Skip to main content
. 2020 Aug 18;10:13959. doi: 10.1038/s41598-020-70613-x

Table 2.

Validation of Xpert MTB/RIF Ultra using culture and/or routine PCR tests as a reference standard.

Sample material Sensitivity Specificity PPV NPV Accuracy
% (95% CI)
TP/(TP + FN)
% (95% CI)
TN/(TN + FP)
%
TP/(TP + FP)
%
TN/(TN + FN)
% (95% CI)
(TP + TN)/(TP + FP + FN + FP)
All samples

90 (56–100)

9/10

99 (93–100)

71/72

90

9/10

99

71/72

98 (91–100)

80/82

Lymph node biopsy

100 (3–100)

1/1

75 (19–99)

3/4

50

1/2

100

3/3

83 (6–100)

4/5

Lymph node aspirate

100 (40–100)

4/4

100 (72–100)

11/11

100

4/4

100

11/11

100 (78–100)

15/15

Pus samples

100 (16–100)

2/2

100 (77–100)

14/14

100

2/2

100

14/14

100 (79–100)

16/16

Other biopsies

100 (3–100)

1/1

100 (75–100)

13/13

100

1/1

100

13/13

100 (77–100)

14/14

Fluid samples

0 (0–98)

0/1

100 (88–100)

28/28

0

N/A

97

28/29

97 (82–100)

28/29

Gastrointestinal lavage

100 (3–100)

1/1

100 (16–100)

2/2

100

1/1

100

2/2

100 (30–100)

3/3

95% confidence intervals for sensitivity, specificity and accuracy were calculated using the exact Clopper–Pearson method.

CI confidence interval, TP true positive, FN false negative, TN true negative, FP false positive.