Journal Logo


Cardiopulmonary Resuscitation in Coronavirus Disease 2019: Rebalancing Risk, Reward, and Autonomy*

Wydo, Salina M. MD1; Peterson, Lars-Kristofer Nelson MD2,3

Author Information
doi: 10.1097/CCM.0000000000004754
  • Free
  • Editor's Choice
  • COVID-19

To say coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has changed the relationship between the general public and critical care medicine grossly understates the effects clinicians have seen in their daily lives. The professionals who work in ICUs globally saw news articles and commentaries about ventilator availability, allocation of scarce resources, and prone positioning in acute respiratory distress syndrome. Out of necessity, ICUs restricted families and visitors, but the world got to peek inside through the news. In March 2020, when cases were rising sharply in New York City and Northern New Jersey, news outlets published that hospitals considered implementing universal do not attempt resuscitation (DNAR) orders for COVID-19 patients (1,2). Critical care clinicians know that DNAR orders only become active in the setting of a cardiac arrest. However, to the general public, the difference between full code and do not resuscitate often means, “will you fight to save my life or not?”

In the United States, admitting clinical teams ask the patient or their substituted decision-maker about resuscitation wishes. In the absence of a patient declining full cardiopulmonary resuscitation (CPR) measures or an advanced directive/physician order of life-sustaining treatment, the default is “full-code,” indicating the patient accepts CPR and intubation with mechanical ventilation in the event of an inhospital cardiac arrest (IHCA). The purpose of resuscitation in both out-of-hospital cardiac arrest (OHCA) and IHCA is to prolong life, allow time to reverse the disease processes causing the physiologic decline, and provide a neurologically acceptable outcome. However, the gulf between expected results and reality is vast.

The public grossly overestimates the success of CPR. In 2000, 1,831 participants in community settings in the United States estimated CPR’s overall survival rate to be 54% (3). With the given specific clinical scenarios, they estimated CPR survival to be 79% if a 75-year old otherwise healthy patient experienced a cardiac arrest (3). A more recent survey of 1,000 participants (noncritically ill patients and their families) evaluated both an OHCA and an IHCA scenario (4). In a 54-year old who had suffered a heart attack at home and required CPR, respondents estimated a 72% chance of survival and predicted a 65% chance of neurologic recovery. For a postoperative 80-year-old patient, survey participants indicated a 57% chance of survival and a 53% chance of full neurologic recovery (4). Observational data of OHCA and IHCA survivorship contrast against these optimistic predictions. In the United States, 2016 estimates of adult OHCA and IHCA survivals with functional neurologic outcomes were 9% and 15%, respectively (5). A disconnect between perception and clinical experience can create conflict between families and treating clinical teams when discussing limitations on resuscitative care. Tension rises when the perceived patient-oriented benefit of resuscitative measures, particularly CPR, approaches zero.

In this issue of Critical Care Medicine, Shah et al (6) share their observations of IHCA of COVID-19 patients at a three-hospital community healthcare system in rural southwest Georgia. The authors included all patients admitted between March 2, 2020, and August 26, 2020, who were COVID-19-positive (confirmed by polymerase chain reaction) and underwent CPR. Casting a broad net, the authors gathered information on the patients’ demographics, social determinants of health, comorbidities, clinical data, and hospital course. To better quantify the patients’ long-term prognosis, the authors used the Charlson Comorbidity Index (CCI) score. Derived in 1987 as a means to quantify 10-year mortality risk in longitudinal studies (7), the CCI has since been validated in recent cohorts of patients (8,9). During the specified time, the study sites admitted 1,094 patients with COVID-19. Of these, 149 died (a mortality rate of 13.6%), and 63 (5.8%) experienced a cardiac arrest, prompting the performance of advanced cardiac life support (ACLS). The 63 patients presented relatively late in their disease course with a duration of symptoms of 11 days. The majority (47, 74.6%) had severe pneumonia at admission. Mechanical ventilation supported 51 of the patients (81%) before their IHCA. Expectedly, these patients demonstrated significant elevations of inflammatory markers. A large majority of the arrests (84.1%) occurred in the ICU with CPR initiated quickly after the recognition of the arrest (< 1 min in the ICU and < 2 min on the floor/emergency department). Severe hypoxia (Pao2/Fio2 ratio < 100) was present in 63.5% of patients, with another 20.6% being moderately hypoxic (Pao2/Fio2, 100-200). Although many of the IHCA patients had significant comorbidities, 28.6% had a CCI of less than 2, predicting a greater than 90% 10-year mortality rate. Although 29% of patients achieved return of spontaneous circulation after their IHCA, there were no survivors to discharge after IHCA. In comparison, over March to August of 2019, the same healthcare system reported 144 IHCAs with a 34.7% survival to discharge rate. Neurologic status was not described in the prior year’s IHCA survivors.

This new contribution adds to a conversation in the literature from two other recent publications (10,11). A cohort from Wuhan, China, was cared for at a tertiary hospital designated to receive COVID-19 patients. Over the 40-day study period, the hospital admitted 761 patients with features of severe COVID-19. Of these patients, 151 experienced a cardiac arrest and 136 records were complete for analysis. Only one patient survived over 30 days with a favorable neurologic outcome (0.7%) (10). The patients who had a cardiac arrest in Michigan, United States, did not fare better. Beaumont Hospital is a large tertiary-care center that admitted 1,309 COVID-19 patients between March 15, 2020, and April 3, 2020. Of these, 60 had a cardiac arrest, with full data available for 54 patients. None of these patients survived despite a historical IHCA survival to discharge rate of 25% (11).

There are significant limitations in the article by Shah et al (6), as well as the prior publications. First, the studies are all retrospective analyses at single centers. Second, there is no ability to infer causality. Do patients with COVID-19 who experience an IHCA do worse due to an idiosyncratic effect of the virus and its attendant complications, or is it a factor of how the healthcare system treats the patients? In the current study and the prior publications, time to CPR initiation was described as relatively short, but there may be uncaptured delays in donning personal protective equipment (PPE). The required PPE may hinder key elements of ACLS including communication, chest compression rate, and depth, and worsen resuscitation team fatigue.

In the era of COVID-19, healthcare workers and employees of essential services confront a potentially fatal disease with no known consistently successful treatment (other than supportive care), and prevention relies on PPE and physical distance. For those working in the prehospital, emergency department, inpatient, and ICU settings, distancing is not always possible. That is particularly true during intubation and CPR. SARS-CoV-2 spreads through aerosolized droplets and contact with contaminated services (12,13). Leading resuscitation organizations stated that both CPR and intubation (often required during CPR) are aerosolizing procedures (14,15). The World Health Organization recommends using N95 masks or their equivalent to protect healthcare workers from exposure during aerosolizing procedures (16,17). Unfortunately, during the initial phases of the pandemic response, PPE scarcity resulted in healthcare worker exposure, infection, and mortality.

At the beginning of the COVID-19 pandemic, data about cardiac arrest outcomes in patients with COVID-19 were not available to guide decision-making in resuscitation. Clinicians on the ground faced the unenviable choice between providing immediate resuscitation with possibly inadequate PPE or delaying resuscitation until appropriate PPE could be donned, risking worse patient outcomes. Institutions needed to balance their obligations to the community, admitted patients, and their staff. Before COVID-19, CPR posed a relatively small occupational risk, usually confined to needlestick injuries or body fluid exposures. Although the outcome of an IHCA was always questionable, the low risk to clinicians combined with the possibility of patient recovery makes empiric resuscitation reasonable. However, the data around COVID-19 and IHCA would indicate the patient-oriented benefit approaches zero, whereas the bodily risk to clinicians is significantly increased. A report of simulated CPR sessions demonstrated that N95 masks might not afford adequate protection against particles similar to those carrying SARS-CoV-2 (17). Infection of inpatient clinicians degrades the institution’s capacity to care for their community and increases the burden of care on noninfected staff, not to mention the impact to the individual and their family. Unlike discussion of the provision of CPR during crisis standards of care, where the resources needed for a postcardiac arrest patient may not be available, the published data force us to consider if performing CPR on a patient hospitalized with COVID-19-related complications is justified in normal conditions.

As with all medical decision-making, choosing to provide CPR in IHCA should be undertaken with our best understanding of the risk-benefit ratio. The fact that this decision occurs during a worldwide pandemic, with the real possibility of resource exhaustion and the legitimate risk of transmission to healthcare providers, only complicates the risk analysis. These factors demand we ask when do we reach the point at which the provision of CPR cannot be justified due to considerations unrelated to the individual patient? Arriving at an answer to this question involves considering the ethical implications, societal values and expectations, and medical science and outcomes.

Despite our best efforts (in the before-COVID era), the medical profession has yet to identify a reliable set of criteria to a priori establish the patient in whom CPR is futile. Short of the patient having a valid DNAR, or irreversible signs of death (e.g., rigor mortis, dependent lividity, decapitation), or having a physiologically irreversible cause of cardiac arrest (e.g., fulminant septic or cardiogenic shock), the American Heart Association recommends administering CPR for all cardiac arrests (18). With current investigations showing a survival from IHCA in patients with COVID pneumonia of 0–0.7%, when is not offering CPR acceptable? A complete discussion of the ethics surrounding DNAR orders is beyond the scope of our commentary; suffice it to say that we have a responsibility not only to respect and uphold patient autonomy but also to consider the risk to others and the fair and just allocation of limited resources.

Generally speaking, we have a responsibility to care for the patient in extremis. Does declining to offer CPR violate this duty to treat? What are the limits of that duty? Where does duty end and supererogation, going above-and-beyond, begin? Sheather and Chisholm (19) argue in a 2020 opinion that the physician’s duties exist not only to the patient in front of them but also to their other patients, community, partners, and spouses. Should a clinician fall ill, the effects reach farther than the care of one patient. Additionally, equitable and just care for the patient assumes the physician has the training and proper resources for the job. What if the healthcare team does not have adequate tools or training to ensure their safety? What if the data regarding the intervention indicate a lack of efficacy despite the increased risk? These factors appear to weaken (but not eliminate) the duty to treat during cardiac arrests.

Ideally, the ethical issues are avoided if the patient/surrogate and physician discuss the goals of care and reach the same conclusion: a patient elected DNAR. Despite the Patient Self-Determination Act of 1991, only 36.7% of American adults have advance directives (20). Having detailed, compassionate, and transparent conversations regarding goals of care can be limited in the emergency setting due to acuity and competing immediate needs. If circumstances afford the luxury of having these conversations, they should happen at admission and be updated frequently throughout the hospital stay, especially if the patient’s condition worsens. The focus should be on identifying the patient’s overall goals of care, rather than on merely achieving a DNAR status. Having this conversation while the patient retains capacity is important.

Because we cannot expect goals of care or a change in code status to be discussed at the time of an IHCA, it is reasonable to ask healthcare systems to assume an “informed assent” model. In an informed assent model, the physician still has the responsibility to communicate context, risks, and benefits to the patient or their surrogate, but the patient/surrogate defers to the physician’s recommendations. This system allows the patient/surrogate to remain involved and preserve autonomy but also respects the physician’s expertise and best clinical judgment (21).

Previously reported perspectives regarding CPR in patients with COVID-19 resulted in three recommendations: acknowledge resource constraints when discussing goals of care and DNAR status, forgo CPR in certain circumstances (particularly when CPR will not address the underlying cause of the arrest), and ensure the safety of healthcare personnel (22). If we are justified in forgoing CPR in certain circumstances, healthcare systems should put forward policies and procedures with the intent to preserve patient autonomy as far as is reasonable, to minimize the risks to the healthcare team, steward scarce resources, and empower treating clinicians to use the best available evidence to decide about offering CPR in the moment. Care should be taken to apply any policies and procedures fairly and equitably (avoiding discrimination by race, ethnicity, sex or gender identity, age, ability, and economic status) and communicated clearly and transparently with the community served (23). The need for attention toward any impact on exiting disparities is all the more critical, because COVID-19 disproportionally affects minority communities in incidence, mortality, and economic loss, magnifying preexisting structural inequities (24).

Regardless of the resolution, patients and surrogates should be clearly told that decisions to withhold life-sustaining treatments, particularly acute resuscitation during a cardiac arrest, is not synonymous with the decision to withhold care. Patients are always entitled to skilled and compassionate care at all phases of their illness, regardless of COVID-19 or DNAR status. COVID-19 does not change the guiding ethical principles that inform our practice. However, emerging science calls for a rebalancing of those principles to give the best care to the patient, the teams supporting them, and their communities.


1. Cha AE. Hospitals consider universal do-not-resuscitate orders for coronavirus patients. The Washington Post. 2020. Available at: Accessed October 15, 2020
2. Sargeant K. Hospitals consider do-not-resuscitate orders for coronavirus patients to protect doctors, nurses. 2020. Available at: Accessed October 15, 2020
3. Marco CA, Larkin GL. Cardiopulmonary resuscitation: Knowledge and opinions among the U.S. general public. State of the science-fiction. Resuscitation. 2008; 79:490–498
4. Ouellette L, Puro A, Weatherhead J, et al. Public knowledge and perceptions about cardiopulmonary resuscitation (CPR): Results of a multicenter survey. Am J Emerg Medicine. 2018; 36:1900–1901
5. Benjamin EJ, Virani SS, Callaway CW, et al. Heart disease and stroke statistics—2018 update: A report from the American Heart Association. Circulation. 2018; 137:116–426
6. Shah P, Smith H, Olarewaju A, et al. Is Cardiopulmonary Resuscitation Futile in Coronavirus Disease 2019 Patients Experiencing In-Hospital Cardiac Arrest? Crit Care Med. 2021; 49:201–208
7. Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis. 1987; 40:373–383
8. Quan H, Li B, Couris CM, et al. Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol. 2011; 173:676–682
9. Radovanovic D, Seifert B, Urban P, et al. Validity of Charlson comorbidity index in patients hospitalised with acute coronary syndrome. Insights from the nationwide AMIS Plus registry 2002–2012. Heart. 2014; 100:288
10. Shao F, Sun P, Tang Z. Cardiopulmonary resuscitation of inpatients with severe COVID-19 pneumonia: The Wuhan experience. Resuscitation. 2020; 152:95–96
11. Thapa SB, Kakar TS, Mayer C, et al. Clinical outcomes of in-hospital cardiac arrest in COVID-19. JAMA Intern Med. 2020; 181:e204796
12. van Doremalen N, Bushmaker T, Morris DH, et al. Aerosol and surface stability of SARS-CoV-2 as compared with SARS-CoV-1. N Engl J Med. 2020; 382:1564–1567
13. Ong SWX, Tan YK, Chia PY, et al. Air, surface environmental, and personal protective equipment contamination by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) from a symptomatic patient. JAMA. 2020; 323:1610–1612
14. Nolan JP, Monsieurs KG, Bossaert L, et al. European Resuscitation Council COVID-Guideline Writing Groups. European Resuscitation Council COVID-19 guidelines executive summary. Resuscitation. 2020; 153:45–55
15. Edelson DP, Sasson C, Chan PS, et al. Interim guidance for basic and advanced life support in adults, children, and neonates with suspected or confirmed COVID-19: From the Emergency Cardiovascular Care Committee and Get With the Guidelines®-Resuscitation Adult and Pediatric Task Forces of the American Heart Association in Collaboration with the American Academy of Pediatrics, American Association for Respiratory Care, American College of Emergency Physicians, The Society of Critical Care Anesthesiologists, and American Society of Anesthesiologists: Supporting Organizations: American Association of Critical Care Nurses and National EMS Physicians. Circulation. 2020; 141:e933–e943
16. World Health Organization. Infection Prevention and Control During Health Care When Coronavirus Disease (COVID-19) Is Suspected or Confirmed. 2020. Available at: Accessed October 21, 2020
17. Hwang SY, Yoon H, Yoon A, et al. N95 filtering facepiece respirators do not reliably afford respiratory protection during chest compression: A simulation study. Am J Emerg Med. 2020; 38:12–17
18. Part 2: Ethical aspects of CPR and ECC Resuscitation. 2000; 46:17–27
19. Sheather J, Chisholm J. Duty to treat: Where do the limits lie? BMJ. 2020; 369:m2150
20. Yadav KN, Gabler NB, Cooney E, et al. Approximately one in three US adults completes any type of advance directive for end-of-life care. Health Aff (Millwood). 2017; 36:1244–1251
21. Curtis JR, Burt RA. Point: The ethics of unilateral “do not resuscitate” orders: The role of “informed assent”. Chest. 2007; 132:748–751; discussion 755
22. Kramer DB, Lo B, Dickert NW. CPR in the Covid-19 era—An ethical framework. N Engl J Med. 2020; 383:e6
23. Curtis JR, Kross EK, Stapleton RD. The importance of addressing advance care planning and decisions about do-not-resuscitate orders during novel coronavirus 2019 (COVID-19). JAMA. 2020; 323:1771–1772
24. Bambra C, Riordan R, Ford J, et al. The COVID-19 pandemic and health inequalities. J Epidemiol Community Health. 2020; 74:964–968
Copyright © 2021 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.