Comparison of Lower Airway Sampling Strategies in Children with Protracted Bacterial Bronchitis (CLASSIC-PBB): a non-randomised, internally controlled, multicentre trial.

BACKGROUND

Protracted bacterial bronchitis (PBB) is a leading cause of paediatric chronic wet cough. Identifying the causative pathogen promotes antibiotic stewardship, but bronchoalveolar lavage taken during flexible bronchoscopy (FB-BAL) is the only validated sampling strategy. We prospectively investigated if cough swab (CS) or induced sputum (IS) was a viable alternative.

METHODS

Children with PBB, referred for FB-BAL, provided a cough swab and induced sputum. Discordance in microbiological yield (failure to identify identical pathogens) between cough swab and BAL, and between induced sputum and BAL was assessed. This was performed for single-lobe (BAL1), two-lobe (BAL2) and six-lobe BAL (BAL6).

RESULTS

135 underwent FB-BAL, 134 produced a cough swab and 111 an induced sputum sample. 30 (22%) cough swab, 67 (60%) induced sputum and 101 (78%) BAL6 were pathogen positive. cough swab/BAL6 discordance was seen in 74% with a discordance OR (95% CI) for cough swab-BAL6+ versus cough swab+BAL6- of 30.3 (9.6 to 95.8). induced sputum/BAL6 discordance was seen in 71% with a discordance OR (95% CI) for induced sputum-BAL6+ versus induced sputum+BAL6- of 4.2 (2.4 to 7.4). The ability of each sampling strategy to identify all pathogens from cough swab, induced sputum and BAL6 was: cough swab 17%, induced sputum 38%, BAL1 40%, BAL2 53% and BAL6 78%. The false negative rates were: cough swab 67%, induced sputum 27%, BAL1 33%, BAL2 16% and BAL6 9%.

CONCLUSIONS

The pathogen yield of both cough swab and induced sputum had high discordance with BAL6, meaning neither is a viable alternative to FB-BAL. However, the tolerability and relatively low false negative rate of induced sputum mean it can be part of early PBB assessment in children who do not warrant FB-BAL.