Methods: Retrospective cohort analysis of patients with culture confirmed PTB at Grady Memorial Hospital in Atlanta, GA from 2008-2015. McNemar’s test was used to compare the sensitivity of CT and CXR. Cox proportional hazards models were fit to test the association between cavitary CT and time to sputum culture conversion.
Results: Among 247 PTB patients, 72% (178/247) had a CT performed a mean of two days within hospital admission. Patients with CT had similar demographics, comorbidities (including HIV and diabetes), CXR findings, and initial AFB sputum smear grade compared to those without CT. Forty-seven percent (82/178) of patients with CT had one or more cavitary lesions. Among patients with non-cavitary CXR, 35% (48/139) had a cavitary lesion on CT. In subgroup analyses of non-cavitary CXR patients, 17% (10/60) of HIV-infected adults, and 11% (5/46) of participants with negative sputum smears had cavitary lesions on CT. The percent of patients achieving culture conversion at 28 and 56 days was 59% and 87% for patients with non-cavitary CT, and 30% and 74% among those with cavitary CT. Cavitary CT was independently associated with a longer time to TB culture conversion after covariate adjustment. The unadjusted and adjusted hazard ratios (HR) for cavitary compared to non-cavitary CT were 0.62 (95%CI 0.45-0.85) and 0.65 (95%CI 0.43-0.99), respectively.
Conclusion: Chest CT was commonly performed among patients with PTB and had better ability to detect cavitary disease than CXR including in pauci-bacillary states. Chest CT may have important clinical utility in that the presence of a cavitary lesion can identify patients who may require more intensive management. Our data support possible use of CT to determine both severity of disease and prediction of culture conversion. Efforts are currently ongoing to further characterize the specific CT findings that predict prolonged time to culture conversion.
A. Hernandez, None