Do Not Resuscitate during Anesthesia and Urgent Procedures

Do-Not-Resuscitate Orders during Anesthesia and Urgent Procedures

NOTE: The UW Dept. of Bioethics & Humanities is in the process of updating all Ethics in Medicine articles for attentiveness to the issues of equity, diversity, and inclusion.  Please check back soon for updates!


Gail Van Norman, MD, Clinical Associate Professor, UW Dept. of Anesthesiology, Adjunct Faculty, UW Dept. of Bioethics & Humanities

Core ClerkshipsFamily MedicineInternal MedicineAnesthesiologySurgery

Related TopicsDo Not Resuscitate Orders I Futility I Advance Care Planning and Advance Directives 


Topics addressed:

  • When should CPR be administered?
  • What should we do with DNR orders in the OR?
  • When should CPR be administered?
  • When can CPR be withheld?
  • Is the outcome from CPR different in the OR than on the medical ward?
  • Does the cause of arrest matter?
  • Why do we agree to do surgery on patients with DNR orders?
  • What should be included in a discussion of DNR orders in the OR with the patient or patient's surrogates?
  • What about emergencies?

It is common to have patients present for surgery, for whom a "Do-Not-Resuscitate" order is written in their chart. Physicians and patients alike suffer from misconceptions about the potential benefits and harms of resuscitation in the operating room (OR), and even the definition itself of resuscitation in the OR requires clarification prior to surgery. Because the OR environment presents patients with a situation in which cardiopulmonary resuscitation (CPR) carries significantly different risks and benefits than on the medical ward, re-discussion of the implications of the DNR order are necessary. (For a discussion of DNR Orders in other medical settings, see main topic: Do Not Resuscitate Orders.)

When should CPR be administered?

Anesthesiologists and surgeons may be reluctant to accept DNR orders on patients undergoing surgery because of the scope of medical practice which constitutes "normal care" in the surgical environment. Many surgeries require intubation and mechanical control of respiration for the duration of surgery, to protect the airway from aspiration, prevent anesthetic-induced hypoventilation, to allow the administration of paralytic agents to prevent muscle contraction during surgery, and for many other reasons. Yet intubation and ventilatory assistance are mainstays of CPR.

It is inaccurate to call anesthesia "ongoing resuscitation," yet the administration of anesthetic agents frequently causes initial changes in the autonomic nervous system, such that hypotension, tachycardia, bradycardia, and temporary cardiac dysrhythmias can result. It is not rare to administer vasopressive medications and antiarrhythmic agents during the course of "normal" anesthetic management. Such medications are often also considered a vital part of effective administration of CPR.

Finally, both invasive and noninvasive technology in the OR permits easy application of therapeutic measures which might seem extreme on the medical ward, such as external or transvenous pacing and defibrillation. Under most other circumstances, such measures would fall almost exclusively within the realm of CPR.

So where do we draw the line between "normal" and "usual" procedures in the operating room, and "extraordinary" procedures which constitute CPR? Many authorities have suggested that the application of chest compressions is an unusual enough occurrence even in the OR setting, that it provides an medical and ethical boundary between CPR and normal anesthetic care.

What should we do with DNR orders in the OR?

In 1992, the American Society of Anesthesiologists (ASA) produced Guidelines for the Ethical Care of Patients with Do Not Resuscitate Orders, and Other Orders Limiting Care in the Operating Room. Out of respect for patient autonomy, or the right of competent, adult patients, to determine their own medical care, no specific definition of CPR was provided in the document. Instead, it requires a discussion with the patient to define medical procedures under anesthesia to which the patient would consent. Shortly after the ASA adopted its guidelines, the American College of Surgeons, and the Association of Operating Room Nurses (AORN) adopted guidelines which drew directly from the ASA's document.

All acknowledged that patients do not check their rights to self-determination at the OR doors, that policies automatically suspending or upholding DNR orders in the OR were ethically suspect, and that rediscussion of the DNR order should occur, whenever possible, prior to undertaking surgery and anesthesia.

When should CPR be administered?

CPR should be administered in the absence of a valid physician's order to withhold it.

When can CPR be withheld?

As with CPR on the medical wards, two general situations arise in which CPR can be withheld in the operating room:

  1. When CPR is judged to be of no benefit to the patient. (See the main topic page, Do-Not-Resuscitate Orders.)
  2. When a patient with intact decision-making capacity (or in the case of those without decision making capacity, an appropriate surrogate decision-maker) indicates that they do no want CPR, even if the need arises.

Is the outcome from CPR different in the OR than on the medical ward?

CPR in the OR carries a very different medical prognosis than CPR administered in other hospital areas. While only 4 to 14% of all patients resuscitated in the hospital survive to discharge, 50 to 80% of patients resuscitated in the OR return to their former level of functioning. This is probably due to several differences between arrest on the medical ward and arrest in the OR. In the OR, the event of arrest is always witnessed, and the proximate cause usually known, allowing rapid, effective intervention which is directed toward the specific cause of arrest. Also, causes of arrest in the OR are often reversible effects of anesthesia or hemorrhage, and not usually due primarily to the patient's underlying disease. Patient and physicians may require correction of the perception that CPR in the OR is just as futile as CPR on the general medical ward.

Does the cause of arrest matter?

Because arrests in the OR are often due to hemorrhage or medication effects, rather than the patient's underlying disease, physicians may feel that their actions "caused" the arrest, and they are ethically obliged to resuscitate the patient, even if the patient has clearly expressed wishes to the contrary. But competent patients, or their appropriate surrogates, have the right to refuse medical procedures and care, even if the care is to counteract the effects of previous medical intervention.

Why do we agree to do surgery on patients with DNR orders?

Many types of surgery provide palliative benefits to patients who either will not survive long-term, or who do not wish resuscitation in the OR. A patient with an esophageal obstruction from cancer might benefit from gastrostomy placement through reduced pain and improved nutritional status, yet not want CPR if cardiac arrest happens in the OR. Requiring such a patient to suspend their DNR orders to be a candidate for surgery uses their discomfort, pain, and desire to benefit from surgery to coerce them into accepting medical care (CPR) they do not want. Patient refusal of some medical therapy, such as CPR, does not ethically justify physicians denying them other medical therapy, such as surgery, that might benefit them.

What should be included in a discussion of DNR orders in the OR with the patient or patient's surrogates?

As discussed above, surgery and anesthesia may require the administration of medical therapies, which under other circumstances might be considered resuscitation. It is an ongoing source of discussion about what constitutes appropriate information and choices to present to patients about to undergo surgery who have DNR orders on their charts.

Since the goal of medical therapy is to provide meaningfulbenefits to the patient, discussion of DNR orders in the OR should center around the patient's goals for surgical therapy. Patients may have fears of "ending up a vegetable" on a ventilator after surgery, for example. In those cases, discussion should center around the positive prognosis for patients who have CPR in the OR, together with reassurance that the patient's stated wishes in their advanced directive regarding ventilatory support would be followed postoperatively after anesthetic effects are ruled out as a cause of ventilatory depression. Most authorities now agree that a "smorgasbord" or checklist "yes-or-no" approach to the various procedures in the operating room is confusing and counterproductive to the purpose of DNR discussions.

Anesthesiologists in particular need to be aware that studies indicate that many patients with DNR orders in their charts (up to 46%) may be unaware that the order exists, even when they are competent. While policies at the University of Washington Medical Center require documentation of discussion of DNR orders with the patient or appropriate surrogates, anesthesiologists and surgeons should nevertheless approach the patient about to undergo surgery with sensitivity to the fact that they may be unaware of their DNR order. If this proves to be the case, a full discussion of the DNR order should be undertaken prior to proceeding.

What about emergencies?

Even in emergencies, physicians have an ethical obligation to recognize and respect patient autonomy. Whenever possible, physicians should obtain input from the patient, or when the patient is incapacitated, from appropriate surrogates, regarding the status of the patient's DNR orders in the OR. In the absence of such input, consensus should be reached among the caregivers about the medical benefits or futility of CPR. In any case, medical care of the patient in the absence of patient input should be directed toward realizing, to the best of the physician's ability and knowledge, the patient's goals.

  • Jackson SH, Van Norman G.  Goals and values directed approach to informed consent in the “DNR” patient presenting for surgery—more demanding for the anesthesiologist?  Anesthesiology 1999; 990:901-5
  • Troug RD, Waisel DB, Burns JP.  DNR in the OR:  A goal-directed approach.  Anesthesiology 1999; 990:289-95
  • Waisel DB, Burns JP, Johnsons JA, Hardart GE, Troug RD.  Guidelines for perioperative do-not-resuscitate policies.  Journal of Clinical Anesthesia.  2002; 14:467-73
  • Ross LF.  Do not resuscitate orders and iatrogenic arrest during dialysis:  should “no” mean “No”?  Seminars in Dialysis.  2003; 16:395-8
  • Stevenson LW.  Rites and responsibility fo rresuscitations in heart failure:  tread gently on the thin places.  Circulation. 1998; 98:619-22
  • Tulsky JA, Chesney MA, Lo B.  How do medical residents discuss resuscitation with patients?   Journal of General Internal Medicine. 1995; 10:436-41