by Craig Klugman, Ph.D.

A nurse in the UK was sanctioned with a ”caution order” (a warning or demerit) on her record for 24 months after she did not perform CPR on a patient who presented as “waxy, yellow and almost cold” when she was called to look at him in his hospital room. The patient was dead. However, since the patient lacked a DNAR order, under UK law and standard of care, she was obligated to begin CPR even though the process has not been known to revive a cold corpse. This incident begs the question as to why performing CPR is the default order and leads me to recommend a policy change: DNAR ought to be the default code status unless a patient or family opts-in for full-code.

CPR is not a procedure that is performed once-in-a-while. Heart disease in the number one cause of death in the United States and the second leading cause of death in the UK. In the United States, there are approximately 735,000 heart attacks per year according to the Centers for Disease Control & Prevention. The American Hospital Association says that 350,000 of those heart attacks occur outside of the hospital.

That CPR became the default code status is an accident of history. This story is one of a medical discovery combined with an effective marketing campaign and the excitement of being able to stave off death. In 1960, an article in JAMA described the first successful external cardiac massage: Fourteen of 20 patients at Johns Hopkins who experienced cardiac arrest survived after receiving CPR. Before that time, all cardiac massage was done directly, by cutting into the chest and grabbing hold of the heart. By 1959, training brochures and procedures had been created by the American Red Cross to teach people how to do “rescue breathing,” a process that was adopted soon thereafter by the major medical associations. In 1963, the American Heart Association endorsed external cardiopulmonary resuscitation (CPR) and in 1966 the National Academy of Sciences held a conference to standardize CPR training and performance. Of course, these recommendations were based on witnessed CPR occurring in the hospital. Today, every health care provider, lifeguard, police office, firefighter, military personnel, and many others are required to be certified in basic life support (BLS). Many health professionals are trained further in advance cardiac lifesaving (ACLS). The American Red Cross and American Heart Association have led a decades-long public service campaign encouraging the public to learn CPR by becoming “certified.”

The DNR has a less clear history. The American Medical Association is reported to have recommended documenting requests against CPR in medical charts as early as 1974. By 1975. The American Heart Association recommend documenting when a physician believes CPR is not indicated after a conversation with patient or surrogate. New York passed the nation’s first DNR law in 1987.

Full code became the default status simply because we could do it, because of licensing requirements and public service campaigns, and because in medicine, death is viewed as the enemy. The technological imperative says that if we can do it, then we ought to do it. This is not based on reason or analysis, but on the simple fact that medicine (rightly) has a bias toward saving lives no matter the consequences or outcome. In fact, The Institute of Medicine found that the chances of surviving a cardiac arrest outside of a hospital with CPR is 6%. Chances of surviving with the brain intact are 2%. If one codes in a hospital, the survival rate with CPR is 24%. Even the American Heart Association’s optimistic rates shows that only 12% of people make it to hospital discharge (but that does not indicate their brain function). The financial cost of successful CPR (meaning surviving for 6 months) is estimated at $555,300 to $1,559,162 (updating the costs estimated in a 1995 study to 2017 dollar values). Obviously, this does not include the cost of treatment for the 60-98% of patients who do not survive CPR attempts but nonetheless undergo it.

A 2012 JAMA article posited that by having DNR as an opt-out to CPR (as opposed to opt-in), in medicine sends a message to patients and families that there may be some benefit to it, otherwise why would a code discussion be raised. The authors, Blinderman, Krakauer and Solomon, recommend a change away from always offering CPR that would require detailed conversations with families and recommendations based on likelihood of benefit.

Over the last 30 years, several interventions have been recommended to help increase the number of DNRs and percent of adults who undergo advance care planning. These suggestions include changing the culture of medicine, training physicians in communication skills, and moving away from pay for procedure (which creates a culture of doing things like CPR). The other reality is that a lot of money is involved with the CPR industry through training and performing the procedures. CPR is always covered by insurance. The automated CPR market alone is estimated to be worth $60 million per year, meaning this is a lucrative industry. With a great deal of federal and private money spent on training and marketing and the cost health care providers pay for biannual BLS and ACLS training, is no incentive for moving to a DNR default.

After 40 years, we are culturally invested in CPR as the default. The ethics of default CPR is a case of prioritizing beneficence over other principles. The DNR/DNAR/AND/No CPR movement has attempted to add autonomy into the mix and let people exercise their right to refuse unwanted treatment through medical orders, advance directives, out-of-hospital DNRs, and POLSTs. The problem is that the code status conversation often does not happen, does not happen well; or people do not always understand what’s involved. After all, on TV, 97% of people survive codes so most people believe that is the case in real life.

What has been lost in all of this is acknowledging the social justice aspects of CPR. Consider that CPR training is correlated with urban, white, non-Hispanic, and higher SEC. Survival rates after arrest in African-Americans is 1/3 that of whites (though in large part this is due to a lower rate of diagnosis and treatment); and whites are more likely to receive CPR from a bystander.

We should also consider rebalancing the principles of bioethics in regards to CPR. Those who do not or cannot exercise autonomy in this matter are subject to a privileging of a questionable beneficence to do something. Even when patients have exercised autonomy, the DNR choice is sometimes ignored, violating one’s self-governance. The problem is that such a view of beneficence ignores the data, that in fact most of CPR does not prolong life, but prolongs death and suffering. CPR breaks ribs, causes deep bruising, and can injure organs. Although a heart may be restarted, after 4 minutes, brain function is compromised meaning that the body that is saved is not the life that has lived in that body. Default CPR, especially on people with co-morbidities, may be ignoring nonmaleficence by causing more physical and existential harm than the statistical benefit.

Instead of forcing people to opt-out of this medical intervention, it is time to place CPR on the same footing as our other death avoiding treatments—patients must consent to it. Thus, rather than make CPR an opt-out procedure, we should make it an opt-in. Most certainly this should be the case for out-of-hospital codes where the survival rates are very low, but also for patients in the hospital for whom there is a low likelihood of survival. Temporary opt-in could be part of the informed consent for surgical procedures or anesthesia where one might code as a result of the procedure and chance of resuscitation is high, or for open-heart surgery where often the heart must be stopped from beating as part of the procedure: What’s the point of doing such surgery if you’ll never restart the heart. There may be times and places where default CPR makes sense, but not in all cases nor in all places. Thus, DNR should be our default code status with resuscitation being a procedure for which one needs consent or an out-of-hospital CPR form.