Procalcitonin for the diagnosis of postoperative bacterial infection after adult cardiac surgery: a systematic review and meta-analysis

Table 2 Sensitivities and specificities at predefined thresholds and corresponding PPVs and NPVs for different prevalences, based on the multiple thresholds model

Threshold	Sensitivity	95%CI	Specificity	95%CI	Prevalence	PPV	NPV	FP*	FN*
1	0.7	0.47–0.85	0.7	0.58–0.8	0.05	0.11	0.98	29	1
					0.1	0.2	0.95	27	3
					0.15	0.29	0.93	26	4
					0.2	0.37	0.9	24	6
					0.25	0.43	0.87	23	7
					0.3	0.5	0.84	21	9
2	0.68	0.48–0.83	0.72	0.63–0.79	0.05	0.11	0.98	27	2
					0.1	0.21	0.95	25	3
					0.15	0.3	0.93	24	5
					0.2	0.37	0.9	22	6
					0.25	0.44	0.87	21	8
					0.3	0.51	0.84	20	10
3	0.67	0.47–0.82	0.73	0.65–0.79	0.05	0.11	0.98	26	2
					0.1	0.21	0.95	24	3
					0.15	0.3	0.93	23	5
					0.2	0.38	0.9	22	7
					0.25	0.45	0.87	20	8
					0.3	0.51	0.84	19	10
4	0.66	0.46–0.82	0.73	0.66–0.8	0.05	0.12	0.98	26	2
					0.1	0.22	0.95	24	3
					0.15	0.31	0.93	23	5
					0.2	0.38	0.9	22	7
					0.25	0.45	0.87	20	8
					0.3	0.52	0.84	19	10
5	0.66	0.45–0.82	0.74	0.67–0.8	0.05	0.12	0.98	25	2
					0.1	0.22	0.95	23	3
					0.15	0.31	0.92	22	5
					0.2	0.39	0.9	21	7
					0.25	0.46	0.87	20	8
					0.3	0.52	0.83	18	10

FN false negative, FP false positive, NPV negative predictive value, PPV positive predictive value
*Number of false positives and negatives in 100 hypothetical cases

ISSN: 1364-8535