A study of infant and child mortality, particularly in developing counties such as Indonesia, is still difficult to carry out due to the limited availability of data. Death registration, which is usually the basis for mortality research in develop countries is very poor and unreliable in Indonesia. Even when reliable, death registration may not provide all the required information about the socio economic, demographic, and cultural conditions of parents of the deceased and their households. Therefore, in developing countries, mortality is usually estimated from child survivorship data collected in censuses and surveys. A number of methods for estimating mortality, especially among infant and children, from census and survey data have been developed over the past five decades. Tese indirect estimation techniques are useful in the absence of a complete and accurate vital statistics registration system. The results of this study, based on the 1990, 2000, 2010, and 2015 Population Censuses and Surveys confirm a trend toward lower infant and child mortality in West Sumatera over the last three decades. The estimates also show that there are considerable urban rural differentials in childhood mortality. In general, infants born in urban areas have lower mortality rate than those born in rural areas. Keywords: rural areas, developing countries, infant and mortality.

Username
Password