Comparative study of diagnostic efficacy of sputum and bronchoalveolar lavage fluid specimens in community-acquired pneumonia children treated with fiberoptic bronchoscopy

BACKGROUND: Community-acquired pneumonia (CAP) is usually diagnosed in children, and the type of respiratory specimen is critical. Differences in pathogens detection between induced sputum (IS) and bronchoalveolar lavage fluid (BALF) have not been evaluated. METHODS: In 2018, paired sputum and BALF samples from CAP hospitalised children with indications for bronchoalveolar lavage (BAL) were subjected to multiplex PCR for the detection of 11 common respiratory pathogens. RESULTS: A total of 142 children with paired sputum and BALF were tested. The overall positivity rate was 85.9% (122/142) for sputum and 80.3% (114/142) for BALF. The two specimens presented almost perfect agreement between the detection on M. pneumoniae, influenza A, influenza B, bocavirus and RSV. In contrast, adenovirus had the lowest kappa value of 0.156, and a false negative rate (FNR) of 66.7%. Rhinovirus had the highest false positive rate (FPR) as 18.5%. The consistent rate was significantly higher in school-age children than those under 1 year old (p = .005). Bacterial co-infection in BALF specimens were observed in 14.8% (21/142). Of the 11 discordant pairs of specimens, 9 cases were sputum(+)/BALF(-) with adenovirus predominating. CONCLUSION: Our findings suggest that the consistency of results between sputum and BALF is pathogen specific. Careful consideration needs to be given to whether sputum can be used as a substitute for BALF when children are young or co-infections with bacteria are suspected.