Table-III.

Performance metrics of different nomogram predictive models.

Model	Sensitivity	Specificity	Accuracy	F1	AUC (95%CI)
Combined model	0.933	0.760	0.783	0.528	0.891(0.818–0.963)
PCT	0.600	0.910	0.870	0.545	0.776(0.628–0.924)
BNP	0.800	0.670	0.687	0.400	0.736(0.600–0.873)
Impaired consciousness	0.867	0.440	0.496	0.310	0.653(0.552–0.755)