In carrying out surveys to estimate disease prevalence or health services coverage in a region, investigators are often interested in identifying communities or other subareas in which the problem of interest is especially severe. In view of the large number of subareas to be examined, the investigator seeks rapid methods of assessment that require minimum sampling in each. Lot quality assurance sampling, developed more than 60 years ago to meet industrial quality control needs, has been found to serve public health purposes as well. Curiously, other more recent and more powerful industrial techniques have not received the recognition they deserve. Several of them are discussed in this paper and likely applications are cited. They fall into two categories: 1) lot acceptance methods to depict static conditions in a population and 2) more dynamic process monitoring methods for continuous surveillance. Under the first category, techniques of double sampling and sequential sampling are described. Reduced and tightened scrutiny, control charts, and exponential smoothing are highlighted as especially promising approaches under the process control category.
