TABLE 3

Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and F-1 score based on selected risk thresholds for specialist and event cohorts, and for utilisation-based and diagnosis-based models

Utilisation-based model, specialist cohort, risk thresholdDiagnosis-based model, specialist cohort, risk thresholdUtilisation-based model, event cohort, risk thresholdDiagnosis-based model, event cohort, risk threshold
0.20.30.40.50.60.20.30.40.50.60.10.20.30.40.50.10.20.30.40.5
n#9627094542511481016768517300108343813975702841014926142755512229
Sensitivity %0.9600.8200.6070.3710.2310.9690.8200.6460.4370.1630.7290.4600.2670.1580.0650.8120.3710.1980.0560.014
Specificity %0.2190.5000.7350.8780.9360.1440.4120.6680.8500.9490.6280.8690.9560.9810.9940.4390.8540.9500.9900.998
PPV %0.4540.5260.6080.6730.7090.4340.4860.5690.6630.6850.1940.3010.4280.5070.5840.1500.2380.3260.4180.448
NPV %0.8910.8040.7340.6740.6420.8740.7720.7360.6900.6260.9500.9290.9140.9050.8970.9500.9170.9060.8950.892
F-1 score0.6170.6410.6070.4790.3480.6000.6100.6050.5270.2630.3060.3640.3290.2410.1160.2540.2900.2470.0990.028

Assessment was based on testing dataset only. #: number of eligible patients whose risk was above each risk threshold.