Score | Discriminatory power (AUROC) | Calibration (Hosmer-Lemeshow χ2 test p value) | ||||
---|---|---|---|---|---|---|
Original model | Recalculated model | Original model | Recalculated model | |||
Development | Validation | Development | Validation | |||
Prytherch score6 | 0.842 (0.818–0.865) | 0.858 (0.827–0.889) | 0.874 (0.841–0.907) | <0.001 | 0.59 | 0.66 |
Froom score7 | 0.862 (0.813–0.910) | 0.930 (0.897–0.962) | 0.882 (0.806–0.957) | − | 0.93 | 0.009 |
Loekito score8 | 0.922 (0.879–0.965) | 0.911 (0.819–1.000) | 0.917 (0.823–1.000) | 0.0007 | 0.79 | 1.00 |
Asadollahi score11 | 0.803 (0.776–0.829) | 0.808 (0.774–0.842) | 0.813 (0.772–0.854) | − | 0.79 | 0.47 |
Area under receiver-operating curve (AUROC) above 0.8 represents good discriminatory power, and p value for calibration above 0.05 represents good calibration