Background
Despite impressive improvements in maternal survival throughout the world, rates of antepartum complications remain high. These conditions also contribute to high rates of perinatal deaths, which include stillbirths and early neonatal deaths, but the extent is not well studied. This study examines patterns of antepartum complications and the risk of perinatal deaths associated with such complications in rural Bangladesh.

Methods
We used data on self-reported antepartum complications during the last pregnancy and corresponding pregnancy outcomes from a household survey (N = 6,285 women) conducted in Sylhet district, Bangladesh in 2006. We created three binary outcome variables (stillbirths, early neonatal deaths, and perinatal deaths) and three binary exposure variables indicating antepartum complications, which were antepartum hemorrhage (APH), probable infection (PI), and probable pregnancy-induced hypertension (PIH). We then examined patterns of antepartum complications and calculated incidence rate ratios (IRR) to estimate the associated risks of perinatal mortality using Poisson regression analyses. We calculated population attributable fraction (PAF) for the three antepartum complications to estimate potential risk reductions of perinatal mortality associated them.