Skip to main content
. 2013 Jan 18;8(1):e54235. doi: 10.1371/journal.pone.0054235

Table 4. Performance characteristics of the score at different cut-offs in 473 patients (n = 138 with PTB and 335 without PTB).

Cut-off Sensitivity Specificity PPV NPV Area under the ROC curve**
≥1 89.1(82.7,93.8) 58.2(52.7,63.5) 46.8(40.6,53) 92.9(88.5,95.9) 0.74(0.7,0.77)
≥2 85.5(78.5,90.9) 63.9(58.5,69) 49.4(42.9,55.9) 91.5(87.1,94.7) 0.75(0.71,0.79)
≥3 57.2(48.5,65.6) 86.9(82.8,90.3) 64.2(55.1,72.7) 83.1(78.8,86.9) 0.72(0.68,0.77)
≥4 13.8(8.5,20.7) 98.5(96.6,99.5) 79.2(57.8,92.9) 73.5(69.2,77.5) 0.56(0.53,0.59)
“Consistent with active TB” reported by readers * 93.4(87.9,97) 27.5(22.6,32.8) 36.1(31.1,41.3) 90.5(82.8,95.6) 0.6(0.57,0.64)

Calculated Score = (UL large opacity*2)+(cavity, any location*2)+(unilateral pleural effusion*1)+(adenopathy, any location*2).

*

This judgment (consistent or not consistent with active TB) was the interpretation of the reader recorded on completion of the CRRS read. in the appropriate block on the CRRS form. (see online repository).

**

The various AUROCs were derived from a dichotomous test at each cut-point.