The estimation of prevalence for the evaluation of a screening program

The evaluation of a screening program in which only screen positives are tested with more definitive diagnostic procedures requires the estimation of prevalence so that the number of screen negatives who are actually diseased can also be estimated. Methodology for incorporating an assessment of the variability of the prevalence rate between as well as within subpopulations is presented and illustrated. The procedure described leads to a negative binomial probability distribution on the number of diseased individuals who were missed by the combined screening-diagnostic protocol.Peer Reviewedhttp://deepblue.lib.umich.edu/bitstream/2027.42/24292/1/0000558.pd