Predicting the microbial cause of community-acquired pneumonia: can physicians or a data-driven method differentiate viral from bacterial pneumonia at patient presentation?

Table 3 Diagnostic prediction performances

	Data-driven approach predictions			Clinician predictions
	Algorithm built from clinical data	Algorithm built from biological and radiological data	Algorithm built from all data sources
Sensitivity	0.75	0.54	0.57	0.86
Specificity	0.71	0.86	0.91	0.54
PPV	0.43	0.86	0.80	0.54
NPV	0.91	0.54	0.77	0.86
Accuracy	0.72	0.67	0.78
AUC	0.72	0.81	0.84
LR + bacterial pneumonia	2.62	3.82	6.29	1.89
LR + viral pneumonia	2.86	1.89	2.12	3.81
LR - bacterial pneumonia	0.35	0.53	0.47	0.26
LR - viral pneumonia	0.38	0.26	0.16	0.53

PPV Positive predictive value, NPV Negative predictive value, AUC Area under the curve, LR Likelihood ratio

ISSN: 1471-2466