Skip to main content
. 2019 Mar 26;14(3):e0213728. doi: 10.1371/journal.pone.0213728

Table 2. Diagnostic accuracy estimates based on different thresholds of pleural fluid adenosine deaminase.

Adenosine deaminase threshold range <36 IU/L 40±4 IU/L 45–65 IU/L >65 IU/L Overall
Number of studies 49 65 56 4 174
Median prevalence of tuberculosis in studies 0.39 0.46 0.41 0.44 0.41
Summary sensitivity 0.91
(0.88–0.94)
0.93
(0.90–0.95)
0.91
(0.88–0.94)
0.86
(0.61–0.96)
0.92
(0.90–0.93)
Summary specificity 0.89
(0.87–0.91)
0.90
(0.87–0.92)
0.90
(0.87–0.92)
0.94
(0.80–0.99)
0.90
(0.88–0.91)
Summary diagnostic odds ratio 86.32
(56.42–132.08)
109.51
(71.52–167.67)
91.57
(55.05–152.33)
96.55
(14.75–632.17)
97.42
(74.90–126.72)
Positive likelihood ratio* 8.60 8.88 8.91 14.62 8.92
Negative likelihood ratio* 0.10 0.08 0.10 0.15 0.09
Positive predictive value at median prevalence* 0.85 0.88 0.86 0.92 0.86
Negative predictive value at median prevalence* 0.94 0.93 0.94 0.89 0.94
Positive predictive value at 20% prevalence* 0.68 0.69 0.69 0.79 0.69
Negative predictive value at 20% prevalence* 0.98 0.98 0.98 0.96 0.98
Positive predictive value at 60% prevalence* 0.93 0.93 0.93 0.96 0.93
Negative predictive value at 60% prevalence* 0.87 0.89 0.87 0.81 0.88

Figures in parentheses are 95% confidence limits

* Values calculated from summary sensitivity and specificity estimates