Past reproductive performance and its correlation with perinatal mortality in the current gestation at teaching hospitals in Addis Ababa, Ethiopia.

Abstract

BACKGROUND: Perinatal mortality remains a challenge in the care of pregnant women world wide, particularly, in low-resource settings. Each year, 8 million perinatal deaths occur through out the globe; 98% of them in developing countries. Data on the frequency and distribution of adverse birth outcomes and their related risk factors are important for planning maternal and child health care services in developing countries. Such information is scarce from developing settings including Ethiopia. OBJECTIVE: To determine the correlation and predictive value of past reproductive performances to the risk of perinatal death occurrence in current pregnancy among mothers who came for delivery at the three hospitals of Addis Ababa. METHODS: Hospital-based case-control study of 390 delivering women conducted at the three teaching hospitals of Addis Ababa, Ethiopia. Control subjects (n = 260) were mothers with live birth and cases (n = 130) were those who experienced perinatal death during current delivery. Information on sociodemographic, past pregnancy outcome, current pregnancy/delivery and its outcome were collected. The difference in proportion was analyzed with Pearson chi square test. The possible confounding effects between independent variables were examined with bivariate and partial correlation analysis. The predictive ability to perinatal outcome of each model as a function of these risk factors was determined. RESULTS: After adjustment for maternal age, parity, address and gestational age of index pregnancy, multivariate analysis showed the following independent variables to be associated with increased risk of perinatal deaths:--maternal illiteracy (OR = 3.0; 95% CI = 0.93 - 9.4) and birth interval < 48 months (OR = 2.3; 95% CI = 0.9 - 6.0) were marginally significant risk factors for perinatal mortality. Birth weight <2500gm (OR = 43.1; 95% CI = 12.0 - 154.6); lack of prenatal care (OR = 10.9; 95% CI = 2.3 - 51.6) and a history of past perinatal loss (OR = 10.6; 95% CI = 3.2 - 35.5) were consistently strong predictors of perinatal mortality in the current delivery. The predictive value (i.e. sensitivity and specificity) of a regression model of perinatal outcome as a function of these independent variables was 80% and 91% respectively. CONCLUSIONS: Multivariate analysis has shown the aforementioned independent variables as predictors of nearly three-fifth of perinatal mortalities in the study population. Hence, by implementing strategies directed to these risk factors, it may be possible to reduce perinatal mortality by at least 60%. Therefore, large community-based studies are recommended to validate the findings and to explain other unmeasured risk factors.