A comparison of fluoroscopic airway screening with flexible bronchoscopy for diagnosing tracheomalacia

Abstract

Flexible bronchoscopy (FB) is the gold standard method of diagnosing tracheomalacia but it is not always feasible in settings with limited resources. Fluoroscopy is sometimes performed as an alternative diagnostic tool but there are no prospective studies that have evaluated the diagnostic accuracy of airway fluoroscopy for tracheomalacia using a-priori definitions. We determined the sensitivity, specificity, and likelihood predictive ratio of airway fluoroscopy compared with FB in children suspected of having an airway abnormality. Airway fluoroscopic examination was undertaken within 2-weeks of a FB in children aged <18-years and reported by a pediatric radiologist blinded to FB data. Fluoroscopic and FB methods and diagnostic criteria were standardized and defined a-priori. Tracheomalacia diagnosed by FB were independently scored (mild, moderate, severe) by 2 pulmonologists in a blinded manner. In 22 children (median age 33 months, range 1-187) evaluated for airway abnormality, tracheomalacia was found in 21 children at bronchoscopy. Of these, fluoroscopy detected tracheomalacia in five children. Airway fluoroscopy was poorly sensitive (23.8%) but highly specific (100%), positive likelihood ratio was 8.6. However, in moderate-severe tracheomalacia, the sensitivity improved to 57.1% but the specificity reduced (93.3%). The agreement between bronchoscopists for tracheomalacia severity was excellent, weighted kappa 0.74 (95% CI 0.77, 0.98). Airway fluoroscopy cannot replace FB which remains the tool for definitively diagnosing airway malacia. However, in absence of other modalities for diagnosis fluoroscopy should be considered in the setting of persistent respiratory symptoms compatible with the clinical picture of tracheomalacia.