Hospital Variation in Time to Epinephrine for Nonshockable In-Hospital Cardiac Arrest

Background: For patients with in-hospital cardiac arrests attributable to nonshockable rhythms, delays in epinephrine administration beyond 5 minutes is associated with worse survival. However, the extent of hospital variation in delayed epinephrine administration and its effect on hospital-level outcomes is unknown. Methods: Within Get With The Guidelines-Resuscitation, we identified 103 932 adult patients (≥18 years) at 548 hospitals with an in-hospital cardiac arrest attributable to a nonshockable rhythm who received at least 1 dose of epinephrine between 2000 and 2014. We constructed 2-level hierarchical regression models to quantify hospital variation in rates of delayed epinephrine administration (>5 minutes) and its association with hospital rates of survival to discharge and survival with functional recovery. Results: Overall, 13 213 (12.7%) patients had delays to epinephrine, and this rate varied markedly across hospitals (range, 0%-53.8%). The odds of delay in epinephrine administration were 58% higher at 1 randomly selected hospital in comparison with a similar patient at another randomly selected hospital (median odds ratio, 1.58; 95% confidence interval, 1.51-1.64). The median risk-standardized survival rate was 12.0% (range, 5.4%-31.9%), and the risk-standardized survival with functional recovery was 7.4% (range, 0.9%-30.8%). There was an inverse correlation between a hospital's rate of delayed epinephrine administration and its risk-standardized rate of survival to discharge (ρ=-0.22, P<0.0001) and survival with functional recovery (ρ=-0.14, P=0.001). In comparison with a median survival rate of 12.9% (interquartile range, 11.1%-15.4%) at hospitals in the lowest quartile of epinephrine delay, risk-standardized survival was 16% lower at hospitals in the quartile with the highest rate of epinephrine delays (10.8%; interquartile range, 9.7%-12.7%). Conclusions: Delays in epinephrine administration following in-hospital cardiac arrest are common and variy across hospitals. Hospitals with high rates of delayed epinephrine administration had lower rates of overall survival for in-hospital cardiac arrest attributable to nonshockable rhythm. Further studies are needed to determine whether improving hospital performance on time to epinephrine administration, especially at hospitals with poor performance on this metric, will lead to improved outcomes.