Table 3.

Statistical assessment of different classifiers [95% CI] at the respective cut-off-values for the XGBoost models

training set	validation parameter	validation set					cut off from training set
		pre-pandemic	1st wave	after 1st wave	2nd wave	after 2nd wave
		06/2014–03/2020	04/2020–05/2020	06/2020–09/2020	10/2020–05/2021	06/2021–10/2021
pre-pandemic (model 1)	Sensitivity	0.722 [0.664–0.774]	0.773 [0.546–0.922]	0.672 [0.540–0.787]	0.760 [0.675–0.832]	0.677 [0.549–0.788]	0.1611
	Specificity	0.927 [0.924–0.930]	0.894 [0.880–0.906]	0.932 [0.925–0.938]	0.921 [0.916–0.926]	0.940 [0.935–0.946]
	PPV	0.086 [0.075–0.098]	0.067 [0.039–0.105]	0.085 [0.062–0.114]	0.091 [0.074–0.110]	0.085 [0.062–0.112]
	NPV	0.997 [0.996–0.998]	0.997 [0.994–0.999]	0.997 [0.995–0.998]	0.997 [0.996–0.998]	0.997 [0.996–0.998]
	F1 Score	0.153 [0.135–0.1723]	0.123 [0.07–0.173]	0.151 [0.107–0.192]	0.163 [0.134–0.191]	0.151 [0.109–0.190]
pre-pandemic + 1st wave (model 2)	Sensitivity	0.730 [0.660–0.792]		0.462 [0.192–0.749]	0.753 [0.685–0.812]	0.662 [0.534–0.774]	0.1805
	Specificity	0.915 [0.912–0.919]		0.927 [0.911–0.941]	0.920 [0.916–0.923]	0.933 [0.927–0.938]
	PPV	0.065 [0.055–0.076]		0.061 [0.023–0.129]	0.082 [0.069–0.096]	0.074 [0.054–0.099]
	NPV	0.998 [0.997–0.998]		0.994 [0.988–0.998]	0.997 [0.997–0.998]	0.997 [0.996–0.998]
	F1 Score	0.119 [0.101–0.138]		0.108 [0.027–0.180]	0.148 [0.128–0.179]	0.134 [0.095–0.166]
whole set (model 3)	Sensitivity	0.258 [0.204–0.319]					0.1296
	Specificity	0.992 [0.991–0.993]
	PPV	0.221 [0.173–0.275]
	NPV	0.994 [0.992–0.994]
	F1 Score	0.238 [0.185–0.288]