Among the factors associated with maternal mortality, the "most mutable" are those to do with health care services. Two examples are the level of use of prenatal care and the rate of cesarean delivery, according to a population-based case-control study conducted in North Carolina.1 Among women who had had a live birth, those who had a cesarean section were more likely than those who had a vaginal delivery to die within one year of childbirth because of the pregnancy or its management (odds ratio, 3.9). However, pregnancy-related mortality was less likely among women who received any prenatal care than among those who did not (0.2).

To explore associations between pregnancy-related death and various aspects of health care service, researchers analyzed data from a North Carolina maternal mortality surveillance system, which matched death certificates, records of live births and fetal deaths, and autopsy or other medical reports. The researchers identified 400 women aged 10-50 who had died within one year of childbirth during 1992-1998, of whom 118 had had a live birth and died of causes directly related to or aggravated by the pregnancy or its management. To obtain unmatched controls, the researchers randomly selected 3,697 women from the 731,217 women who had had live births registered in the state in 1992-1998, ensuring that study and control groups were equally distributed over the seven-year period.

Overall, 55 deaths per 100,000 live births occurred during the study period and were attributable to any cause, and 21 per 100,000 live births were attributable to pregnancy-related causes.

Similar proportions of women in the study and control groups had received maternity care coordination assistance (30% and 23%, respectively), maternal and child nutritional services (45% and 42%), and prenatal care in a public rather than a private clinic (23% and 22%). Furthermore, similar proportions had received prenatal care that was classified as adequate according to the standards set by the American College of Obstetricians and Gynecologists (75% and 82%). However, cesarean deliveries were significantly more common among the study group than among the controls (52% vs. 16%), and the receipt of any prenatal care was more common among the control women than among those who died (99% vs. 94%). Univariate logistic regression confirmed that the likelihood of pregnancy-related death was associated with having had a cesarean rather than a vaginal birth (unadjusted odds ratio, 5.6) and with having received any prenatal care (0.2).

Further regression analyses revealed several confounding factors: Six medical conditions—eclampsia, pregnancy-induced hypertension, hypertension not induced by pregnancy, heart disease, fever during labor and diabetes—as well as older age and preterm birth (i.e., occurring before 37 weeks) were associated with significantly increased odds both of cesarean delivery and of pregnancy-related death. After adjustment for these factors, the odds of pregnancy-related death remained significantly higher among women who had had a cesarean birth than among women who had delivered vaginally (odds ratio, 3.9).

In addition, an annual income of less than $10,000 and an education below high school level were associated with an increased likelihood of pregnancy-related mortality and decreased likelihood of receipt of prenatal care. When these two confounders were accounted for, the odds of dying within one year of childbirth because of the pregnancy remained 80% lower among women who had received any prenatal care than among women who had not obtained such care (odds ratio, 0.2)

Finally, in a comparison of mortality by cause of death among all women who had had a live birth during the study period, rates for cesarean deliveries were consistently higher than those for vaginal deliveries. Overall, the rates of mortality attributable to pregnancy were approximately 36 per 100,000 cesarean deliveries and nine per 100,000 vaginal deliveries. The analysts therefore estimate that cesarean births quadruple a woman's risk of pregnancy-related death (relative risk, 3.9).

On the basis of these findings, the researchers suggest that the Healthy People 2010 objective of reducing maternal mortality to about three deaths per 100,000 live births "can be achieved through system changes." In particular, they continue, "improving use of prenatal care and lowering the cesarean delivery rate could potentially reduce pregnancy-related mortality in the United States."

—T. Lane

1. Harper MA et al., Pregnancy-related death and health care services, Obstetrics & Gynecology, 2003, 102(2):273-278.