Comparison of perinatal outcome in rural Tanzania as obtained from a prospective community-based survey and hospital data
Walraven, G.E.; Mkanje, R.J.; van Roosmalen, J.; van Dongen, P.W.; Dolmans, W.M.
Tropical and Geographical Medicine 46(1): 11-13
In rural Kwimba District, Mwanza region, Tanzania, researchers compared data on perinatal outcome from an August, 1989-March, 1991, community-based survey (421 deliveries) with 1989-1990 hospital data on perinatal outcome from Sumve District Hospital (3056 deliveries). The survey took place in 5 villages in the area of the primary health care (PHC) program of the hospital. Adverse perinatal outcomes in the hospital study occurred at a higher rate than those in the community survey. The perinatal mortality rate for the hospital was 96/1000 births, while it was 68/1000 for the community survey (p = .08). The low birth weight (LBW) rates were 14.6% and 9.9%, respectively (p = .01). The twin delivery rate for the hospital was 39/1000 births compared to 14/1000 births for the community (p = .02). Stillbirths made up 81% of perinatal mortality cases in the hospital, while they made up just 41% of those cases in the community (p .001). Early neonatal deaths in the hospital study were likely underreported due to early discharge (1.93% vs. 4.04% for community survey). Based on these results, hospital-based data cannot be extrapolated to communities. Hospital rates for perinatal mortality, LBW, and twinning should indeed be higher than community rates in hospitals that are an integral part of a PHC network.