Editor’s Note: The following guest post by Daniela Lamas is reprinted with permission from Now@NEJM, a blog for physicians about NEJM.
The woman slumps to the floor beside her office desk… No one sees her go down… A colleague hears a thud and rushes in….. He calls her name… No response… He checks for a pulse: nothing… “Call 911!” he yells.
EMS arrives quickly, and finds the co-worker panicked, but performing good quality chest compressions he learned in a life support class.
The emergency medical technicians take over CPR while setting up the defibrillator, as dictated by current standard of care. But when in this un-witnessed arrest should they pause, analyze the rhythm and shock if indicated?
At this point the recommendations become shaky. Analyze first? That was the standard until 2005, when guidelines from the American Heart Association-International Liaison Committee on Resuscitation were revised to suggest, instead, that EMS personnel provide 2 minutes of CPR before the first analysis of cardiac rhythm. The theory behind this change is that increasing myocardial perfusion could result in more effective first defibrillation. But this hasn’t been clearly borne out in studies, which have offered inconsistent evidence. Thus, in 2010, in the face of conflicting data, the AHA committee revised their guidelines to say there was actually no clear evidence to back up early or later analysis.
All this leaves those EMS personnel at the scene of this out-of-hospital, un-witnessed arrest with some uncertainty. Should they analyze the rhythm as soon as possible, after a brief period of CPR? Or should they wait for a full round of CPR before first analysis? In a situation with such dismal outcomes, would this change make any difference at all?
The answer is: No, there’s no benefit to waiting. That is based on results of a large-scale, randomized trial of adults with out-of-hospital cardiac arrest, “Early versus Later Rhythm Analysis in Patients with Out-of-Hospital Cardiac Arrest,” published in last week’s issue of NEJM. In this study, Stiell and colleagues enrolled 9933 patients with out-of-hospital cardiac arrest in sites throughout the United States and Canada. The early analysis group was assigned to receive 30-60 seconds of EMS-administered CPR before rhythm analysis. In the later analysis group, EMS was to administer 180 seconds (3 minutes) of CPR directly upon arrival at scene.
The primary endpoint was survival to hospital discharge with “satisfactory functional status” – which study investigators defined as a score of 3 or less on the Rankin scale, which measures disability or dependence in daily activities. (A score of 3 describes moderate disability, requiring some help but able to walk unassisted, for instance).
There was no significant difference in the primary outcome between the two study groups: 6.0% of the patients with shorter CPR, and 5.9% of the patients with longer CPR left the hospital with a satisfactory functional status. When the authors looked at survival as a function of actual time to rhythm analysis (regardless of how the patients had been assigned), they again found that the chance of a satisfactory outcome didn’t improve with more time to first rhythm analysis. Interestingly, the rate of survival actually declined with longer EMS-administered CPR in those who had received CPR first from a bystander and whose first analyzed rhythm was ventricular tachycardia or ventricular fibrillation.
The authors conclude, “Overall, our data suggest that the administration of 2 minutes of CPR by EMS personnel before the first analysis of rhythm…is unlikely to provide a greater benefit than CPR of shorter duration.”
What about no CPR at all, prior to first analysis? That question remains unanswered, as the study investigators write that they “deliberately insisted on some CPR for the early-analysis group, in the belief that good patient care required cardiopulmonary support while the defibrillator was being prepared.”
While the data showed no benefit to waiting a longer time for rhythm analysis, it’s important to note that 36% of the patients in the study did not receive the assigned compression-to-analysis time. Does this negate the results? The authors argue that this limitation points to the difficulties inherent in performing a randomized controlled trial in a life-or-death, often messy and uncontrolled situation with so many potential variables.
In an accompanying editorial, Arthur B. Sanders, an emergency medicine physician who researches efficacy of resuscitation, addresses precisely this challenge. He writes that perhaps randomized-controlled trials are not the best way to answer these questions: “The urgency of the setting, the heterogeneity of initiating causes and clinical features of the arrests, the fact that informed consent must be waived…all makes trials problematic. It may be more useful to consider out-of-hospital cardiac arrest as a public health problem rather than as a disease process.”
Which leaves us with our patient. EMS is at the scene. How to give her the best chance to leave the hospital with as little disability as possible?
They’ve been performing CPR for a minute now. Should they pause and analyze the rhythm? It seems, on the basis of this study, that there’s no benefit in waiting, and it’s probably time to see what’s going on.
Question: How would you acquire more data on resuscitation? If randomized controlled trials aren’t the way to go, what would you suggest?