Abstract
Background: CXR plus sputum microscopy & culture (TBC) for M.tuberculosis (MTB) can identify co-existing pulmonary TB (PTB) in subjects presenting with extra-pulmonary tuberculosis (EPTB). In practice, EPTB subjects with normal CXR may not have sputum sampling performed, leading to underestimation of pulmonary disease.
Objectives: To determine the use of CXR & TBC in patients with EPTB & their diagnostic utility to detect co-existent PTB.
Study design: Retrospective clinical and demographic data for TB patients at our hospital (1.1.06 - 31.12.08) obtained from the London TB Register were linked to hospital microbiology & HIV test data. Baseline CXR (B-CXR) were scored by 2 respiratory radiologists blind to patient diagnosis.
Results: Of 308 cases (median age 39y, 52.9% female), 155 were notified as EPTB only. 143 (92.3%) had B-CXR, with abnormal (abN) features identified in 67 patients (46.9%): 48 (33.6%) consistent with possible PTB & 9 (6.3%) probable PTB, independent of HIV status. Sputum samples were obtained from 54 patients (37.8%). This was less likely in those with a normal B-CXR (21.4% vs. 56.2% abN, p<0.001). TBC was MTB positive in 9 patients (16.7%): 7 with abN B-CXR (5 possible PTB, 2 probable PTB) & 2 with normal B-CXR.
Conclusion: Most EPTB patients had a B-CXR, 40% of which were consistent with possible/probable active PTB. Sputum samples were infrequently obtained, though when performed MTB yield was high. Clinical diagnosis of EPTB using B-CXR review alone may underestimate co-existing PTB. To identify infectious cases & improve TB control, sputum collection & TBC should be performed in all patients.
- © 2011 ERS
