July 2002, Vol. 8, No. 1

Women's Health
Bridging the Safe Motherhood Knowledge Gap

Reduction in maternal deaths, the goal of safe motherhood programs over the last decade and a half, has proved elusive. This may be due in part to the fact that safe motherhood programs often have been created on the basis of ideas that initially seemed good but had not been demonstrated to be effective. Two recent studies led by Council researchers have highlighted the need for identifying effective approaches before implementing them on a large scale. 

Self-reporting of symptoms 
Ninety-nine percent of maternal deaths occur in developing countries, where fewer than 50 percent of women give birth in health facilities. In these resource-poor settings, life-threatening obstetric conditions often go unidentified. Women do not know when they require special obstetric care, and in rural areas where health services are scarce, pregnancy complications are frequently not properly managed. 

Attempts to use risk factors to predict which women are likely to experience serious complications have been unsuccessful. Researchers had hoped that women’s self-reports of symptoms could be used as a means to indicate life-threatening complications requiring referral and management. 

In conjunction with the Kintampo Health Research Centre, Ghana, Council researchers conducted a retrospective study to determine whether it was possible to use women’s self-reported obstetric symptoms to identify pregnancy complications. The investigators interviewed and assessed 340 consenting women who presented themselves to the Holy Family Hospital from the seventh month of pregnancy to 42 days postpartum. 

Researchers used clinical examination and laboratory testing to determine women’s health status. Self-reported symptoms were then compared with clinical findings of obstetric complications. Sixty-seven women (20 percent) experienced one or more complications. Complex algorithms correctly identified the majority of complicated and uncomplicated pregnancies. These tools, however, missed 24 percent of cases requiring emergency obstetric care and incorrectly identified 25 percent of uncomplicated cases as requiring special obstetric management. Simple individual questions that could be used at the community level did not predict obstetric complications. 

These findings are consistent with results from studies conducted in rural settings in other developing countries. Together, these studies show that the symptoms that best identified each obstetric condition varied across cultures. Therefore, algorithms to identify which women require referral would need to be developed and tested individually for each location. Given current knowledge and diagnostics, health programs are unlikely to develop successful systems to identify and refer women for obstetric management in areas with limited health care services. 

Causes of maternal death
In developing countries where medical services are scarce, many women deliver at home. In general, understanding of the medical and social causes of maternal death is incomplete because it is based on reports, called “verbal autopsies,” of relatives and neighbors who have no medical training. In Mexico, the Council conducted a large “verbal autopsy” study to identify the causes of maternal death and disease. Data pertaining to 145 maternal deaths that occurred in 1995 in the states of Guerrero, Querétaro, and San Luis Potosí were reanalyzed. 

Using the same data, but comparing different systems of classification, the researchers’ calculations showed large variations in the attributed causes of women’s deaths. The analyses, for example, demonstrated that postpartum hemorrhage might play a role in anywhere between 14 percent and 52 percent of maternal deaths. 

In addition to demonstrating the inherent limitations of the verbal autopsy methodology, the study also reconfirmed that mortality among poor women with little access to medical care is substantially higher than that among wealthier women who have better access to institutional delivery and postpartum care. 

“This study suggests that although we know, in principle, the major medical causes of maternal death, we do not know the relative contribution of these various causes,” says Population Council program associate Nancy L. Sloan, lead author of the article. However, she contends, program managers may not always need this knowledge to save women’s lives. Over half of the 145 maternal deaths were associated with more than one complication. But preventing or managing any one of these conditions—postpartum hemorrhage, for example—could potentially save lives by reducing the gravity of the other obstetric conditions women experience.

“While we must continue our efforts to ensure adequate emergency obstetric care for all,” says Beverly Winikoff, director of the Council’s reproductive health program, “there may be ways to prevent life-threatening obstetric conditions that could feasibly and significantly reduce deaths.” Population Council researchers are currently conducting a study of a strategy for reducing the incidence and severity of postpartum hemorrhage among women having home births in rural areas. 

Sources
Sloan, Nancy L., Esi Amoaful, Paul Arthur, Beverly Winikoff, and Sam Adjei. 2001. “Validity of women’s self-reported obstetric complications in rural Ghana,” Journal of Health, Population and Nutrition 19(2): 45–51. 

Sloan, Nancy L., A. Langer, B. Hernandez, M. Romero, and B. Winikoff. 2001. “The etiology of maternal mortality in developing countries: What do verbal autopsies tell us?” Bulletin of the World Health Organization 79(9): 805–810. 

Outside funding
Canadian International Development Agency, UNICEF, and The World Bank

(Return to issue contents)