Conclusions cannot yet be drawn

EDITOR—Sherlaw-Johnson et al evaluated policies for withdrawing women from the cervical cancer screening programmes before the recommended age of 64, using a mathematical model.1 Their results were obtained with specific and uncertain model assumptions, which were insufficiently subjected to validation and sensitivity analysis.

From the description of the model in cited earlier papers, most new cases of cervical intraepithelial neoplasia seem to originate at younger ages. The duration is assumed to be independent of age and very long on average (50 years for cervical intraepithelial neoplasia grade III). This implies that most invasive cancers occurring over age 50 started as cervical intraepithelial neoplasia before age 50, which could thus be detected by screening before age 50. Hence this model is bound to predict only small increases in incidence when women …