Sensitivity and specificity of tuberculosis signs and symptoms screening and adjunct role of social pathology characteristics in predicting bacteriologically confirmed tuberculosis in Myanmar

Table 6 Overall performance, calibration, and discrimination ability of predicted propensity score in various TB screening models to predict bacteriologically confirmed TB

Model performance measures	Various TB screening models to predict bacteriologically confirmed TB
	Combined models		Social pathology model	TB-SS models		Validated data of social pathology model (Bootstrap resampling = 500)
	A	B	C	D	E	C
Overall performance
Nagelkerke R²	12.9%	11.8%	11.5%	2.7%	0.6%	11.9% (95% CI 9.8–13.8%)
Brier score	0.0059	0.0059	0.006	0.006	0.006	0.0058 (95% CI 0.005–0.006)
Calibration
Hosmer–Lemeshow test	0.005	0.604	0.787	1	1	0.502
Slope	1	1	1	1	1	1
Intercept	0	0	0	0	0	0
Eavg	0.001	0.000	0.000	0.000	0.000	0.0006 (95% CI 0.0003–0.001)
Emax	0.190	0.138	0.093	0.000	0.000	0.126 (95% CI 0.03–0.295)
Discrimination
AUC	81.7	80.7	80.5	63.7	55.2	80.8 (95% CI 78.3–83.0)
DeLong p value	0.0136	0.3606	Reference	0.0001	0.0001
Discrimination slope	0.018	0.015	0.015	0.002	0.001	0.012 (95% CI 0.004–0.018)

Combined model A includes significant social pathology characteristics and any TB-SS variable. Combined model B includes significant social pathology characteristics and any cough variable. Model C includes significant social pathology characteristics. TB-SS model D includes any TB-SS variable. TB-SS model E include any cough variable
AUC area under the curve, Eavg average absolute error, Emax maximal absolute error, 95% CI 95% confidence interval by bootstrap resampling = 500

ISSN: 1349-4147