Timing the key to meaningful survival: The importance of early epinephrine in cardiac arrest

Treating time-sensitive conditions is a daily occurrence for emergency physicians and for prehospital personnel, including first responders, emergency medical technicians, and paramedics. We prioritize administering antibiotics early in sepsis, expediting a stroke patient to CT, or activating the catheterization laboratory quickly for an acute myocardial infarction because reducing time to these interventions improves patient outcomes.1-5 The ultimate time-sensitive condition is cardiac arrest given that the probability of death rises every minute that passes without CPR or defibrillation.6-8 Epinephrine is commonly administered in cardiac arrest, but data are limited showing a clear survival benefit. Unlike defibrillation, which is clearly understood to have a time dependent relationship with survival,6 epinephrine administration historically has been viewed as a binary intervention without consideration of the impact of the timing of administration on survival. The American Heart Association guidelines recommend use of epinephrine in cardiac arrest and that it is reasonable to administer epinephrine as soon as feasible in patients with a nonshockable rhythm.9 As clinicians, we generally ask the general question of whether epinephrine uniformly improves functional survival, without regard to factors such as duration of no-flow time or resuscitation time bias. While this is an important question, this binary outcome does not allow for circumstances of the arrest that might confound the answer. Sakamoto et al. in this month's issue of AEM asks a more specific question: is timing of epinephrine administration in nonshockable cardiac arrest associated with survival outcomes? In this observational study, the authors analyzed out-of-hospital cardiac arrest (OHCA) data from 95 hospitals in Japan with time-dependent propensity score and risk set matching, with nearly 37,000 eligible patients. The time interval between EMS contact and epinephrine administration was categorized into 5-min time intervals, with a primary outcome of functional neurological survival at 30 days. The authors found that there was a significant association with functional survival if epinephrine was administered within 10 min of EMS contact. There was also a significant association with overall survival at 30 days if epinephrine was administered within 20 min. Notably, the median time from EMS personnel contact to epinephrine administration was 21 min. Ultimately, less than 20% of the patient cohort received epinephrine within 10 min of arrival. A potentially significant limiter here is Japanese EMS personnel are unable to place intraosseous (IO) lines. This study is important as it is one of the largest cohorts of nonshockable patients specifically looking at the question of survival in relation to epinephrine timing in cardiac arrest and accounting for resuscitation time bias. The results are clear and make intuitive sense; the sooner a nonshockable patient can get epinephrine, the higher the chance of functional survival. Accordingly, these findings are consistent with previous observational study. Hansen et al.10 analyzed just over 32,000 adult and pediatric OHCA patients with nonshockable rhythms, demonstrating most patients received epinephrine >10 min after EMS arrival, and each minute delay was associated with unfavorable neurological outcomes. The PARAMEDIC-2 trial, performed in the United Kingdom and published in 2018, randomized adult OHCA patients to intravenous epinephrine or placebo, with a primary outcome of 30-day survival.11 The trial found that epinephrine did result in improved 30-day survival, but no difference was observed in favorable neurological survival between groups. A key finding of this trial was that the median time from emergency call to administration of the trial agent was about 21 min. In a subsequent paper, the authors examined the time to treatment with epinephrine on 30-day survival in witnessed arrest patients.12 The main finding was that survival decreased with increasing time to treatment with both epinephrine or placebo. The randomized trial design allowed for observation of an unbiased treatment effect in both study arms; unsurprisingly the best outcomes were seen early after arrest regardless of what the patient was given. The question that naturally arises from this previous work is whether earlier administration of epinephrine leads to increased neurologically intact survival and if so, how can epinephrine be administered earlier? One approach is to change the route of administration. Palatinus et al.13 recently addressed this question, publishing the experience of the Salt Lake City Fire Department with intramuscular (IM) epinephrine. This was a before-and-after study of adult nontraumatic OHCA patients from January 2010 to May 2024, with the intervention starting in November 2019. At that time, paramedics started administering an initial 5-mg IM epinephrine dose to the lateral thigh, simultaneous with attempted placement of an IV line. Further epinephrine was given in 1-mg doses either IO or IV. With this approach, a reduction of approximately 3.5 min in time to administration of epinephrine was observed in the IM cohort of patients. In the total cohort of 1405 patients, IM epinephrine was associated with the primary outcome of survival to hospital discharge as well as survival to hospital admission and functional neurological survival. Given the nonrandomized nature of this study, definitive conclusions cannot be established, but it raises the question of whether the adoption of an IM-first approach may have benefit in OHCA where limited time and specifically trained personnel make rapid IV or IO access and medication administration particularly challenging. This has significant implications for EMS, including the possibility of earlier treatment options for basic life support–trained first responders prior to paramedic arrival as well as the possibility of publicly accessible autoinjectors analogous to the EpiPen® for anaphylaxis that could be placed with public access automated external defibrillators. Further work, including a randomized trial is warranted based on these preliminary but promising data. The obvious limitation of both studies by Sakamoto and Palatinus is that both are observational, nonrandomized studies and therefore prone to the biases inherent in all studies of this type. It is easy as a disciplined researcher and reader of the literature to dismiss these findings as unproven or coincidence. However, some of these study questions could never be ethically performed. As Sakamoto et al. discuss, for example, it would not be ethical to randomize OHCA nonshockable patients to early or late epinephrine. This question is likely only to be answered in the current format, large observational data sets. Further, cardiac arrest research is extraordinarily difficult with unique challenges. The desired outcome, functional neurological survival, is infrequent. On average, less than 10% of Americans walk out of the hospital after cardiac arrest.14 The patient population is heterogeneous; there are many causes of arrest. It is impossible to prospectively identify and consent patients for a study. Many sites are usually needed to get the sample size needed for the desired power. It is a time-intensive and expensive effort to coordinate sites and to train and maintain the study protocol with study personnel. Maintaining CPR quality metrics during a study period is paramount, as this is not only best patient care and the basis of any resuscitation but can also be a significant effect modifier on the intervention if done poorly. These studies require significant investment and most often result with equivocal findings, leaving the resuscitation community to wonder if we are asking the right questions and if we are studying them the right way. The current stagnant survival rates tell us what we are doing is unfortunately not working. New research perspectives and strategies are needed. As new resuscitation ideas and interventions emerge for study, we must consider resuscitation bias, the window of treatment effect, and the population in which it would be the most effective. The authors have no conflicts to disclose. Data sharing is not applicable to this article as no new data were created or analyzed in this study.