Comparison of Xpert MTB/RIF, Microscopy and AFB Culture in Bronchial Washings in Sputum Scarce Cases of Suspected Pulmonary Tuberculosis

This cross-sectional study will be conducted in Ojha Institute of Chest Diseases and Department of Medicine, Dow Medical College, Karachi both affiliated with Dow University of Health Sciences, Karachi, Pakistan after IRB approval. Patients were informed of study protocol, benefits and risks and written consent will be taken. All patients of either gender with clinical suspicion of PTB on history, examination and CXR and not expectorating sputum or less than 1 ml will be included in the study. Patients with extra pulmonary tuberculosis, those with age < 12 years will be excluded. Demographic data of selected patients including age, gender, contact with PTB patient will be collected. Data will be entered in study proforma.

Selected patients will undergo bronchoscopy using Olympus 180 series video bronchoscope. BW will be done by instilling 60 ml of normal saline via bronchoscope and 3 aliquots of BW samples will be collected by suctioning. One part of washing will be immediately inoculated in AFB Culture media and second will be submitted for Xpert MTB/RIF testing. Third part will be centrifuged, and sediment will be subjected to AFB microscopy after staining with Ziehl-Neelsen technique.

Results of microscopy and Xpert MTB/RIF will be collected next day, and treatment will be started if MTB comes positive, while results of AFB Culture will be collected after 6 weeks and treatment will be started if not already started on previous reports. Sensitivity/Resistance pattern will be recorded for both Xpert MTB/RIF and ABF Culture.

Sample size was calculated using reported both for smear microscopy and Xpert MTB/RIF. Using the reported sensitivity of Xpert MTB/RIF of 82.4% by Rakotoarivelo R, et al and keeping the power at 95% and alpha at 0.05 the sample size for Xpert MTB/RIF is estimated as 25. For smear microscopy the reported sensitivity of 70.47% by Chowdhury R et al was used and sample size of 72 was calculated using same parameters as those for Xpert MTB/RIF. Higher sample size of 72 was taken for conduct of study. Sample size calculations were done using PASS software version 15.0. Frequency of positive tests will be calculated. Mean age will be calculated according to gender and compared by Student's t-test. Sensitivity & specificity of tests will be calculated using online calculator by VassarStats. Frequency of positive yield of Xpert MTB/RIF will be compared to that of microscopy by χ2- test. P-value of <0.05 was considered significant.