Cost efficient sampling design for quality control of a birth defects registry

Quality control studies of registries are frequently conducted. We employed a cost effective sampling scheme for assessing the quality of case identification and abstracting in one birth defects registry. The general sampling method was a stratified two-stage design, and the optimal sample size for each stratum was chosen to minimize cost, which was defined as time for data collection. The resulting sample sizes depended on the variability in the number of abstracts between and within data collection facilities, and the amount of time needed to complete each data collection step. The most time effective scheme was to visit several facilities each for a short period of time rather than fewer facilities for a longer period of time. Cost efficient sampling strategies, such as the method used here, can be applied to ensure precision in registry quality control analyses and other public health studies.