Doctors Urged to Separate CPR from Supportive Care in End-of-Life Discussions

Doing CPR on dying patients with uncorrectable conditions is futile, says cardiologist John Mandrola.

Published on Mar. 10, 2026

Dr. John Mandrola, a cardiac electrophysiologist, argues that doctors should separate discussions about do-not-resuscitate (DNR) orders from conversations about providing supportive end-of-life care. He says that cardiopulmonary resuscitation (CPR) was designed to be used in specific situations like heart attacks, not as a default treatment for patients nearing the end of life due to incurable conditions. Mandrola believes framing DNR orders this way could increase patient acceptance of them and reduce the need for unilateral decision-making by doctors.

Why it matters

Mandrola's perspective challenges the common practice of automatically making patients "full code" unless they explicitly request a DNR order. He believes this default setting often leads to futile and potentially harmful CPR being performed on dying patients, when the focus should instead be on providing compassionate supportive care. Changing the conversation around DNR could improve end-of-life care.

The details

Mandrola notes that as a specialist, he is often asked to see chronically ill patients with multiple complex conditions whose cardiac symptoms or signs, like arrhythmias, are a manifestation of their severe illness. Yet these patients are often designated as "full code," meaning CPR and defibrillation will be performed if they go into cardiac arrest. Mandrola says that in these cases, nearly everyone involved knows the CPR would be futile or even harmful, but no one has had the difficult conversation to change the patient's code status.

  • In the 1960s, surgeon James Jude and colleagues first described the use of closed-chest cardiac massage, which later became known as CPR.
  • A year later, Jude's team reported on a larger series of 138 episodes of cardiac arrest, showing that CPR could restore cardiac action in 78% of cases and allow more than half of those patients to return to their pre-arrest cognitive status.

The players

John Mandrola

A cardiac electrophysiologist who practices in Louisville, Kentucky and writes and podcasts for Medscape. He espouses a conservative approach to medical practice and participates in clinical research.

James Jude

A surgeon who, along with colleagues, first described the use of closed-chest cardiac massage in the 1960s, which later became known as CPR.

Tammy Pegg

A cardiologist from New Zealand who emailed Mandrola with the idea that CPR should be framed as a rescue intervention for acute, fixable conditions, not a default treatment for dying patients.

Got photos? Submit your photos here. ›

What they’re saying

“If CPR can be remembered as Jude described, as only bridging chest compressions, and if a do-not-attempt-CPR order does not restrict access to other evidence-based interventions deployed appropriately in the peri-arrest period, then maybe this would result in increased patient acceptance of these orders and reduce the requirements for unilateral decision-making.”

— Tammy Pegg, Cardiologist (Email to John Mandrola)

“Most [cardiac arrest] events in the 21st-century hospital arise from serious illness with deterioration and activation of early warning scores.”

— Tammy Pegg, Cardiologist (Email to John Mandrola)

What’s next

Mandrola believes that if clinicians can successfully separate discussions about supportive end-of-life care from decisions about CPR, it could lead to more patients accepting DNR orders and reduce the need for doctors to make unilateral decisions about resuscitation.

The takeaway

Mandrola's perspective challenges the common practice of automatically making patients "full code" unless they explicitly request a DNR order. He believes framing DNR orders as a way to avoid futile and potentially harmful CPR, while still providing compassionate supportive care, could improve end-of-life discussions and decision-making.