Effects of community factors on infant and child mortality in rural Bangladesh.

Data from the 1975-76 Bangladesh Fertility Survey (BFS) were used to analyze the influence of community factors on infant and child mortality in rural Bangladesh. The BFS community data provide information on the relative accessibility of health and maternal and child health/family planning (MCH/FP) facilities for each rural village in the sample. In the 1st stage of the analysis infant and child mortality rates were calculated for births classified by demographic and sociodemographic variables and by the accessibility of health and MCH/FP services. In the 2nd stage the rates for the demographic and socioeconomic variables were calculated with adjustments for the associations among these variables. In the 3rd stage the strength of the relationships between community variables and the infant and child mortality rates were considered with controls for the demographic and socioeconomic characteristics of the child and its parents. The level of infant and child mortality in rural Bangladesh remains very high. The mortality trends during the last 30 years show some decline yet the decline has not been substantial. The BFS data indicate that infant mortality fell between 1951-55 and 1966-70 but then remained stable between 1966-70 and 1971-74. In the BFS a total of 6513 ever-married women below age 50 residing in 5855 households were interviewed. Among the significant determinants of infant and child mortality length of the previous interval emerged as dominant. Among rural women the range in rates between births following an interval of less than 18 months and those following an interval of more than 36 monthis was 185:37 for neonatal mortality 86:29 for postneonatal mortality 271:66 for infant mortality and 140:50 for child mortality. The results suggest that if all mothers in Bangladesh spaced their children by at least 3 years infant and child mortality would be reduced by 1/3 or more. The birth order of a child had a very strong effect on survival in the 1st year of life but the effect diminshed thereafter. Education had a significant effect on mortality but the effect was confined to child mortality. The effect of religion on mortality increased with age of the child. Muslim children experienced lower mortality than did Hindu children. In analysis of effects of the community variables the distance to a family planning clinic and the distance to a primary school emerged as the strongest correlates of mortality. Birth spacing should get national priority in the information education and communications and training programs and there should be continued investment in the maintenance and expansion of family planning clinics and hospitals.