Abstract

How to assess and interpret a prognostic study Prognosis is the prediction of the evolution of a given disease or event of interest and the frequency of occurrence of specific outcomes. The ideal study design to assess prognosis are cohort studies. However many prognostic studies have case-control or case series designs. Prognostic studies should include all participants with the event of interest of a determinate geographical region, to avoid bias. The measurement precision and exactitude of intervening variables and outcomes, the identification of confounding variables and an adequate follow up period, have a great influence on the methodological quality of prognostic studies. Prognosis is usually expressed as survival, disease free or remission free rates. This article provides basic methodological concepts that should be considered when evaluating a prognostic study.