If at First You Don’t Get ROSC: Dose, Dose Again…*

Improving survival after cardiac arrest, with favorable neurologic outcomes, is the holy grail of cardiac arrest resuscitation science. One of the principal tools in this quest has been epinephrine, which has repeatedly been demonstrated to improve return of spontaneous circulation (ROSC), although not always improving survival to hospital discharge, particularly with good neurologic outcome, when compared with placebo (1,2). It is clear that increasing the dose of epinephrine is not beneficial in improving outcomes (3). Yet the impact of providing lower doses of epinephrine in more frequent intervals, which has the net effect of providing more epinephrine while minimizing the size of the bolus dose, has been poorly studied. Current pediatric cardiac arrest guidelines recommend dosing epinephrine at 3–5-minute intervals based on limited data and "no studies of pediatric OHCA on frequency of epinephrine dosing" (4). This lack of data has fueled several recent observational studies in pediatric and adult cardiac arrests aimed at identifying the optimal dosing interval for epinephrine. Recent studies have had mixed findings about what epinephrine dosing interval provides the best associations with outcome. A 2017 retrospective analysis of 1630 pediatric in-hospital cardiac arrests (IHCAs) from the Get With The Guidelines-Resuscitation (GWTG-R) registry found that longer average dosing intervals of epinephrine administration (1–5 min vs. > 5 to <8 min and 8 to < 10 min/dose) were associated with improved survival to hospital discharge (5), but only after heavy statistical adjustment. The crude associations in this study all demonstrated worsened outcomes with more infrequent epinephrine dosing than every 1–5 minutes. The analogous adult IHCA study from GWTG-R showed the shortest epinephrine dosing interval (1–3 min) was optimal (6). A secondary analysis of the Therapeutic Hypothermia After Pediatric Cardiac Arrest In-Hospital (THAPCA-IH) trial found epinephrine dosing interval of 3 to less than 5 minutes were associated with the best 12 months survival compared with both shorter or longer dosing intervals (7). It is noteworthy that over half the children in THAPCA-IH were placed on extracorporeal support, a setting in which less frequent epinephrine dosing after the initial (10 min) resuscitation is associated with better outcomes (8). In a 2019 study of 15,909 adult patients with out-of-hospital cardiac arrest, Grunau et al (9) found that a shorter average epinephrine dosing interval was associated with improved survival with favorable neurologic status. Finally, Kienzle et al (10) single-center retrospective analysis from 2021 found that epinephrine dosed at intervals of 2 minutes or less was associated with increased odds of survival with favorable neurobehavioral outcome compared with more delayed dosing. The range of associations seen in these recent studies invite further analysis, which is what Kienzle et al (11) have provided us in this issue of Critical Care Medicine. This is a post hoc analysis of the prospective multicenter hybrid stepped-wedge cluster-randomized trial of a quality improvement bundle (The ICU-RESUScitation Project [ICU-RESUS]) (12). The trial included 18 participating pediatric general and cardiac ICUs from high-volume quaternary care centers. Importantly, the interventional bundle emphasized education regarding intra-arrest and post-arrest physiologic targets. In other words, ICU-RESUS was a study where physiology was brought to the forefront at high performing centers. The targets were higher diastolic blood pressure from an arterial line and end-tidal co2 from in line capnography. Resuscitating teams were therefore trained to interpret real time physiologic monitoring data from their patients and adapt CPR to optimize targets. It is not surprising that these centers did very well in achieving high quality CPR measured by compression depth and rate and minimal interruptions. The design of ICU-RESUS is the optimal place to look at divergence from guidelines within an observational study. Highly trained adequately resourced teams were given physiologic targets previously demonstrated in animal (13) and human (14) studies to optimize CPR outcomes. One would expect that the teams would therefore modify their resuscitation not on the basis of random chance, rather in an effort to achieve the physiologic targets. The earlier than guideline recommended dosing of epinephrine in this study (11) and its single-center precursor (10) was 25%, which is substantially higher than 15% in GWTG-R (9). The "bias by indication" the authors allude to in their discussion is a well-trained team running an exceptional resuscitation and choosing when to accelerate epinephrine dosing to get results: faster and more frequent ROSC. Equally important is the finding that in those patients where guideline compliant (less frequent) dosing of epinephrine was provided, the more important outcomes of hospital survival and neurologic outcomes were comparable as were post-resuscitation markers of perfusion such as lactate and 6-hour Vasoactive-Inotrope Score. Bolus epinephrine is not without adverse effects, and in some populations early or more frequent use may worsen outcomes (8,15). Other findings from this study support the concept of hemodynamically titrated epinephrine dosing. Intervals less than 3 minutes were associated with favorable neurologic outcome in the subset of patients on a vasoactive infusion. This is the group one would most expect to need a vasoactive agent as they were already on one. Sensitivity analyses examining other divergences from Pediatric Advanced Life Support guidelines (4) such as dosing at intervals of less than or equal to 2 vs. greater than 2 minutes or less than 3 vs. 3–5 vs. greater than 5 minutes were also associated with better outcomes. Here, again, the context is important—this was a study where physiology was being taught and prioritized, so the assumption should be that teams titrated to physiologic effect. As with all observational studies, the obvious critique is that unmeasured confounders rather than sound physiologic targeting by well-trained teams are mediating the observed associations. In theory, this problem could be overcome by randomization. But any randomized study of epinephrine dosing must beware the very real risk of practice misalignment (16). Few adept ICU physicians with access to continuous arterial blood pressure and capnography would feel comfortable delaying epinephrine dosing on a patient with low blood pressure and end-tidal co2. Likewise, providing additional epinephrine to a patient where these measures are adequate and ROSC is anticipated, or where defibrillation is needed (15) or extracorporeal circulation planned (8) is not sensible based on existing knowledge. Yet a randomized clinical trial (RCT) with insufficient flexibility in design (a common RCT malady) could force patients into a protocol-driven strategy that does not account for the physiology being observed in real time. Such randomization has the potential to misalign the patient with the optimal intervention based on their physiologic phenotype thus biasing the study to the null. In that light, the observational design of the present study (11) may be the best evidence to support increased flexibility in the resuscitation guidelines when they are reissued next year.