Increasing the accuracy of the Expanded Programme on Immunization's cluster survey design.

The Expanded Programme on Immunization (EPI) of the World Health Organization developed a cluster survey design for the rapid assessment of children's vaccination coverage in developing countries. Because of its simplicity and familiarity, the EPI methodology has been used for additional purposes, including the detection of relatively small changes in vaccination coverage levels that are already high and the estimation of other population parameters when moderate accuracy is required. This article suggests techniques for improving the accuracy of the EPI cluster survey method, including (a) segmenting sample clusters, if necessary, so that reasonably sized areas can be counted, mapped, and listed; (b) selecting an equal probability sample of housing units within a cluster; (c) selecting a fixed number of housing units per sample cluster; and (d) performing weighted analyses. These procedures will produce accurate estimators and corresponding standard errors in EPI cluster surveys.