Abstract
Rationale The epidemiology of continuous ketamine use in the management of critically ill adults receiving invasive mechanical ventilation (MV) in the U.S. is unknown. Objective To quantify the temporal trends and variation across U.S. hospitals in continuous ketamine use. Methods We performed a retrospective cohort study of adults (age ≥18) receiving MV who were discharged (alive or dead) from U.S. hospitals contributing data to the Premier Healthcare Database between January 2008 and September 2018. We used mixed effects multivariable logistic regression modeling (fixed effects: patient and hospital characteristics; random effect: discharge hospital) to evaluate the associations of discharge quarter and discharge hospital with continuous ketamine use (defined as a charge for intravenous ketamine on ≥2 consecutive calendar days). Results We identified 2,059,599 MV adults across 842 hospitals; of these, 7,927 (0.4%) received continuous ketamine. Ketamine use increased over time from 0.07% of all patients in quarter 1 2008 to 1.1% of all patients in quarter 3 2018. After adjustment, the odds of receiving continuous ketamine were consistently increased starting in quarter 4 2011 (odds ratio [95% confidence interval]: 1.83 [1.09-3.06] vs quarter 1 2008, p=0.023) with >10-fold increased odds starting in quarter 2 2017. Out of 842 hospitals, 486 (57.7%) used continuous ketamine on at least one cohort patient during the study period. Across these hospitals ever using ketamine, median use was 0.2% (IQR 0.08%-0.5%) with 5 hospitals using continuous ketamine in >5% of patients. The adjusted median odds ratio for discharge hospital was 3.72 (95% confidence interval: 3.37-4.13) which was higher than the odds ratio for any patient or hospital covariable other than discharge quarter. Conclusions In U.S. hospitals, use of continuous infusion ketamine increased markedly over time in critically ill patients receiving MV, with substantial variability between hospitals. Given the unique properties of ketamine as a sedative, further research is needed to assess its safety and utility in critically ill populations.
