immunisation practices on women of childbearing age or pregnant women in communities with similar, or higher, levels of risk of neonatal tetanus, to the two study sites." Demicheli et al. 2005, abstract.

  • 43

    Kidney et al. 2009, Pg 5, Table 2.

  • 44

    Possible problems with study design include minor selection bias, too short follow up period, and contamination of intervention (treatment not limited to targeted intervention). Kidney et al. 2009, Pg 5, Table 1 and 2.

  • 45

    "The cost per newborn life saved was US$3442 ($4397 including health-service strengthening costs) and per life year saved $111 ($142 including health-service strengthening costs). This value compares favourably with the World Bank's recommendations that interventions less than US$127 per disability-adjusted life year saved are some of the most cost effective." Manandhar 2004, Pg 977.

  • 46

    World Health Organization, "Why Do So Many Women Still Die in Pregnancy or Childbirth?"

  • 47

    "Basic delivery kits can increase awareness and use of clean delivery practices. The kits are designed for use in the home by untrained and trained birth attendants (TBAs) and women delivering alone. Basic delivery kits contain supplies that are essential for supporting clean delivery practices and providing clean cord care immediately after birth. While basic delivery kits are designed for use in the home, they can also be used in resource-poor medical facilities such as health posts or health centers." Program for Appropriate Technology in Health 2001, Pg 1.

  • 48

    Program for Appropriate Technology in Health 2001, Pgs 74-75.

  • 49

    "Newborns whose mothers used the delivery kit were 13.1 times less likely to develop cord infection than infants whose mothers did not use the kit. Furthermore, women who used the kit for delivery were 3.2 times less likely to develop puerperal sepsis than women who did not use the kit...Single-use delivery kits, when combined with education about clean delivery, can have a positive impact on the health of women and their newborns by significantly decreasing the likelihood of developing puerperal sepsis or cord infection." Winani et al. 2007, abstract.

  • 50

    "A stepped-wedge, cross-sectional study was conducted in 10 surveillance sites across two rural districts of Mwanza Region, Tanzania. A total of 3262 pregnant women between the ages of 17 and 45 years were enrolled in the study. Village health workers administered questionnaires to each mother at 5 days postpartum and inspected the infants' umbilical cord stumps for signs of infection." Winani et al. 2007, abstract.