In-house nucleic acid amplification tests for the detection of Mycobacterium tuberculosis in sputum specimens: meta-analysis and meta-regression

Table 3 Meta-regression analysis to determine sources of heterogeneity

Covariate	Coefficient	P-value	Relative diagnostic odds ratio (RDOR)	95% confidence interval
Intercept	3.248	0.0052	----	----
Threshold (S)	- 0.072	0.4522	----	----
Case control vs. cross-sectional design	- 0.292	0.5219	0.75	(0.30 – 1.85)
Blinded vs. unblinded studies	0.441	0.2795	1.55	(0.69 – 3.48)
Chemical vs. other DNA extraction methods	- 0.209	0.5877	0.81	(0.38 – 1.74)
*IS6110* vs. other target sequences	1.055	0.0074	2.87	(1.34 – 6.16)
Nested vs. regular amplification protocols	1.196	0.0135	3.31	(1.29 – 8.49)
Probe vs. gel-UV detection methods	0.157	0.7222	1.17	(0.49 – 2.81)
Sputum smear positive/both vs negative	0.242	0.8146	1.27	(0.16 – 9.88)

Intercept: constant term in the model
S: indicator of threshold (logit TPR+logit FPR); TPR: true positive rate; FPR: false positive rate RDOR: relative diagnostic odds ratio (obtained by exponentiating the model coefficients)

ISSN: 1471-2180