The hidden hypothesis: A disseminated tuberculosis case

•A Bayesian approach to clinical decision making should be recommended.•Test interpretation should account for the test imperfect diagnostic performance.•Further studies to investigate genetic susceptibility to tuberculosis are needed. 77-year-old former smoker admitted because of fatigue and abdominal distention. Past medical history positive for two previous hospitalizations for pericardial and pleural effusions (no diagnosis achieved). At admission erythrocyte sedimentation rate was 122mm per hour. Baseline investigations revealed ascitic, pleural and pericardial effusion. Effusions were tapped: neoplastic cells and acid-fast bacilli (AFB) were not identified, aerobic and mycobacterial culture resulted negative. QuantiFERON TB-Gold test was negative. Total body PET-CT and autoimmunity panel were negative. A neoplastic process was considered the most likely explanation. Before signing off the patient to comfort care, a reassessment was performed and an exposure to tuberculosis during childhood was documented. Because of constrictive pericarditis, pericardiectomy was performed: histologic examination showed chronic pericardial inflammation without granulomas, but Ziehl-Neelsen stain identified AFB and PCR was positive for Mycobacterium tuberculosis complex. Patient was started on anti-TB therapy with resolution of the effusions in the following months. Genes associated with defects in innate immunity were sequences and dentritic cells were studied, but no alterations were identified. A Bayesian approach to clinical decision making should be recommended. Interpretation of diagnostic tests should take into account the imperfect diagnostic performance of the majority of these tests. Further studies to investigate genetic susceptibility to tuberculosis are needed.