Incidence of cognitive errors in difficult airway management: an inference human factors study from the Pediatric Difficult Intubation Registry.

Background: Cognitive errors are known contributors to poor decision-making in healthcare. However, their incidence and extent of their contribution to negative outcomes during difficult airway management are unknown. We aimed to identify cognitive errors during paediatric difficult airway management using data from the Pediatric Difficult Intubation (PeDI) registry, to determine patient and clinician factors associated with these errors, and their contribution to complications.

Methods: We conducted a retrospective analysis of the PeDI registry data including cases with at least three intubation attempts. Cognitive error definitions were adapted to airway management, and predefined clinical endpoints were used to identify cognitive errors. A subanalysis was performed for children weighing <5 kg. Our primary outcome was the overall incidence of cognitive errors. Secondary outcomes included the incidence of specific cognitive error subtypes, associations with patient and clinician factors, and the relationship between cognitive errors and complications.

Results: Cognitive errors were identified in 17.4% (487/2801) of cases, with fixation errors being the most common (11.5%), followed by omission bias (5.9%) and overconfidence bias (4.5%). Non-anaesthesiologist clinicians had the highest odds of cognitive errors. The presence of at least one cognitive error was independently associated with a higher risk of complications (adjusted odds ratio, 1.86 [95% confidence interval, 1.53-2.27]; P<0.001), and multiple errors increased the likelihood of severe complications (adjusted odds ratio, 2.48 [95% confidence interval, 1.24-4.94]; P=0.01).

Conclusions: Cognitive errors occurred in nearly 20% of paediatric difficult airway encounters and were linked to increased complications. Further research should refine error definitions and develop mitigation strategies to improve outcomes.