Table 3.
Characteristic | N | CXR consistent with tuberculosis | TST induration ≥ 10 mm | Positive QFT-G | |||
---|---|---|---|---|---|---|---|
n | aOR* (95% CI) | n | aOR* (95% CI) | n | aOR* (95% CI) | ||
Total | 1,246 | 272 | 721 | 352 | |||
Age group | |||||||
18–20 years | 139 | 9 (6.5%) | 1.0 | 49 (35.3%) | 1.0 | 25 (18.0%) | 1.0 |
21–30 years | 293 | 19 (6.5%) | 1.01 (0.44–2.32) | 143 (48.8%) | 2.17 (1.40–3.35) | 77 (26.3%) | 1.66 (1.002–2.759) |
31–40 years | 294 | 36 (12.2%) | 2.06 (0.96–4.42) | 199 (67.7%) | 4.14 (2.67–6.43) | 89 (30.3%) | 1.97 (1.19–3.25) |
41–50 years | 268 | 80 (29.9%) | 5.87 (2.84–12.15) | 171 (63.8%) | 3.49 (2.24–5.46) | 91 (34.0%) | 2.29 (1.39–3.79) |
51–64 years | 207 | 99 (47.8%) | 12.94 (6.23–26.89) | 137 (66.2%) | 4.22 (2.61–6.81) | 60 (29.0%) | 1.81 (1.07–3.07) |
≥65 years | 45 | 24 (64.4%) | 25.12 (10.06–62.70) | 22 (48.9%) | 1.92 (0.95–3.92) | 10 (22.2%) | 1.26 (0.55–2.89) |
Sex | |||||||
Female | 842 | 148 (17.6%) | 1.0 | 439 (52.1%) | 1.0 | 226 (26.8%) | |
Male | 404 | 124 (30.7%) | 1.69 (1.24–2.29) | 282 (69.8%) | 2.09 (1.60–2.73) | 126 (31.2%) | N.S. and NIM |
BCG status | |||||||
Not vaccinated | 735 | 191 (26.0%) | 398 (54.1%) | 1.0 | 215 (29.3%) | ||
Vaccinated | 511 | 81 (58.8%) | N.S. and NIM | 323 (63.2%) | 1.62 (1.25–2.09) | 137 (26.8%) | |
Chest exam | |||||||
Normal | 1239 | 266 (21.5%) | 1.0 | 718 (57.9%) | 351 (28.3%) | ||
Abnormal | 7 | 6 (85.7%) | 31.23 (3.04–320.68) | 3 (42.9%) | N.S. and NIM | 1 (14.3%) | N.S. and NIM |
M. avium reactivity | |||||||
No | 884 | 217 (24.5%) | 550 (62.2%) | 1.0 | 268 (30.3%) | 1.0 | |
Yes | 362 | 55 (15.2%) | N.S. and NIM | 171 (47.2%) | 0.59 (0.46–0.77) | 84 (23.2%) | 0.71 (0.53–0.95) |
CXR: chest radiographs; TST: tuberculin skin test; QFT-G: QuantiFERON-TB Gold test; n: number in subset of N; aOR: adjusted odd ratio; 95% CI: 95% confidence interval; N.S.: not significant; NIM: not in model; BCG: bacille Calette-Guérin; TB: tuberculosis; QFT: QuantiFERON-TB test.
*Multivariate models were created using factors with P values ≤0.2 in univariate analysis and <0.05 in stepwise logistic regression until the best fitting, parsimonious model was identified. Model fit was evaluated using the Hosmer-Lemeshow test. Negative and indeterminate QFT-G results were coded as “not positive.” The variable “Prior TB” was not included in the model because all 7 subjects with prior TB had CXR findings consistent with TB disease and its inclusion prevented convergence of the model. Bold font indicates statistically significant adjusted odds ratios (aORs).