Table 3. Posterior distributions derived from two Bayesian latent class models (BLCMs) to estimate the sensitivity and specificity of five tests to diagnose pulmonary tuberculosis in 344 captive bears in Cambodia and Vietnam.

Test Parameter	Posterior distributions median (95% Bayesian credible interval)
	Initial BLCM assuming conditional independency between the tests	Final BLCM allowing for dependency between BAL microscopy and BAL PCR
BAL microscopy
Sensitivity	0.56 (0.28, 0.83)	0.49 (0.24, 0.77)
Specificity	0.997 (0.99, 1.00)	0.998 (0.99, 1.00)
BAL PCR
Sensitivity	0.59 (0.31, 0.86)	0.50 (0.25, 0.78)
Specificity	0.99 (0.97, 1.00)	0.99 (0.97, 1.00)
BAL culture
Sensitivity	0.81 (0.51, 0.99)	0.69 (0.38, 0.96)
Specificity	0.996 (0.99, 1.00)	0.997 (0.99, 1.00)
Thoracic radiography
Sensitivity	0.93 (0.75, 0.998)	0.95 (0.76, 0.998)
Specificity	0.94 (0.91, 0.96)	0.95 (0.91, 0.98)
DPP (positive: strong reactions only)
Sensitivity	0.65 (0.38, 0.88)	0.62 (0.36, 0.86)
Specificity	0.95 (0.92, 0.97)	0.96 (0.93, 0.98)
DPP (positive: strong and weak reactions)
Sensitivity	0.82 (0.55, 0.98)	0.80 (0.53, 0.97)
Specificity	0.79 (0.74, 0.83)	0.80 (0.75, 0.84)
Covp	n/a	0.18 (0.08, 0.24)
Covn	n/a	0.001 (0, 0.007)

Prior distribution for all parameters was β (1;1)

BAL: bronchoalveolar lavage

DPP: Dual path platform VetTB serological assay

n/a: not applicable

Covp: covariance between the sensitivity of the tests among TB positive bears

Covn: covariance between the specificity of the tests among TB negative bears