Sunday, December 23, 2007

CPR and EBM

I was about to cite a couple of new articles and an editorial from Circulation to seize a SITYS (see I told you so) moment when I found that California Medicine Man beat me to it. It concerned predictions I made in a series of posts starting in October of 2005 about changes I thought were overdue in the guidelines for CPR and emergency cardiac care. California Medicine man beat me not only to the SITYS (not that that was his intent) but also to a post back in 2005 which predated mine by several months (back before I started blogging).

The posts concerned data on CPR from a group of Arizona investigators spearheaded by Dr. Gordon Ewy, Chief of Cardiology at the University of Arizona College of Medicine and author of the cited editorial. In it Ewy notes poor survival rates following CPR and asks:

Why have survival rates not improved? One possibility is that the guidelines are not optimal. The guidelines advocate the same approach for 2 entirely different pathophysiological conditions: respiratory arrest in which severe arterial hypoxia and hypotension eventually lead to secondary cardiac arrest, and primary cardiac arrest in which the arterial blood is fully saturated with oxygen at the time of the arrest.

Despite that rationale (and it’s pretty strong pathophysiologic rationale) to adopt compression only resuscitation for certain types of cardiac arrest, traditional CPR with compressions and rescue breathing has persisted through multiple iterations of the guidelines. Ewy, himself an early leader in this field, has been telling us for years that we’ve been doing it all wrong. Why have the guideline writers ignored his pleas? I believe it’s an unintended consequence of evidence based medicine (EBM).

Don’t get me wrong. I’m an advocate of EBM. There’s nothing in EBMs basic principles, which tell us to seek the best and most current evidence available to guide treatment, to stop guideline writers from adopting Ewy’s recommendations. Rather, the problem seems to be a popular distortion of EBM which summarily rejects pathophysiologic rationale, animal studies and "low level" human data, anything, in fact, other than a randomized controlled trial (RCT). I’m a fan of the RCT but I recognize that there are some questions which are extremely difficult to subject to that method of investigation. In such cases we must go with the best available information.

So I have a somewhat different take on CPR and EBM from that of California Medicine Man. He seems to think EBM will come to the rescue and bring about needed guideline changes. That may be true. But I feel EBM (well, not really EBM but the popular distortion of EBM I referenced above) has stalled important updates to CPR protocols. I made that point earlier this year in a Medscape Roundtable piece about EBM controversies:

Sometimes, EBM proponents' insistence on patient outcome-based studies slows the wheels of medical progress. The American Heart Association 2005 guidelines for emergency cardiac care provide a recent example. Investigators at the University of Arizona, Tucson, have been publishing evidence for years in support of changes in emergency cardiac care for adult victims of out-of-hospital cardiac arrest.[8,9] Despite this evidence, the guideline authors, dutifully trying to be evidence-based, failed to adopt the new methods because of a reluctance to base changes on anything other than randomized trials. However, the evidence, although based on "low-level" physiologic rationale, was compelling. More recently, the Arizona investigators were vindicated by direct evidence that the new procedures save lives. This situation represents a failure of EBM, not due to any shortcomings of EBM itself, but due to the misappropriation of its principles.

Ewy expresses a similar view in the editorial even though prospective randomized trials comparing the two methods are finally underway:

Will these reports finally be enough to result in guideline changes in the near future, or as suggested by the authors of the Swedish study, should this decision await the published outcome of randomized trials of CCC and CPR now underway? My conclusion is that guideline changes should be made as soon as possible because they are long overdue.

So that’s the question. If the RCTs now underway prove that traditional CPR with rescue breathing is superior to compression only CPR for bystander-witnessed primary cardiac arrest the EBM extremists will be vindicated.

What does the evidence show so far? California Medicine Man’s posts summarize it pretty well. Not mentioned is this study of implementation of the changes in a region in Wisconsin, showing that the new protocol is associated with improved survival as well as data presented by Ewy’s group at the American Heart Association meetings in 2006 showing markedly improved survival in Phoenix resulting from compression only CPR. In summary, we have pathophysiologic rationale, strong data from animal studies as well as comparison studies in humans using historical controls which show either equivalency or superiority of compression only CPR for survival and neurologic outcomes.

The zealots of EBM like to trot out the CAST study and trials on estrogen replacement therapy as examples of the hazards of over reliance on low level data. Those are important lessons, to be sure, but one has to wonder whether the insistence on RTCs is always appropriate or, as I asked in my Roundtable, whether we’ve gone too far with EBM.

No comments: