Methods: We performed a retrospective cohort study of all inpatients diagnosed with active TB at Yale-New Haven Hospital from 2005 – 2014. Medical records were reviewed for demographics, clinical data, and management. Chi-square and nonparametric testing were used to perform univariate analyses.
Results: 82 patients were included in our study. Median age was 36 years (IQR 25-52), 47 (57.3%) were male; the majority were foreign-born (68.3%) or had traveled outside of the US (13.4%). A substantial proportion (34.1%) were immunocompromised (HIV, diabetes mellitus, malignancy, immunosuppressive medications, cirrhosis). 64.6% had pulmonary TB (PTB) only, 17.1% had extrapulmonary TB (ETB) only, and 18.3% had both PTB and ETB. While 87.9% had pansensitive TB, 12.1% were found to have drug resistance on susceptibility testing. 18.9% had negative PPD, and nearly half (48.8%) were acid-fast bacilli smear negative. 41.5% received empiric anti-TB treatment without laboratory confirmation of TB. Of those with PTB, immunosuppressed patients were more likely to present with normal chest radiograph (p = 0.02).
Conclusion: At our institution, the majority of TB cases occurred in young and foreign-born individuals, and many were immunocompromised. A substantial proportion of our study group had negative PPD, normal chest radiographs, and smear negative samples and anti-TB therapy was often initiated based on clinical suspicion. Our review suggests that when performing an evaluation for TB, microbiologic evaluation and drug susceptibility testing is crucial, and that chest computed tomography may be more beneficial for evaluation of TB in immunocompromised patients.
M. Villanueva, None
S. Shenoi, None