Columbia University Medical Center

When a Patient Refuses Treatment, What Should Doctors Do?

It’s a surprisingly common dilemma in medicine: A hospital patient who lacks capacity because of dementia, mental illness, or other conditions refuses a diagnostic test or treatment that the doctors feel is in the patient’s best interests.

Should the physician deliver treatment against the patient’s wishes? Remarkably, there is little guidance for physicians, even though the situation arises frequently. To address this gap, Kenneth Prager, MD, and Jonah Rubin, MD’16, physicians and ethicists at Columbia University Irving Medical Center and NewYork-Presbyterian, recently created a series of questions to guide others faced with these ethically complex situations.

Q: With no published guidelines available until now, how do physicians decide whether to go ahead and treat a patient who lacks decisional capacity and refuses medical intervention?

KP: This is one of the most frequent reasons for requesting an ethics consult at many hospitals. When ethics consultants are not available, physicians often end up making a unilateral decision with no guidance. Depending on the physicians’ clinical judgment, awareness of the ethical issues, and knowledge of the law, their decisions may or may not be ethically appropriate.

Part of the issue is there is nothing in the medical literature to help physicians deal with this challenging situation. There are guidelines to help determine if a patient has decisional capacity. And there are fairly clear policies and laws concerning the ethics and legality of delivering psychiatric care to patients who refuse it. But there is nothing out there to help health care professionals approach the problem of delivering medical treatment against the wishes of patients who lack decisional capacity.

Q: When is it obvious that a physician should ignore the patient’s wishes and deliver treatment? And when is it not so obvious?

KP: A simple example of when treatment over a patient’s objection would be appropriate is if a psychotic patient who had a life-threatening, easily treatable infection was refusing antibiotics for irrational reasons. Treatment would save the patient’s life without posing significant risk to the patient.

When treatment is not likely to be as effective and might cause serious complications, or when the risk to the patient is not as clear, the ethical issues are more complex.

JR: A less obvious example concerns a patient who is blinded by cataracts and wants to have his sight restored but refuses to have cataract surgery. Given the patient’s wish to see again and the low risk and high success rate of cataract surgery many doctors would agree that it is appropriate to treat over objection. But some doctors might conclude that it’s inappropriate to ignore the patient’s refusal because the patient was already blind and the procedure would just reverse a harm that’s already been done—not prevent one from happening.

Our questions are designed to help physicians navigate this gray area.

When treatment is not likely to be as effective and might cause serious complications…the ethical issues are more complex.

Q: With patients in the “gray area,” how do your guidelines help physicians make decisions?

JR: We identified seven core questions that provide an ethical framework for making such decisions. The questions are mostly intuitive and address several dimensions. They ask what anyone would want to know before undergoing a medical procedure or intervention.

One key component of our guide is that it provides structure. These discussions can frequently become disorganized, and it’s easy to be swayed by the last point or lose sight of all the issues. So we developed an algorithm that walks the user step-wise through the core issues. After going through these questions, the user can derive a comprehensive ethical conclusion based on all, not just some, of the key components we’ve identified.

The first few questions consider the imminence and severity of the harm expected to occur by doing nothing as well as the risks, benefits, and likelihood of a successful outcome with the proposed intervention. Other questions consider the psychosocial aspects of this decision—how will the patient feel about being coerced into treatment? What is the patient’s reason for refusing treatment? The last question concerns the logistics of treating over objection: Will the patient be able to comply with treatment, such as taking multiple medications on a daily basis or undergoing frequent kidney dialysis?

Q: How did you arrive at these questions?

KP: Our guide is based on notes from thousands of ethics consults that I conducted over the course of 25 years. Over time, I developed a checklist of questions to ask in each case that I felt would be helpful in arriving at an ethically acceptable approach to the problem. We think that our guide will prevent physicians from glossing over some of these questions and will give them the structure that had been lacking in these deliberations.

It also became clear that the issue of logistics—is it actually possible to treat the patient who resists?—could make the question of treatment moot. As Dr. Rubin stated, one cannot force three times weekly dialysis sessions on a resistant patient even if it means that the patient will die without the treatment.

Q: How can you be confident that your guide will help physicians in these situations?

JR: Our guide doesn’t suggest that there is a right or wrong answer in every case. Two groups may use the same questions and end up with different conclusions, as in the example of the blind patient who refuses to have cataract surgery. This demonstrates that our approach—as titled—is a guide to ensure that physicians address all of the relevant points.

KP: The only way to test the questions is to gather qualitative input from people who use them. We hope that physicians will put these guidelines to the test and share their feedback with us so that we can modify them as needed.

Jonah Rubin, MD, is a resident in internal medicine at NewYork-Presbyterian/Columbia University Irving Medical Center.


Kenneth Prager, MD, is a professor of medicine at Columbia University Vagelos College of Physicians and Surgeons, chair of the Medical Ethics Committee at NewYork-Presbyterian/Columbia University Irving Medical Center, and director of clinical ethics at CUIMC.


Their article, published online in Mayo Clinic Proceedings, is titled “Guide to Considering Non-Psychiatric Medical Intervention Over Objection for the Patient without Decisional Capacity.” The authors report no financial or other conflicts of interest.