Abstract

PIP: The author stresses that uncritical reliance on the institutional maternal mortality rates in developing countries will provide spurious indications of improvement in this area. There appears to bean an important, although imprecisely known, differential in coverage between deliveries and maternal deaths. In 1 area, the institutional maternal mortality rate was 10 times higher among unbooked than among booked deliveries. Moreover, caution should be used in transposing maternal mortality estimates based on life table data from Europe and North America from the early 20th century to present-day Africa. Health statistics should be used to monitor health status; analyzed reduction in maternal mortality should be analyzed carefully to ensure they are valid.