At Switzerland’s Lausanne University Hospital, several physicians expressed concerns about how CPR was discussed with geriatric patients. A group of researchers set out to investigate the subject further by listening to recordings of these CPR discussions.1 “The results were not surprising. There were scarcity of explanations [or] explanations focusing on the steps of the procedure,” says Anca-Cristina Sterie, PhD, chair of geriatric palliative care at Lausanne University Hospital.

Initially, certain patients said their preferences were clear; in reality, these patients were not well-informed. Sterie and colleagues concluded physicians needed to assess what patients actually understood about the situation before asking about their preferences. “This discussion is an opportunity to clarify what makes sense for the patient and what doesn’t,” Sterie says.

Sterie and colleagues are using the findings to develop a training course for physicians on how to discuss CPR with their patients. “Physicians have an ethical responsibility to give information that helps patients establish their preferences. But how this information is structured is also important,” Sterie says.

Explanations might frame CPR as desirable or undesirable. Another ethical question is whether CPR should be discussed at all with patients who would not benefit from the procedure. “We argue that there is a benefit to discussing this topic with all patients, as long as it’s done with a ‘goals of care’ perspective and not as a binary issue — ‘Do you want CPR or not?’” Sterie explains.

Conversations about CPR are extremely common at the time of admission, but can “exacerbate confusion and distrust during a critical time in the patient’s care,” says Paul J. Hutchison, MD, MA, HEC-C, assistant professor of medicine and bioethics at Loyola University Chicago’s Stritch School of Medicine in Maywood, IL. Patients who are admitted often are acutely ill and highly stressed. “The uncertainty of their condition, and their complete reliance on the nurses and physicians caring for them, underscores their vulnerability,” Hutchison observes.

If patients at low risk for cardiac arrest are asked if they would want CPR, some take it as an ominous sign that their condition is dire. “They may even feel as if the clinician caring for them is not entirely committed to their improvement or survival,” Hutchison suggests. This is an obstacle to trust in the clinician-patient relationship.

There are ethical concerns if a conversation about CPR clearly is appropriate, as with critically ill patients. “Explaining CPR in a value-laden or incomplete manner may result in the selection of a code status that does not accurately reflect the patient’s authentic values and preferences,” Hutchison says.

Clinicians must be aware of their own biases regarding CPR so they can present the choice objectively. A statement such as, “If you survived, you wouldn’t have any quality of life” is an example.

“While it is important to portray the success rate of CPR as accurately as possible, this should be done in a way that avoids introducing value judgments,” Hutchison explains. Code status conversations are appropriate for many patients, but many jurisdictions do not require an explicit conversation about CPR on admission. “These conversations should be selective based on the clinician’s best judgment,” Hutchison says.

Under the federal Patient Self-Determination Act, hospitals are required to ask patients at the time of admission about advance directives and living wills. “This is a safe question to ask, even if it is asked by the registration staff or the intake nurse because it seeks to understand whether a patient has already made decisions,” Hutchison says.

If the patient has not created an advance directive, it is an opportunity to let the hospital know about it. If not, says Hutchison, “then it is up to the clinician to choose the most appropriate time to address these questions with the patient.”

In the ED, while the patient is unstable, it is not the best time for a half-hour discussion on the specifics of CPR and mechanical ventilation. Instead, a clinician might ask about existing advance directives or treatment limitations to give patients the chance to indicate previously decided preferences.

Many patients have never considered these complex therapies, and are now arriving to the hospital in a state of significant vulnerability. In cases like that, “it is appropriate to treat them as full code,” Hutchison says.

A more thoughtful conversation can occur after a day or two, once the patient’s condition stabilizes. “This approach aims to minimize patient anxiety and optimize the conditions under which CPR conversations occur,” Hutchison says. “Most importantly, it augments the trust established between clinician and patient at the time of hospital admission.”

Many patients’ perceptions of CPR are affected by what they see on TV, notes Cheyn Onarecker, MD, MA, chair of the healthcare ethics council at Trinity International University Center for Bioethics & Human Dignity in Deerfield, IL. Popular medical dramas depict successful CPR about 70% of the time.2 Family members usually believe resuscitation will be successful in most cases. “The actual numbers, though, tell a different story,” Onarecker says.

Overall, CPR for hospitalized patients resulted in survival to discharge in 22.3% of cases in 2009, according to one study of 84,625 hospitalized patients in cardiac arrest.3 Another study of 86,426 patients with in-hospital cardiac arrest found a median discharge survival rate of 34%.4

The success rate also depends on the patient’s underlying illnesses. For example, adults with cancer survive to discharge only 6% of the time; for cancer patients already in the ICU, survival to discharge decreases to only 2%.5 Additionally, some patients who were independent before CPR require nursing home admission after resuscitation, or suffer neurological deficits, kidney failure, and rib fractures.

“Patients and family members should be aware that many patients who survive suffer significant complications after resuscitation,” Onarecker says.

Discussing CPR should not scare people with worst-case examples, nor should it paint an overly optimistic outcome. Rather, the goal is to present a realistic picture of what CPR would mean for a particular patient, says Monica L. Gerrek, PhD, co-director of MetroHealth System’s Center for Biomedical Ethics in Cleveland. That can change during the hospitalization as the patient’s prognosis becomes clearer.

“Such conversations can be very challenging,” notes Gerrek, an assistant professor of bioethics at Case Western Reserve University.

Even knowing their loved one has a dismal chance of a successful resuscitation, surrogate decision-makers might feel pressure from other family members to request CPR. Surrogates may feel they are giving up on the patient otherwise. “A compassionate and empathetic approach can assist the surrogate in making the tough but appropriate decision to avoid CPR when a poor outcome is predicted,” Gerrek says.

That is not just relevant upon admission, but at any time. “If patients and surrogates aren’t given adequate information regarding what happens during an attempt at CPR or of how successful it might be in a particular patient’s case, then they can’t make an informed decision about it,” Gerrek explains.

Lack of information results in some families asking for “everything” to be done, even in situations where significant harm is likely to occur. “This puts providers in a very difficult position, of feeling obligated to perform actions that they feel violate their Hippocratic Oath and are in conflict with the principle of nonmaleficence,” Gerrek says.

For the sake of everyone involved, code status conversations should cover what medical interventions the order indicates should or should not be administered.

“The question 'Do you want everything done?' generally is not a very useful question,” Gerrek says. “Everything” does not convey what interventions a DNR order includes.

Code status conversations also should include details about under what circumstances medical interventions should or should not be administered. Once a DNR is in place, what exactly will happen? All parties should discuss this, along with details about how the provider anticipates a patient’s course will proceed and potential complicating factors.

After consenting to a DNR order, patients might experience an unexpected, unrelated medical issue. “These types of situations can cause confusion for not only the patient or surrogate, but also for medical providers tasked with treating the patient, who in many cases were not involved in discussions about the DNR,” Gerrek reports.

Learning how to engage in code status conversations is as important as learning how to perform medical procedures, according to Elizabeth Dzeng, MD, MPH, PhD, an assistant professor of medicine at the University of California, San Francisco.

“The implications of not communicating well are also as dangerous as if one hasn’t been trained in putting in a central line and there are complications due to incorrect placement,” Dzeng warns.

Some clinicians are reluctant to provide their medical opinion because of worries about compromising patient autonomy. “But everything we do in medicine needs to be taken in light of the risks and benefits and the patient’s prognosis,” Dzeng stresses.

Clinicians would not ask patients in completely neutral terms whether they want a procedure that has no chance of working or would inflict serious harm. Dzeng says it is similarly problematic when doctors say something like: “In the event your heart were to stop, would you want us to do CPR?”

“Any conversation around resuscitation status should take into account a patient’s goals and values, what is important to them in life, and what is the minimum acceptable quality of life they would be OK with,” Dzeng offers.

This should be taken into account, along with the prognosis and the likelihood of survival after CPR. “Oftentimes, a decision is made around the broader goals of care,” Dzeng adds. “The decision around CPR naturally follows from that.”


  1. Sterie AC, Jones L, Jox RJ, Truchard ER. ‘It’s not magic’: A qualitative analysis of geriatric physicians’ explanations of cardio-pulmonary resuscitation in hospital admissions. Health Expect 2021; Mar 8. doi: 10.1111/hex.13212. [Online ahead of print].
  2. Portanova J, Irvine K, Yi JY, Enguidanos S. It isn’t like this on TV: Revisiting CPR survival rates depicted on popular TV shows. Resuscitation 2015;96:148-150.
  3. Girotra S, Nallamothu B, Spertus J, et al. Trends in survival after in-hospital cardiac arrest. N Engl J Med 2012;367:1912-1920.
  4. Girotra S, Nallamothu BK, Tang Y, et al. Association of hospital-level acute resuscitation and postresuscitation survival with overall risk-standardized survival to discharge for in-hospital cardiac arrest. JAMA Netw Open 2020;3:e2010403.
  5. Reisfield G, Wallace S, Munsell M, et al. Survival in cancer patients undergoing in-hospital cardiopulmonary resuscitation: A meta-analysis. Resuscitation 2006;71:152-160.