RESUMEN
BACKGROUND: The diagnosis of paediatric tuberculosis (TB) remains difficult in resource-: poor settings. OBJECTIVE: To evaluate induced sputum collection and examination using microscopy, culture and Xpert(®) MTB/RIF assay for the diagnosis of pulmonary TB (PTB) in a Tanzanian hospital vs. PTB diagnosis using clinical scoring tools alone. METHODS: We conducted a cross-sectional study from October 2013 to April 2014 at our hospital in northwestern Tanzania. Children presumed to have TB were assessed using four TB score charts and sputum examination. Sputum samples were analyzed using fluorescence microscopy, solid culture and Xpert. The number of cases microbiologically confirmed was compared to the number of TB cases suspected based on TB score charts. RESULTS: A total of 192 patients were enrolled. Sputum specimens were successfully obtained in 187 (97.4%) patients without any major complications. Ten (5.2%) children were confirmed to have PTB by sputum examination. More than half (50-90%) of the confirmed cases were not detected by score charts alone. CONCLUSION: Sputum induction is both safe and feasible in a severely resource-limited hospital, and can lead to microbiological PTB diagnosis that would not be detected by clinical criteria alone.