A variety of methods have been proposed to examine the accuracy of diagnostic (or prognostic) tests when data are presented in a 2 × 2 format (see Table 1). Rule-in (true positive, or TP) and rule-out (true negative, or TN) accuracy are usually best considered separately, but there are circumstances in which a summary of the overall performance is needed. There are four common methods to calculate overall performance. These are the Youden index, the predictive summary index (PSI), the receiver operator curve (ROC), and the overall accuracy. The ROC, which can be considered a special case where overall performance of a test across a range of cut-off scores is needed, is not discussed in this entry. Each of these major methods has an associated ...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles