From Normal Distribution to Diagnostic Test

Authors

  • Anteo Di Napoli Comitato Tecnico-Scientifico RIDT, Roma
  • Michele Nichelatti Servizio di Biostatistica, Dipartimento di Ematologia e Oncologia, Ospedale Niguarda, Milano
  • Francesco Franco Regione Lazio, Roma

DOI:

https://doi.org/10.33393/gcnd.2015.840

Keywords:

Normal distribution, Sensibility, Specificity, ROC curves

Abstract

Distribution curves of laboratory parameters in healthy and sick subjects are quite similar, leading to a “grey zone” referred to subjects that could be not correctly identified and classified. To identify a cut-off value, ROC (Receiver Operating Characteristics) curves can be applied; they represent the graphic relationship between sensibility (true positive) and specificity (false positive), depending on cut-off values variability. ROC curve subtends AUC (Area Under the Curve); AUC's width is directly related to discerning power of any validated statistical test. (Epidemiology_statistics)

Most read articles by the same author(s)

1 2 3 > >>