Study: Physician CPR policy not disproportionately applied
Approach can "protect seriously ill patients from harm"
There is a long-standing concern that policies empowering physicians not to offer cardiopulmonary resuscitation (CPR) are disproportionately applied to vulnerable populations, including the elderly, the disabled, and racial or ethnic minorities. However, a study conducted at Massachusetts General Hospital in Boston suggests otherwise.1
"We found no evidence to support this concern at our hospital," says Andrew Courtwright, MD, PhD, the study’s lead author. In 2007, the hospital’s clinical ethics consultation committee introduced a "Do No Harm" policy. This permits not offering CPR in some circumstances, after careful exploration of the patient’s values and medical situation.
Researchers examined case records from 2007 to 2013 and found 134 cases involving a disagreement over whether to offer CPR. They found no association between age, race, or functional status and a recommendation by the ethics committee not to offer CPR.
Other findings include the following:
Orders to withhold CPR were more common among critically ill patients and were associated with high in-hospital and 90-day mortality.
In about 25% of the cases the committee did not support the view of the clinicians, and instead recommended that CPR be offered.
In 45 of the cases (33.6%), the patient or surrogate agreed to a Do Not Resuscitate (DNR) order after an initial ethics consultation.
"We concluded that a carefully formulated hospital policy on not offering CPR, combined with consultation from a clinical ethics committee, can simultaneously maximize respect for patients’ values and protect seriously ill patients from harm at the end of their lives," says Courtwright, a physician at Massachusetts General Hospital’s Institute for Patient Care.
At times, patients and surrogates request that CPR be provided in the event of cardiopulmonary arrest even when clinicians believe it would be non-beneficial, or more harmful than beneficial. "Little is known about the use and impact of policies that allow physicians to order CPR to be withheld in these circumstances," says Courtwright.
In the researchers’ view, there are times when CPR is not a reasonable treatment option in the context of the patient’s values and illness. For example, if a patient will die imminently from widespread, untreatable, metastatic cancer and resultant multi-organ system failure, and if exploration of the patient’s values revealed no religious, cultural, or personal reason why CPR would be more beneficial than harmful in this situation, CPR would not be a reasonable treatment option.
"In this situation, we think the most ethical course is not to offer CPR and to compassionately explain why it will not be offered," Courtwright says.
As more numerous, complex, and powerful life-sustaining treatments become available, he says, it becomes increasingly important to determine carefully whether any treatment being considered for an individual patient has a reasonable chance of bringing more benefit than harm.
"These decisions can be complex," says Courtwright. "They require nuanced understanding of the patient’s values and medical situation, as well as of the capabilities of the treatment in question."
In some cases, a DNR order might be appropriate even without surrogate consent, says Thaddeus Mason Pope, JD, PhD, director of the Health Law Institute and associate professor of law at Hamline University School of Law in Saint Paul, MN. "For some patients, CPR is very unlikely to work to restore pulse or to get the patient out of the ICU [intensive care unit]," he notes.
In Texas and Virginia, the ability to write a DNR without consent is a function of state law governing disputes over "appropriate" life-sustaining treatment. "But it is unclear what the clinical thresholds are for triggering a determination that CPR would be non-beneficial or inappropriate," says Pope. "There is high variability in practice."
Even if consent is not required, consultation usually is. "Clinicians should be open and transparent about what they are doing and why," says Pope. "The family should not discover’ a DNR order on the chart." Recent cases in the U.S. and the United Kingdom have clarified the duty to consult.2
"A related issue, though separate from the issue of unilateral DNR orders, is the issue of defaults," says Pope. "This is less intrusive on individual liberty." Even if it were determined that some populations of critically ill patients should be presumptively DNR, that would still leave them opportunity to opt into CPR.
"In other words, patients would have to opt into CPR, instead of opting out of CPR," Pope says. Recent research suggests this would ensure more patients get treatment consistent with their preferences and values.3
"But changing such a well-ingrained custom is going to be very, very hard," says Pope.
Unilateral DNR orders are "only one small example of problems with the way we talk about end-of-life care," says Alana Sagin, MD, an instructor for the Palliative Care Service at Hospital of the University of Pennsylvania in Philadelphia.
"When thinking about the topic of unilateral DNR orders, we really need to look at how we present CPR to people in general," says Sagin.
Disagreements between the medical team and the family or patient involving DNR orders and the appropriateness of CPR occur for many reasons. Unrealistic expectations are one.
"Television and other media often portray an overly optimistic depiction of outcomes after CPR," says Sagin. "Patients may not have a realistic sense of outcomes."
Other reasons for disagreements include lack of prognostic awareness, unresolved family conflicts, mistrust of the medical system, and difficulty coping with a poor prognosis. "Disagreements often arise because many physicians leave the patient and family to make the decision about CPR without much guidance," adds Sagin.
Patients and families wonder, "Why would my doctor offer this if it isn’t a good idea?" "As physicians, we should make recommendations about CPR to our patients, just as we do with other medical interventions," says Sagin.
For instance, it’s known that CPR outcomes are very poor in certain populations. "With our very sick patients, we probably should be recommending against CPR more often if it is unlikely to help them meet their goals," says Sagin.
On the other hand, if the goal is prolongation of life for any length of time and with any quality of life, CPR may be appropriate. "In any case, the discussion about CPR should take place within a larger discussion about the patient’s general goals of care and prognosis," says Sagin.
With good communication and careful eliciting of patient values, the need for unilateral DNR enforcement is "very unlikely to be needed," says Sagin. "Ethics or palliative care consults can help to facilitate communication in these difficult cases, and help bring about a consensus."
- Courtwright AM, Bracket S, Cadge W, et al. Experience with a hospital policy on not offering cardiopulmonary resuscitation when believed more harmful than beneficial. Journal of Critical Care. Published online October 8, 2014. DOI: http://dx.doi.org/10.1016/j.jcrc.2014.10.003.
- Tracey v. Cambridge University Hospital  EWCA Civ 822.
- Halpern SD, Loewenstein G, Volpp KG, et al. Default options in advance directives influence how patients set goals for end-of-life care. Health Affairs 2013; 32(2):408-417.
Andrew Courtwright, MD, PhD, Institute for Patient Care, Massachusetts General Hospital, Boston, MA. Phone: (919) 699-1729. Fax: (617) 726-7557. Email: firstname.lastname@example.org.
Thaddeus Mason Pope, JD, PhD, Director, Health Law Institute, Associate Professor of Law, Hamline University School of Law, Saint Paul, MN. Phone: (651) 523-2519. Fax: (901) 202-7549. Email: email@example.com.
Alana Sagin, MD, Palliative Care Service, Hospital of the University of Pennsylvania, Philadelphia. Email: Alana.Sagin@uphs.upenn.edu.