This study aimed to compare test characteristics of standard (lateral and posteroanterior or anteroposterior) chest radiographs with and without special views (expiratory or bilateral decubitus) in the emergency department evaluation of children with suspected airway foreign bodies.Methods
From 1997 to 2008, 328 patients with a suspected airway foreign body had standard and special view chest radiographs: 192 with left and right decubitus views, 133 with expiratory views, and 3 with both. Patients were excluded for cardiorespiratory disease, chest wall deformity, visible airway foreign bodies on standard views, or spontaneously expelled airway foreign bodies. After blinded radiologist review, standard plus special view test characteristics were compared to standard views.Results
Nine upper airway and 70 tracheobronchial airway foreign bodies were identified by direct visualization or bronchoscopy, and the remainder were ruled out by bronchoscopy (50 patients) or clinically (199 patients). The sensitivity and specificity of the radiographs were, respectively, decubitus cohort, standard views, 56% and 79% and standard+decubitus views, 56% and 64%; expiratory radiograph cohort, standard views, 33% and 70% and standard+expiratory views, 62% and 72%. For standard plus decubitus views versus standard views alone, the relative sensitivity was 1.0 (0.56/0.56; 95% confidence interval [CI] 0.81 to 1.23) and the relative 1–specificity was 1.76 (0.36/0.21; 95% CI 1.3 to 2.37). For standard plus expiratory views versus standard views alone, the relative sensitivity was 1.87 (0.62/0.33; 95% CI 1.23 to 2.83) and the relative 1–specificity was 0.93 (0.28/0.3; 95% CI 0.6 to 1.44).Conclusion
The addition of decubitus to standard views increases false positives without increasing true positives and lacks clinical benefit. The addition of expiratory to standard views increases true positives without increasing false positives, but test accuracy remains low and the clinical benefit is uncertain.