Predictors of stillbirth in sub-saharan Africa

Objective: To describe the incidence and predictors of stillbirth in a predominantly human immunodeficiency virus (HIV)-infected African cohort. Methods: Human Immunodeficiency Virus (HIV) Prevention Trials Network (HPTN) 024 was a randomized controlled trial of empiric antibiotics to reduce chorioam nionitis-related perinatal HIV transmission. A proportion of HIV-uninfected individuals were enrolled to reduce community-based stigma surrounding the trial. For this analysis, only women who gave birth to singleton infants were included. Results: Of 2,659 women enrolled, 2,434 (92%) motherchild pairs met inclusion criteria. Of these, 2,099 (86%) infants were born to HIV-infected women, and 335 (14%) were born to HIV-uninfected women. The overall stillbirth rate was 32.9 per 1,000 deliveries (95% confidence interval [CI] 26.1-40.7). In univariable analyses, predictors for stillbirth included previous stillbirth (odds ratio [OR] 2.3, 95% CI 1.2-4.3), antenatal hemorrhage (OR 14.4,95% CI 4.3-47.9), clinical chorioamnionitis (OR 20.9, 95% Cl 5.1-86.2), and marked polymorphonuclear infiltration on placental histology (OR 2.9, 95% CI 1.7-5.2). When compared with pregnancies longer than 37 weeks, those at 34-37 weeks (OR 1.7, 95% CI 0.8-3.4) and those at less than 34 weeks (OR 22.8, 95% CI 13.6-38.2) appeared more likely to result in stillborn delivery. Human immunodeficiency virus infection was not associated with a greater risk for stillbirth in either univariable (OR 1.5, 95% CI 0.7-3.0) or multivariable (adjusted OR 1.11, 95% Cl 0.38-3.26) analysis. Among HIV-infected women, however, decreasing CD4 cell count was inversely related to stillbirth risk (P=.009). Conclusion: In this large cohort, HIV infection was not associated with increased stillbirth risk. Further work is needed to elucidate the relationship between chorioamnionitis and stillbirth in African populations.