Using routine surveys to measure mortality: a tool for programme managers.

Aid donors and recipients have begun to demand timely, population-based information for programme planning and for measuring health programme performance. Results from trials in Jordan, Syria, Djibouti and People's Democratic Republic of Yemen show that widely-used routine surveys for estimating vaccination coverage can be adapted to collect data on health indicators such as child and maternal mortality. Estimation methods must be robust and fieldwork well-supervised. Adding questions about total children ever born and surviving, the survival of the preceding birth, and the survival of sisters to such surveys, population-based estimates of the trend and recent level of childhood mortality and of the lifetime risk of maternal death can be obtained. These trials indicate that the need to monitor selected health indicators could be met through inexpensive, low-technology surveys.