Derivation cohort (n=8288)* | Validation cohort (n=2029) | |
Nagelkerke R2 | 0.11 | 0.12 |
Brier score | 0.04 | 0.05 |
c-statistic (95% CI) | 0.71 (0.68 to 0.73) | 0.70 (0.64 to 0.76) |
Calibration-in-the-large | NA | 0.05 |
Calibration slope | NA | 1.01 |
Nagelkerke R2 can range from 0 to 1, with 0 denoting that model does not explain any variation and 1 denoting that it perfectly explains the observed variation. The Brier score can range from 0 for a perfect model to 0.25 for a non-informative model with a 50% incidence of the outcome. c-statistic can range from 0.50 for a non-discriminative model to 1 for a perfect model with ≥0.70 as a reasonable discriminative ability. The calibration-in-the-large indicates whether predicted probabilities are systematically too low (value >0) or too high (value <0). The calibration slope indicates whether the model is overfitted (estimated risks too extreme, value <1) or underfitted (estimated risks too close to baseline risk, value >1). A model is perfectly calibrated if the calibration-in-the-large is 0 and calibration slope is 1.
*There were missing data in three cases.
NA, not applicable.