Table 3 The results of pooled estimates of VAE criteria for the detection of VAP
| Â | Studies (n) | Patients (n) | Estimates (%) | 95 % Confidence interval (%) | Heterogeneity | Publication biase | Effect model | ||
|---|---|---|---|---|---|---|---|---|---|
I 2 (%) | P | t value (Egger’s test) | P | ||||||
Sensitivity | |||||||||
 VAC | 11 | 1633a | 41.8 | 17.7, 65.9 | 99.2 | <0.01 | −3.53 | <0.01 | Random |
 IVAC | 6 | 1323a | 36.3 | 14.4, 58.3 | 98.4 | <0.01 | −1.74 | 0.16 | Random |
 Possible VAP | 2 | 248a | 14.4 | 10.1, 18.8 | 70.6 | 0.07 | NA | NA | Fixed |
 Probable VAP | 2 | 248a | 1.6 | 0.1, 3.2 | 51.9 | 0.15 | NA | NA | Fixed |
Specificity | |||||||||
 VAC | 9 | 23112b | 84.5 | 76.6, 92.4 | 99.8 | <0.01 | −1.50 | 0.18 | Random |
 IVAC | 6 | 14459b | 94.0 | 91.4, 96.7 | 97.7 | <0.01 | −0.99 | 0.38 | Random |
 Possible VAP | 1 | 8325b | 97.3 | 96.9, 97.6 | NA | NA | NA | NA | NA |
 Probable VAP | 1 | 8325b | 99.3 | 99.1, 99.5 | NA | NA | NA | NA | NA |
Positive predictive value | |||||||||
 VAC | 9 | 3572c | 23.2 | 12.6, 33.9 | 99.1 | <0.01 | 1.25 | 0.25 | Random |
 IVAC | 5 | 1563c | 47.2 | 16.1, 78.3 | 99.5 | <0.01 | 0.63 | 0.57 | Random |
 Possible VAP | 1 | 243c | 7.4 | 4.1, 10.7 | NA | NA | NA | NA | NA |
 Probable VAP | 1 | 62c | 6.5 | 0.3, 12.6 | NA | NA | NA | NA | NA |
Negative predictive value | |||||||||
 VAC | 9 | 20927d | 95.6 | 94.7, 96.5 | 98.2 | <0.01 | −7.30 | <0.01 | Random |
 IVAC | 6 | 14211d | 93.3 | 91.5, 95.0 | 99.1 | <0.01 | −3.89 | 0.02 | Random |
 Possible VAP | 1 | 8165d | 99.2 | 99.0, 99.4 | NA | NA | NA | NA | NA |
 Probable VAP | 1 | 8346d | 99.1 | 98.8, 99.3 | NA | NA | NA | NA | NA |