Sputum induction for microbiological diagnosis of childhood pulmonary tuberculosis in a community setting.

Setting: Sputum induction has increasingly enabled microbiological confirmation of pulmonary tuberculosis (PTB) in hospitalised children, but it has not been evaluated in a community setting.

Objective: To investigate the yield, feasibility and safety of sputum induction for the diagnosis of TB in children in a primary health care facility.

Design: A prospective study in a primary health care clinic in South Africa from April 2007 to June 2009. Consecutive children with clinically suspected PTB, with a household adult PTB contact or human immunodeficiency virus infected with respiratory symptoms were enrolled. History, clinical examination, tuberculin skin test and chest X-ray results were recorded. Two sequential induced sputum specimens were obtained for smear and culture.

Results: A total of 270 children were enrolled (median age 38 months); sputum induction was successful in 269 (99%); 65 (24%) children were clinically diagnosed, of whom 11 (16.9%) were microbiologically confirmed. An additional 18 children not clinically diagnosed had microbiological confirmation of PTB and were placed on TB treatment thereafter, increasing the diagnostic yield by 21.6%, from 65 to 83 cases. Sputum induction procedures were well tolerated; no major adverse events occurred.

Conclusion: Sputum induction is feasible and safe in a community setting. Sputum induction was useful for making a microbiological diagnosis, increasing the number of children diagnosed and treated for PTB.