RESUMO
SETTING: Five health care systems in Texas. OBJECTIVE: To describe the epidemiology of inadequate isolation for pulmonary tuberculosis leading to tuberculosis (TB) exposures from confirmed TB patients and the patient factors that led to the exposures. DESIGN: A retrospective cohort and case-control study of adult patients with TB resulting in exposures (cases) vs those TB patients who did not result in exposures (controls) during January 2005 to December 2012. RESULTS: There were 335 patients with pulmonary TB disease, 199 cases and 136 controls. There was no difference between groups in age (46 ± 14.6 vs 45 ± 17 years; P > .05), race, or substance abuse. Cases were more likely to be transplant recipients (adjusted odds ratio [AOR], 18.90; 95% CI, 1.9-187.76), have typical TB chest radiograph (AOR, 2.23; 95% CI, 1.1-4.51), and have positive acid-fast bacilli stains (AOR, 2.36; 95% CI, 1.31-4.27). Cases were less likely to have extrapulmonary disease (AOR, 0.47; 95% CI, 0.24-0.95). CONCLUSIONS: TB exposure resulting from inadequate isolation is frequent in health care settings. Extrapulmonary involvement resulted in earlier airborne isolation. Being a transplant recipient, having chest radiograph findings typical for TB, and sputum positivity acid-fast bacilli upon staining were associated with increased risk of inadequate isolation.