Predictors of Laryngospasm During 276,832 Episodes of Pediatric Procedural Sedation

Study objective: Laryngospasm is a rare but potentially life-threatening complication of sedation. The objective of this study was to perform a predictor analysis of biologically plausible predictors and the interventions and outcomes associated with laryngospasm. Methods: Secondary analysis of prospectively collected data from consecutively sedated patients, less than or equal to 22 years of age, at multiple locations at 64 member institutions of the Pediatric Sedation Research Consortium. The primary outcome was laryngospasm. The independent variables in the multivariable model included American Society of Anesthesiologists category, age, sex, concurrent upper respiratory infection, medication regimen, hospital sedation location, whether the procedure was painful, and whether the procedure involved the airway. The analysis included adjusted odds ratios (aORs) and predicted probabilities. Results: We analyzed 276,832 sedations with 913 reported events of laryngospasm (overall unadjusted prevalence 3.3:1,000). A younger age, a higher American Society of Anesthesiologists category, a concurrent upper respiratory infection (aOR 3.94, 2.57 to 6.03; predicted probability 12.2/1,000, 6.3/1,000 to 18.0/1,000), and airway procedures (aOR 3.73, 2.33 to 5.98; predicted probability 9.6/1,000, 5.2/1,000 to 13.9/1,000) were associated with increased risk. Compared with propofol alone, propofol combination regimens had increased risk (propofol+ketamine: aOR 2.52, 1.41 to 4.50; predicted probability 7.6/1,000, 3.1/ 1,000 to 12/1,000; and propofol+dexmedetomidine: aOR 2.10, 1.25 to 3.52; predicted probability 6.3/1,000, 3.7,/1,000 to 8.9/1,000). Among patients with laryngospasm, the resulting outcomes included desaturation less than 70% for more than 30 seconds (19.7%), procedure not completed (10.6%), emergency airway intervention (10.0%), endotracheal intubation (5.3%), unplanned admission/increase in level of care (2.3%), aspiration (1.1%), and cardiac arrest (0.2%). Conclusion: We found increased associations of laryngospasm in pediatric procedural sedation with multiple biologic factors, procedure types, and medication regimens. However, effect estimates showed that the laryngospasm prevalence remained low, and this should be taken into consideration in sedation decisionmaking.