Skip to main content

Advertisement

Table 4 Performance measures for the expert-based and statistically derived prognostic models

From: Prediction of persistent shoulder pain in general practice: Comparing clinical consensus from a Delphi procedure with a statistical scoring system

  Expert model dicho Expert model cont Statistical model
calibration slope 1.017 0.986 1.019
R2 N 0.098 0.13 0.162
AUC 0.656 0.679 0.702
95% CI 0.612 - 0.700 0.636 - 0.722 0.660 - 0.745
Optimism 0.029 0.028 0.014
AUCcorrected 0.627 0.651 0.688
  1. Expert model dicho = prognostic model using dichotomized expert selected predictors
  2. Expert model cont = prognostic model using continuous expert selected predictors
  3. R2 N = Nagelkerke's explained variance
  4. AUC = Discriminative ability as indicated by the Area Under the ROC Curve
  5. 95% CI = 95% confidence interval for the AUC
  6. Optimism = the models' estimated deterioration when applied to new subjects
  7. AUCcorrected = the optimism corrected Area Under the ROC Curve