An evaluation of screening policies for cervical cancer in England and Wales using a computer simulation model.

Several screening policies have been recommended for implementation in England and Wales in the last 20 years, although no evidence as to their relative effectiveness or efficiency has been provided. Using a computer simulation model, the outcomes expected from those policies had they been implemented over a 30 year period (1961-90) have been examined. The original policies based on five-yearly testing of women aged over 35 appear to be the most cost-effective, and extension of screening to younger age groups leads to loss of efficiency. Attempts to use non-screening health care contacts in order to take cervical smears (eg, during pregnancy, family planning, at gynaecology clinics) produce few advantages and considerably complicate the establishment of regular testing for the individual. The achievement of higher attendance rates is as important to the outcome of screening as concentration on more intensive or complex policies.