Skip to main content
. 2019 Jan 4;14(1):e0208141. doi: 10.1371/journal.pone.0208141

Table 4. Misclassification Table.

Time Test Sample Size Event Proportion Model AuROC Best cut-off Specificity Sensitivity PPV NPV
1 year 18896 0.036 CS Cox 0.807 0.041 0.79 0.71 0.11 0.99
CS Boosting 0.817 0.037 0.77 0.73 0.11 0.99
LGT Cox 0.828 0.037 0.75 0.76 0.10 0.99
LGT Boosting 0.838 0.035 0.76 0.77 0.11 0.99
3 years 14605 0.112 CS Cox 0.784 0.095 0.73 0.72 0.25 0.95
CS Boosting 0.799 0.091 0.76 0.71 0.27 0.95
LGT Cox 0.804 0.095 0.75 0.74 0.27 0.96
LGT Boosting 0.815 0.090 0.76 0.73 0.28 0.96
5 years 11334 0.206 CS Cox 0.775 0.151 0.74 0.70 0.41 0.90
CS Boosting 0.790 0.138 0.75 0.70 0.42 0.91
LGT Cox 0.794 0.151 0.75 0.71 0.42 0.91
LGT Boosting 0.805 0.128 0.73 0.74 0.41 0.92

(CS) cross-sectional; (LGT) longitudinal; (PPV) positive predictive value; (NPV) negative predictive value.