Sputum microbiology is essential for pulmonary TB diagnosis. Detecting TB organism, M. tuberculosis, in a stained sputum smear/GeneXpert assay provides evidence of the disease. Sputum cultures are then set up. Our observation lately is that bronchoscopy is done to obtain a diagnostic specimen which would be unnecessary if sputum examination was done first, as routine. Compared to natural sputum examination, bronchoscopy is more expensive and invasive, and might lead to serious complications, especially in the elderly, besides the potential for TB exposure during the procedure. The aim of this evaluation was to quantify the extent of bronchoscopy use in TB diagnosis setting in Arkansas.
We reviewed pulmonary TB case records available at the TB Control Program (n = 383) diagnosed in the period between January 1st, 2009 and December 31st, 2015. Cases that had bronchoscopy were identified. Timing of sputum sample collection in relation to undergoing bronchoscopy was examined. A logistic regression model was used to predict the likelihood of having bronchoscopy in diagnostic work-up.
Among cases reviewed, 107 [27.9%, 95% CI (23.6%, 32.8%)] had bronchoscopy. Of those, 67 [62.6%, 95% CI (52.7%, 71.6%)] had the procedure before a sputum sample was collected, 21 [19.6%, 95% CI (12.8%, 28.7%)] had it after a sputum sample was collected, 4 [3.8%, 95% CI (1.2%, 9.9%)] had both done on the same day, and 15 [14.0%, 95% CI (8.3%, 22.4%)] had no sputum sample collected. In cases with bronchoscopy, sputum smear and culture were positive for M. tuberculosis in 39 [36.5%, 95% CI (27.5%, 46.4%)], smear negative but culture positive in 30 [28.0%, 95% CI (20.0%, 37.7%)], smear positive but culture negative in 2 [1.9%, 95% CI (0.3%, 7.2%)], and both negative in 21 [19.6%, 95% CI (12.8%, 28.7%)]. Of the 15 who did not have sputum sample collected, at least 10 [66.7%, 95% CI (38.7%, 87.0%)] were coughing. In a multivariable model, older age, female gender, certain public health regions, and negative sputum smear result were significant predictors of the likelihood of having bronchoscopy.
If bronchoscopy was solely indicated to diagnose pulmonary TB, at least one-third of the procedures performed were avoidable; as sputum examination identified those cases. Thus, a precise clinical judgment should govern the decision of subjecting a patient to bronchoscopy for the sole purpose of diagnosing pulmonary TB.
J. Selig, None
J. Bates, None
L. Mukasa, None