(November 2003) Arguing that too many pregnant women in developing countries lack access to good medical facilities and care, particularly when emergencies arise, some public health workers are pursuing fresh approaches to ending the death and suffering associated with pregnancy and childbirth in the world’s poorest communities.

The latest UN estimates highlight the magnitude of the problem. Roughly half a million women died of risks associated with pregnancy and childbirth in 2000, with some 95 percent of the deaths occurring in Asia (253,000) and sub-Saharan Africa (251,000). Four percent were in Latin America and the Caribbean, and less than 1 percent in more developed countries, according to new estimates by three UN agencies.1

Even though Asia and Africa had an almost equal number of deaths, the risk of maternal death is highest by far in Africa, where countries struggle to provide health services for large, dispersed, mainly rural populations and the average number of children per woman is close to 6.2 Here, a woman has a 1 in 16 chance of dying in pregnancy or childbirth over her lifetime, compared with a 1 in 94 chance in Asia, the UN agencies report. In Europe, where the average number of children per woman is less than 2 and medical care is readily available, a woman has a 1 in 2,400 chance of dying of maternity-related causes.

Most maternal deaths result from excessive bleeding, infections, hypertensive disorders, obstructed labor, or complications from unsafe abortions. Sadly, these emergencies are often difficult to predict.3 In many poor communities, women with complications face delays in deciding to seek care, getting to the appropriate health facility, and receiving treatment once they get there. Efforts to reduce these deaths have included strengthening access to prenatal care, improving women’s nutrition, trying to identify high-risk pregnancies early, and training traditional birth attendants.

None of these approaches has had a major impact on maternal mortality. A major challenge is saving lives in poorly functioning health systems. An April 2003 UN background paper on maternal and child health says that strategies including training and deploying skilled attendants and referring complicated cases to emergency facilities rarely try to find out whether the facilities function.4

Emergency Care is Key to the New ‘Averting Maternal Death and Disability Program’

A four-year-old program based at the Columbia University Mailman School of Public Health and funded by the Bill & Melinda Gates Foundation is trying a new approach. The Averting Maternal Death and Disability (AMDD) Program stresses the importance of improving health systems, particularly emergency obstetric care facilities. AMDD staff note that, even in the best circumstances, women will die if they cannot get to functioning obstetric facilities on time when sudden complications occur.

“Focusing on emergency obstetric care does not negate the importance of other kinds of activities, such as working with the community to encourage families to promptly seek care for women who are ill. Nor does it mean that all pregnant women need to deliver in medical facilities,” says Deborah Maine, the AMDD program director. Calling this a “first-things-first” approach, she argues that emergency obstetric services must first function well before the community is mobilized to seek care.5

The AMDD program, which partners with UN agencies, nongovernmental organizations, and government ministries of health in developing countries, conducts more than 50 projects in about 43 countries, assessing the way health facilities and systems function and the quality and use of emergency obstetric care.

Interest in new approaches like AMDD’s is being fueled by frustration with persistently high levels of maternal deaths in less developed countries as well as by weaknesses in traditional ways of tracking the problem. Determining maternal mortality ratios — the number of deaths for every 100,000 live births in a population — relies on the gathering of accurate data on the deaths and their causes. These are difficult to come by for a variety of reasons: Many deaths occur outside of health systems and are not recorded; health workers may not always know the causes of death; collecting the data is costly; and calculating the numbers is complex.

Using UN Goals to Measure Progress

The urge for new approaches has increased in light of the need to measure countries’ progress toward the Millennium Development Goals set by world leaders in 2000. These include a 75 percent reduction, between 1990 and 2015, in the maternal mortality ratio.

“Now is the time to take up the challenge,” says Wendy Graham, the main researcher for the Initiative for Maternal Mortality Programme Assessment (IMMPACT). “Health sector reforms are being implemented in a number of less developed countries. It has long been recognized that maternal health services are dependent on the functioning of the entire health system. … Data are needed to ensure that the most effective and cost-effective intervention strategies for safe motherhood are integrated into essential service packages, and also to track the impact of the reform process using realistic and affordable information systems.”6

To achieve these goals, attention has focused on measures released in 1997 by the World Health Organization (WHO), the United Nations Children’s Fund (UNICEF), and the United Nations Population Fund (UNFPA). Known as “UN process indicators,” they provide policymakers and those designing maternal health programs with new kinds of information. While maternal mortality ratios point to levels and frequency of maternal deaths, the UN process indicators monitor the progress of health programs in reducing these deaths. According to the UN agencies, the indicators provide information that can help plan programs, highlight problems, and evaluate programs’ success.7

These indicators are already in use to assess health programs in developing countries. For the AMDD program, use of the indicators, following UN guidelines, has provided a range of information on emergency obstetric care facilities, including the following:

  • The number and distribution of emergency obstetric care facilities. To address problems related to timely access to care, the indicators suggest that every 500,000 people need at least one facility that offers comprehensive care and four that offer basic care. Basic care includes antibiotics, manual removal of the placenta, and assistance in vaginal delivery of the baby. Comprehensive care includes all these services as well as Caesarean sections and blood transfusions.
    In Bhutan, a country characterized by mountainous terrain, dispersed communities, and difficult communication, UNICEF and AMDD have supported the government in improving the geographic distribution of emergency obstetric care. In this population of some 900,000 people where an estimated 255 women die for every 100,000 live births,8 only eight facilities offered emergency obstetric care in 2000. By making policy changes and upgrading operating theaters, labor rooms, and other key facilities at district hospitals, the government boosted that number to 17 by mid-2003.9 The government plans to have 29 emergency obstetric care facilities functioning by 2007.
  • The proportion of cases admitted to emergency obstetric facilities. The UN agencies estimate that 15 percent of all pregnant women develop complications; the agencies therefore expect that percentage of pregnant women to seek emergency services.
    A survey to evaluate emergency obstetric care services in five of Cameroon’s 10 provinces revealed that the proportion of births delivered in emergency obstetric facilities — less than 6 percent — was well below the 15 percent recommended level. This knowledge has led the Ministry of Public Health, with support from UNFPA, to develop a pilot project in two districts to train staff, set up a referral system, and improve the monitoring and evaluating of services.10
  • The met need for emergency care. All women with obstetric complications should be treated in emergency obstetric facilities, according to the guidelines.
    A project supported by AMDD and UNFPA in Rajasthan, India, helped train 53 teams of doctors and paramedics in emergency obstetric care as it sought to increase access to basic emergency services near rural communities. As a result, the number of facilities staffed to provide such care increased from 26 in 2000 to 52 in 2003, and the percentage of all women with complications who received treatment in emergency facilities climbed from roughly 9 percent to 14 percent (this percentage is still very low, but it does show progress).11
  • The number of deaths among women admitted to the facilities with complications. Deaths should not exceed 1 percent of the women admitted, according to the UN guidelines.
    A project supported by AMDD and UNICEF aimed to reduce maternity-related death and disability in three districts in Sindh Province, Pakistan. Of the 24 facilities that the project targeted for strengthening, only two provided 24-hour emergency care, seven days a week. By the end of 2002, 19 facilities were providing round-the-clock emergency care, and fatality rates had dropped 50 percent.12

While UN agencies do not recommend abandoning traditional measures of maternal mortality, such as the maternal mortality ratio, they note that the process indicators can answer such important questions as why women are dying and how the deaths can be prevented. In the meantime, “Further research is needed to identify cost-effective and reliable ways of measuring maternal mortality,” notes Maternal Mortality in 2000: Estimates Developed by WHO, UNICEF, and UNFPA.

Yvette Collymore is senior editor at PRB. 


  1. World Health Organization (WHO), Maternal Mortality in 2000: Estimates Developed by WHO, UNICEF, and UNFPA (Geneva: WHO, 2003), accessed online at www.who.int/reproductive-health/
    publications/maternal_mortality_2000/maternal_mortality_2000.pdf, on Nov. 1, 2003.
  2. Carl Haub, 2003 World Population Data Sheet (Washington, DC: Population Reference Bureau, 2003).
  3. WHO, Mother-Baby Package: Implementing Safe Motherhood in Countries (Geneva: WHO, 1994).
  4. Lynn Freedman et al., “Background Paper of the Task Force on Child Health and Maternal Health,” Millennium Project: Commissioned by the UN Secretary General and Supported by the UN Development Group (April 18, 2003).
  5. Address delivered at the AMDD Network Conference 2003, Kuala Lumpur, Malaysia, Oct. 21-23.
  6. Wendy J. Graham, “Now or Never: The Case for Measuring Maternal Mortality,” The Lancet 359 (Feb. 23, 2002): 701-704.
  7. UNICEF, Guidelines for Monitoring the Availability and Use of Obstetric Services (New York: UNICEF, 1997).
  8. National Health Survey 2000, obtained from WHO’s “Country Health Profile: Bhutan,” accessed online at http://w3.whosea.org/cntryhealth/bhutan/bhustatics.htm, on Nov. 18, 2003.
  9. Hemlal Sharma, “Setting up Emergency Obstetric Care Facilities in Resource-Poor Setting: Experience from Bhutan” (PowerPoint presentation made at the AMDD Network Conference 2003, Kuala Lumpur, Malaysia, Oct. 21-23, 2003).
  10. Suzanne Maiga Konaté and Amadou Traore, “Strengthening Emergency Obstetric Care in Cameroon” (book of abstracts for the AMDD Network Conference 2003, Kuala Lumpur, Malaysia, Oct. 21-23, 2003).
  11. Venkatesh Srinivasan, Hemant Dwivedi, and Dinesh Agarwal, “Increasing Access to Basic Emergency Obstetric Services: Process Adopted and Experiences in Rajasthan, India” (PowerPoint presentation made at the AMDD Network Conference 2003, Kuala Lumpur, Malaysia, Oct. 21-23, 2003).
  12. Talat Rizvi, “Availability of Skilled Staff for 24-Hour Emergency Obstetric Care: Meeting the Challenge in Pakistan” (PowerPoint presentation made at the AMDD Network Conference 2003, Kuala Lumpur, Malaysia, Oct. 21-23, 2003).