Objective: Most studies of ketamine administered to children for procedural sedation are limited to emergency department use. The objective of this study was to describe the practice of ketamine procedural sedation outside of the operating room and identify risk factors for adverse events.
Design: Observational cohort review of data prospectively collected from 2007 to 2015 from the multicenter Pediatric Sedation Research Consortium.
Setting: Sedation services from academic, community, free-standing children's hospitals and pediatric wards within general hospitals.
Patients: Children from birth to 21 years old or younger.
Measurements and main results: Describe patient characteristics, procedure type, and location of administration of ketamine procedural sedation. Analyze sedation-related adverse events and severe adverse events. Identify risk factors for adverse events using multivariable logistic regression. A total of 22,645 sedations performed using ketamine were analyzed. Median age was 60 months (range, < 1 mo to < 22 yr); 72.0% were American Society of Anesthesiologists-Physical Status less than III. The majority of sedations were performed in dedicated sedation or radiology units (64.6%). Anticholinergics, benzodiazepines, or propofol were coadministered in 19.8%, 57.9%, and 35.4%, respectively. The overall adverse event occurrence rate was 7.26% (95% CI, 6.92-7.60%), and the frequency of severe adverse events was 1.77% (95% CI, 1.60-1.94%). Procedures were not completed in 39 of 19,747 patients (0.2%). Three patients experienced cardiac arrest without death, all associated with laryngospasm.
Conclusions: This is a description of a large prospectively collected dataset of pediatric ketamine administration predominantly outside of the operating room. The overall incidence of severe adverse events was low. Risk factors associated with increased odds of adverse events were as follows: cardiac and gastrointestinal disease, lower respiratory tract infection, and the coadministration of propofol and anticholinergics.