Optimizing ages for cervical smear examinations in followed healthy individuals

Abstract

A method is described and illustrated for calculating near optimal planned lifetime schedules of cervical smear examinations on an individual-by-individual basis. The method can be used to minimize the sum (discounted present value) of lifetime costs for prevention, management, treatment, and follow-up, plus the expected cost or loss if death occurs prematurely. A stochastic model of health, and of disease and death if they occur, is the basis of a computer program for finding the required examination ages. The program has built-in options of data and parameter values, which the user may override with preferred other values. The built-in data include estimates drawn from the British Columbia screening program, and, for loss due to death, from Social Security Administration data. Early cervical neoplasia is treated as being potentially nonprogressive to an extent that is unpredictable in the affected individual. Illustrations are presented of the comparative effects on optimal examination schedules, and on costs and return on preventive cytology regarded as an investment, of relatively low vs relatively high likelihood of disease onset occurring, of various levels of costs, including an effect of inflation, and of false-positive reports in cytology.