The objective of this study was to derive pooled estimates of the diagnostic performance of three tools, including history, physical exam, and radiography, commonly used to identify airway foreign bodies (AFB) in children relative to the gold standard of endoscopy.
Medline, EMBASE, and all Cochrane Collaboration Evidence-based Medicine databases were searched and a meta-analysis of published articles was performed using a random-effects modeling approach.
Nineteen articles were identified, with a total sample size of 2,665. History, including sudden onset of symptoms, witnessed choking, and coughing with or without cyanosis, is a sensitive tool (sensitivity 95%) with a high positive predictive value (PPV) (91%), but has limited specificity (35%) and a low negative predictive value (NPV) (54%). The physical exam, including cough, localized wheezing, and decreased localized breath sounds, was less sensitive (74%), but had higher specificity (90%) with a high PPV (97%) and a similar NPV (52%). Finally, the diagnostic performance of anterior-posterior and lateral chest radiography was lower (sensitivity 73%, specificity 60%, PPV 86%, and NPV 48%).
Diagnostic performance of the physical exam, historical symptoms, and radiography were limited relative to the gold standard of airway endoscopy. Historical symptoms had the greatest sensitivity and physical exam had greatest specificity. Of the three tools, radiography yielded a lower overall diagnostic performance. There is a need for development of safe, more accurate, and less invasive diagnostic tools to identify AFB in children.