Interrater Reliability of a Modified Bronchoscopy Scoring Tool in Children With Cystic Fibrosis

Flexible bronchoscopy (FB) is widely used in the management of children with Cystic Fibrosis (CF) to visualize airway abnormalities, assess inflammation and detect infection. While previous scoring systems have been proposed to quantify visual airway findings in general pediatric populations, no standardized tool exists for assessing airway inflammation specific to CF.

We modified the previously proposed pediatric bronchoscopy scoring tool by adding four features relevant to CF pathology: mucus plugging, secretion viscosity, bleeding, and vascular drawing (abnormal or enhanced visualization of airway mucosal vessels, reflecting neovascular remodeling associated with inflammation). Eighty bronchoscopy recordings (50 CF, 30 non-CF) were retrospectively scored by four raters blinded to the clinical information, and ten visual features were assessed: six from the previously proposed score (secretion amount and color, mucosal edema, erythema, ridging, and pallor) and the four CF-specific additions. Inter-rater reliability was assessed using Gwet's AC2 coefficient.

Agreement between raters varied across features. Mucus plugging (AC2 = 0.94) and bleeding (0.80), both CF-specific, were among the most reliably scored features, while secretion viscosity (0.47) and vascular drawing (0.50) showed the lowest agreement. The expanded score demonstrated comparable or improved reliability for overlapping features from earlier scoring systems.

The modified bronchoscopy score demonstrated moderate to excellent inter-rater reliability and added clinically relevant features specific to CF. It may serve as a standardized method to assess bronchoscopy for pediatric CF lung disease, although further validation is needed for features with lower inter-rater reliability.