Table 3.

C-index of trained models for test dataset predictions

	Before feature selection	After feature selection	Excluding cholesterol measurements + substituting BP
Number of features	608	47	45
CPH model C-index	0.7431 [0.7422–0.7441]	0.7443 [0.7441–0.7445]	0.7409 [0.7407–0.7411]
DeepSurv model C-index	0.7461 [0.7452–0.7469]	0.7446 [0.7441–0.7452]	0.7388 [0.7382–0.7393]

A 95% confidence interval is shown in square brackets. Cholesterol measurements include total cholesterol and cholesterol ratio, systolic blood pressure was substituted with heart rate.

BP, blood pressure.