Estimation of early childhood mortality using preceding birth technique in a community-based setting.

The preceding birth technique, an indirect method for estimating early childhood mortality based on ascertaining the survival of previous children, was applied in rural Ethiopia within a system of demographic surveillance. Trained community health workers acted as interviewers and transmission flow of the routine health information system was used for data reporting. Data were collected for a median 12 month period from 39 sampled villages of Ticho district in Arsi region. From 1064 mothers having at least a second birth, the proportion dead of the last birth was 0.138 (95% CI 0.117-0.150), corresponding approximately to the probability of dying between birth and age two. It approximates to an infant mortality rate of about 100 x 1000 live-births. From 974 mothers having at least a higher-order birth, the proportion dead of the second to last birth was 0.203 (95% CI 0.178-0.228), equivalent to the risk of dying between 0 and age 5. In spite of the progressive migration of the surveyed population, the process and the outcome of the study suggest that PBT may well be inserted into a primary health care information system run by trained community health workers.