Abstract
Background. Population-based estimates of maternal mortality from Pakistan are inadequate to define the magnitude of the problem or provide information on clinical causes and determinants.
Methods. Surveys were conducted in selected clusters in Karachi, Balochistan and North West Frontier Province from 1989-1992. Pre-coded questionnaires were administered to 38,563 households to ascertain household characteristics, complete pregnancy histories and deaths of household members in the five years preceding the survey. Verbal autopsy questionnaires were then conducted to establish cause of death to women in the reproductive age group. Descriptive, bivariate and multivariable analyses were carried out to determine the association between the background variables, biological and women's status indicators and maternal mortality using a nested case-control design.
Results. Overall, the estimated maternal mortality ratio combining the data from the different sites was 433 per 100,000 livebirths. The estimated maternal mortality ratios per 100,000 livebirths ranged from a low of 281 in Karachi to a high of 673 in Khuzdar [Balochistan], Hemorrhage (52.9%), puerperal sepsis (16.3%) and eclampsia (14.4%) were the leading causes for direct maternal deaths. Logistic regression identified the important risk factors as poor housing construction material (OR=2.1; 95% CI= 1.3,3.2), distance of 40 or more miles from nearest hospital (OR = 1.3; 95%CI=0.9,1.8), grandmultigravidity (OR=1.6; 95% CI= 1.1,2.4) and prior fetal losses (OR=5.3; 95% CI=3.8,7.4).
Conclusion. Focusing on special groups of pregnant women with targeted programs such as training, monitoring and supervision of birth attendants for the provision of oxytocics, will go a long way in decreasing the proportion of maternal deaths attributed to direct, avoidable causes.