Skip to main content

Table 3 Misclassification table

From: Adapted time-varying covariates Cox model for predicting future cirrhosis development performs well in a large hepatitis C cohort

Follow-up interval, laboratory window (years) Test sample (n)* Event proportion AUROC Brier score Best cut-off Specificity Sensitivity PPV NPV
1, 2 34,081 0.030 0.815 0.028 0.057 0.737 0.749 0.080 0.990
3, 2 30,117 0.161 0.811 0.133 0.059 0.767 0.719 0.373 0.934
5, 2 28,604 0.329 0.794 0.274 0.056 0.750 0.703 0.579 0.838
1, 4 25,934 0.026 0.796 0.031 0.105 0.731 0.712 0.066 0.990
3, 4 20,765 0.083 0.781 0.069 0.104 0.732 0.703 0.191 0.965
5, 4 18,164 0.230 0.792 0.172 0.105 0.723 0.721 0.438 0.897
  1. *Sample size is based on the selected representative split