Yield of preoperative findings in pediatric airway foreign bodies - A meta-analysis.

IMPORTANCE

Foreign body (FB) aspiration into the airway is a significant cause of pediatric morbidity and mortality, yet the clinical presentation is diverse and dynamic. There are conflicting recommendations which pre-procedural findings support performing a bronchoscopy, the gold standard for diagnosis and removal of FBs, however a procedure that entails general anesthesia and possible risks.

OBJECTIVE

Decision whether to proceed to a bronchoscopy may be challenging. Our goal was to enhance decision-making by analyzing the diagnostic values of the different pre-procedural findings in this setting.

DATA SOURCES: A comprehensive search was performed in PUBMED, EMBASE and Cochrane Review databases to find studies from the last 19 years that reported pre-procedural history, physical examination and radiological findings in patients who had bronchoscopies.

STUDY SELECTION: Studies were included of pediatric populations if they contained bronchoscopy results (positive and negative for foreign body) with a breakdown according to pre-intervention findings.

DATA EXTRACTION AND SYNTHESIS: Titles and abstracts retrieved from our search were screened. Thereafter, full-texts were carefully reviewed and selected for inclusion if the aforementioned criteria were met. PRISMA guidelines for systematic review and meta-analyses were followed.

MAIN OUTCOME(S) AND MEASURE(S): Cumulative weighted prevalence, sensitivity, specificity, positive and negative predictive values of each pre-procedural finding were calculated, as well as for the "classic triad" (history of an acute event, wheezing, and unilateral decreased breath sounds). Calculation for other combinations of findings, or optimally, constructing a weighted score based on all the findings for each specific patient were not possible to perform, as the specific data breakdown is rarely reported.

RESULTS

Fifteen studies met inclusion criteria, totaling 5606 patients who underwent bronchoscopies. All studies but one were single center based and all except one were retrospective. No single finding has both positive and negative predictive values over 50%. The "classic triad" has 90% specificity, however only 35% sensitivity.

CONCLUSIONS

The data is very heterogeneous with regard to pre-procedural findings and how best to guide treatment according to them. This meta-analysis provides cumulative weighted metrics for each finding, to optimize decision-making for the individual patient. Future reporting of data should be enhanced, so that combinations of findings for a specific patient can be used to optimize management.

LEVEL OF EVIDENCE: 4.