Posted on May 19, 2019 at 4:07 PM
by Daniel J Brauner, MD
Wait a minute! I made up the term DNE—Do Not ECMO—not because I thought we needed another RULE, but to motivate a cautionary tale about CPR that may help to avoid the order becoming necessary in the first place.
Attention to the changing indications for resuscitation over its more than 200 year history reveals what in retrospect was a natural experiment performed by the American Medical Association (AMA). The AMA conceived of the Current Procedural Terminology or CPT in response to their fears about increasing government involvement with the medical enterprise. Their goal was to maintain the “private practice” model of medicine under government sponsorship by creating a list of relative values to charge for procedures. First published in 1966, in time for the new Medicare program, CPT produced a wide range of effects, perhaps most importantly its influence on the standard of care, dictating what doctors do. One paradigmatic example is revealed in the history of resuscitation. A review of its various incarnations prior to CPR shows that its indications were limited to those not at the end of terminal conditions. This remained true for CPR as revealed by its developers, who emphasized the limited indications for the procedure throughout the 1960s. It was the publication of the massively larger (by 70%) second edition, released in 1970, that let loose the true power of their intervention. In a testament to this power, CPR became the default for all patients in cardiac arrest in the early 1970s, shortly after it was listed as a billable procedure: “CARDIOPULMONARY RESUSCITATION FOR CARDIAC ARREST—-96000” in the second edition of the CPT manual.
The current standard for ECMO, especially when applied with or as a substitute for CPR, both called ECPR, is reminiscent of how CPR was practiced in the 1960s. In all the many studies which consistently show ECPR to be superior to CPR alone for in-hospital resuscitation, extreme caution is taken to distinguish the indications for ECPR from that of CPR. In these studies ECPR is only to be applied to patients who have a reversible etiology of their cardiac arrest that is unexpected and not those at the end of terminal conditions. However, given its superiority over CPR alone, the fear among many is that ECMO will become part of the CPR protocol, making it a default for all cardiac arrest. In our paper published in the AMA’s Journal of Ethics, my co-author Christopher Zimmermann and I suggest that one way to avoid this becoming a reality is to not create a CPT code linking ECMO with cardiac arrest, as happened to CPR in our cautionary tale. Patients and their families will be much more empowered to make decisions about whether they will avoid or undergo ECMO during their critical illness, and clinicians will retain the ability to offer or recommend ECMO based on their clinical judgment, if ECMO does not become a default. If ECMO is not added to the list of default procedures to which patients must say, NO!—in order to avoid their application at the end of their lives, we can avoid the need for a DNE order.
The larger implications of this cautionary tale is the power of CPT to control practice in ways that are not evidence based or beneficial for many of the patients to whom these procedures are applied. Perhaps it is time to acknowledge the downside of the experiment and move beyond CPT codes.