Review of methodological workflow, interpretation and limitations of nomogram application in cancer study

Abstract

Nomograms have been increasingly adopted in cancer studies for predicting a variety of clinical outcomes. Compared to conventional prediction approaches, nomograms present an excellent graphical easy-to-interpret visualization and by combining patients’ characteristics and statistical models constructed based on diagnostic tests, higher accuracy could be achieved, facilitating the clinical decision-making process. However, with the increasing number of nomograms being proposed, misinterpretations may often occur, introducing a large degree of uncertainties in clinical situations. This review article aims to provide a methodological summary for nomogram applications, construction, and interpretation, with a demonstrative example of nasopharyngeal carcinoma patient data in the field of radiation medicine. Common misunderstandings and limitations are also discussed to underscore the applicability of nomograms in different scenarios. The overarching goal of this review article is to provide general audiences with fundamental and essential knowledge toward comprehending and implementing nomograms in cancer studies.

Keywords

• Nomogram
• Cancer
• Decision making
• Receiver operating curve
• Prognosis