In order to provide effective health care to patients, clinicians must rely on their ability to accurately diagnose disease and to prognosticate the outcomes. Prognostic studies have received considerable attention in health science and medicine in relation to patient outcomes. However, little effort has been spent on evaluating prognostic accuracy. The purpose of this paper is to present a comprehensive review of the methods for assessing prognostic accuracy in patient outcome prediction. The strengths and limitations of these approaches are critically appraised. We argue that we need to consider incorporating accuracy assessment for predicting patient outcomes both in clinical practice and in research.