The Added Value of Bronchoalveolar Lavage for Pulmonary Tuberculosis Diagnosis in High-Risk Hospitalized Patients with Negative Sputum Samples

Abstract

Hospitalized patients with a high suspicion of pulmonary tuberculosis (HS-PTB) are isolated until a definite diagnosis can be determined. If doubt remains after negative sputum samples, bronchoscopy with bronchoalveolar lavage (BAL) is often sought. Still, evidence of the added value of BAL in this patient population is scarce. To address this issue, we included consecutive HS-PTB patients with negative sputum samples who underwent BAL between 2017 and 2018. Chest X-rays (CXR) and CT scans were evaluated by a chest radiologist blind to the final diagnosis. Independent predictors for PTB were assessed by multivariate regression, using all positive PTB patients between 2017 and 2019 (by sputum or BAL) as a control group (n = 41). Overall, 42 HS-PTB patients were included (mean age 51 ± 9, 36% female). BAL was a viable diagnostic for PTB in three (7%) cases and for other clinically relevant pathogens in six (14%). Independent predictors for PTB were ≥2 sub-acute symptoms (adjusted OR 3.18, 95% CI 1.04–9.8), CXR upper-lobe consolidation (AOR 8.70, 95% CI 2.5–29), and centrilobular nodules in chest CT (AOR 3.96, 95% CI 1.20–13.0, p = 0.02). In conclusion, bronchoscopy with BAL in hospitalized patients with HS-PTB had a 7% added diagnostic value after negative sputum samples. Our findings highlight specific predictors for PTB diagnosis that could be used in future controlled studies to personalize the diagnostic evaluation.