How has the group of patients used in the analysis been selected, and in particular how has the decision to verify the test by the gold standard been made? A common error is to select patients in some manner for verifi cation of a previous diagnosis; this usually leads to positive tests being overrepresented in the verifi ed sample and the sensitivity being infl ated. It is also common for investigation to assume that unverifi ed cases are disease-

free, which can lead to infl ated specifi city estimates. The best way to avoid such bias is to construct a prospective study in which all patients receive defi nite verifi cation of disease status.