CMDA's The Point

CPR in the Times of COVID-19

April 7, 2020

by Christine C. Toevs, MD

Cardiopulmonary resuscitation (CPR) is universally applied in cardiac standstill, unless a physician order is given to Do Not Resuscitate (DNR). CPR is the only procedure that can be performed without a physician order; a nurse cannot give aspirin, start an IV or feed a patient without an order. However, CPR is the automatic default when the heart stops. This universal application has created several ethical issues, and the current pandemic now has us questioning if CPR should be the automatic default.

Unilateral DNR is the situation where the physician orders DNR without discussion with the patient or family. The term is most often applied in clinical situations when the patient is frail, elderly and demented, and the physician places a DNR order on admission to the hospital, as part of their admission orders. Several states have outlawed this exact scenario.

CPR is a medical procedure just like any other medical procedure. CPR carries risks (yes, it does break ribs and the sternum while also puncturing lungs, liver and heart) and restoration of spontaneous circulation (ROSC) is often limited to clinical situations of primary cardiac dysfunction, usually in combination with early defibrillation (thus, the near universal AICD in public places). CPR in the hospital is applied in the ICU on patients with multisystem organ failure, for which it was not designed and is rarely successful. For the limited patients who have ROSC after in-hospital CPR, less than 11 percent survive to hospital discharge.

Application of CPR is a medical decision, and there are clinical situations where CPR is not indicated, such as life-threatening hemorrhage. If the surgeon cannot stop the hemorrhage, then CPR does not restore circulation. In our autonomy-driven world, we expect this to be a decision of the patient or family. Offering CPR in a clinical condition for which it is not indicated is giving the family a choice that is not a choice. There is no choice to be made about whether to apply chest compressions if the medical condition cannot be treated with chest compressions. Asking the family if they want CPR in these situations creates moral distress, leading them to think they have made the decision to withhold treatment that resulted in the death of their loved one.

Given the near universal application of CPR, and the expectation that modern medicine can conquer death, the ethics literature is now debating an “opt-in” default for CPR, rather than an “opt-out.” The ICU literature (such as this article and this article) specifically talks about these clinical situations and how to have a discussion with a family about code status. Goals of care discussion should include code status in the ICU, and all admissions to the ICU should have a goal of care discussion. All choices are not morally or ethically equal, and to present all medical treatments in the ICU as a menu of choices disrespects the patient and offers false hope about the ability of medicine to restore the patient to health. The current recommendations in the ICU regarding code discussions are to frame the discussion about what it will really take to meet the patient’s goals (time on ventilator; other interventions needed, including dialysis and surgery; and expected time in recovery, including disposition, need for skilled nursing facility, etc.). This is a more reasoned and commonsense approach than to ask, “You want your loved one to live, don’t you?” Discussing whether CPR will work in their loved one’s clinical situation is part of goals of care. Families are often relieved when the physician makes recommendations about what will work and won’t work, rather than the physician offering every medical treatment that is available. Physicians complain that “families want everything done,” but physicians create this problem by not discussing ICU treatments in the context of goals of care, while reassuring families that we are trying to save a life.

In the case of COVID-19, there has been some discussion in the media about the application of universal DNR and the unfairness of the physician making the decision without speaking to the family. In the non-medical world, there is a lack of understanding of CPR, as TV shows portray a success rate of greater than 70 percent. This TV success can yield a perception that, without CPR, the physician is just letting the patient die. For a patient infected with COVID-19, CPR should only be performed with personal protection equipment (PPE) in place. A high-risk exposure occurs when a healthcare professional not wearing PPE provides CPR to a patient with COVID-19. Current CDC recommendations are that the healthcare professional be excluded from work for 14 days after the exposure. Therefore, to perform CPR in a patient with COVID-19, the resuscitation code team must put on PPE. This takes time, and time matters in CPR because “every second counts.” The delay in providing CPR decreases its already low effectiveness. For the healthcare professional to provide CPR without PPE is reckless, and it has serious consequences to all patients, as they will be out of work for at least 14 days. One can see how universal application of DNR in all patients with COVID-19 is being considered in many hospitals.

I propose a more nuanced approach. Rather than a blanket DNR for all patients with COVID-19, a discussion with patients when they are diagnosed, and with the family when they are in the ICU, is crucial. Information about the low likelihood of success of CPR and the reasons why will give the family an understanding that there are limitations to medical treatments, but medicine is not “giving up” on their loved one. The physician can explain that CPR will not be offered in this situation, and letting the family agree or “opt-out” is reasonable. CPR is a medical treatment, and the limited prospect for success must be considered, given the need for PPE and the inherent delay in this process.


  1. Avatar Robert Cranston on April 23, 2020 at 11:43 am

    Great article, Christine. I agree completely.


  2. Avatar Jerry Wittingen on May 11, 2020 at 7:40 pm

    This is a reasonable, and wise, approach.

  3. Avatar sandy christiansen on May 11, 2020 at 8:15 pm

    Thank you for addressing this nuanced issue.

  4. Avatar Eugene Shively MD FACS on May 12, 2020 at 8:53 am

    Good approach Christine. It rural Kentucky we frequently have a problem with CPR . Futile care often leads to unnecessary suffering. Lots physicians do not discuss CPR and DRN orders with their patients. This often leads to a disaster.

  5. Avatar Mary Cairns M.D. on May 12, 2020 at 1:19 pm

    Thank you for this article, glad to see CMDA publishing on this issue.

Leave a Comment