PURPOSE: The diagnosis of pulmonary tuberculosis (PTB) can be challenging in patients unable to spontaneously produce sputum. Current evidence indicates that induced sputum collection is the least invasive, most cost-effective method of diagnosis and should be utilized prior to bronchoscopy. We reviewed the records of 337 patients with confirmed PTB treated in Northern Alberta between 1997 and 2006 to determine if local practice patterns reflect this evidence. We also collected information on the sensitivity of the diagnostic methods used in our health region.
METHODS: Microbiologic data was collected from the Provincial Laboratory Database and demographic information from the patients’ charts at TB Services. Statistical analysis was done using the two-tailed T-test or chi-squared test as indicated.
RESULTS: 8.5% of patients had induced sputum collected, compared with 35.8% of patients who underwent bronchoscopy. The sensitivity for a culture positive sample was 57.7% for sputum induction, compared to 69.1% for bronchoscopy and 85.3% for spontaneous sputum samples. As expected, increasing the number of samples sent for microbiology increased the yield. Among bronchoscopy patients, bronchial wash samples had a higher sensitivity than bronchoalveolar lavage. Bronchial brush did not improve sensitivity. Post-bronchoscopy sputum was collected in 23.6% of bronchoscopies and had a sensitivity of 80.8%. Only 17% of patients referred for bronchoscopy had 3 or more sputum samples collected. Of these, only 7 (6.3%) were culture positive at bronchoscopy. A similar number (6.4%) had positive sputum cultures when a 4th or 5th sputum sample was sent.
CONCLUSION: Despite supporting evidence induced sputum collection remains underutilized in our health region. Further efforts are needed to improve access to this less invasive test.
CLINICAL IMPLICATIONS: Use of sputum induction in a ”real world“ setting may be less sensitive, less well standardized and more difficult to access than previous research suggests. Bronchoscopy shows few advantages over less invasive methods and should be reserved for select cases.
DISCLOSURE: Sharla-Rae Olsen, No Financial Disclosure Information; No Product/Research Disclosure Information