J Epidemiol Community Health 39:364-366 doi:10.1136/jech.39.4.364
  • Research Article

Estimating the sensitivity of a screening test.


A commonly used estimate of a screening test's sensitivity, based on the prevalence at screening and the incidence of interval cancers, is shown to be logically unsound and quantitatively poor. An alternative simple estimate is proposed, based only on incidence rates.