Every day, physicians grapple with wrenching life-and-death decisions. They often must weigh many factors (some conflicting), such as patient wishes, laws, right and wrong, one's sense of duty… There are enough shades of gray such that the best course of action may not be obvious—and often there is no optimal course of action at all.

Medscape's Ethics Report premiered in 2010. Here, in our fourth report, more than 7500 physicians from more than 25 specialties shared their often clashing views on issues about life, death, and patient suffering, with some notable shifts in attitude over the years.

A growing number of US physicians now endorse the idea of physician-assisted dying (or physician-assisted suicide) for terminally ill patients. In 2010, less than half of US doctors said they would consider helping terminally ill patients end their lives. That share has increased significantly, and today well over half of the physicians we surveyed endorse the idea of physician-assisted dying. Of note, the change is offset by a decline in the percentage of doctors opposing physician-assisted dying, not by a decline in those who say, "It depends."

"I believe that terminally ill patients should have the right to die on their own terms and in peace," writes a cardiac surgeon.

"It is shocking that this is a controversial topic. Medicine is a secular institution and we must honor the wishes of our patients," says a family doctor.

"Absolutely not," writes an internist. "It is fraught with the risks for abuse of vulnerable populations such as the elderly, poor, and disabled, and destroys trust between physicians and patients by making the physician judge, jury, and executioner."

"A very slippery slope," says a gastroenterologist.

Male and female physicians favor physician-assisted dying equally, although more men than women oppose the idea (31% vs 26%).

In recent years, some patients with terminal illnesses—such as Brittany Maynard in 2014—have publicly expressed their wish to use physician-assisted dying to end their lives rather than continue with prolonged suffering and treatment. These highly publicized cases have made an impression on the thinking of many people.

"The movement toward physician-assisted dying, among both men and women, has been toward more public support, so this isn't a surprise," says Arthur Caplan, a professor of bioethics and director of the Division of Medical Ethics at New York University Langone Medical Center. "Physicians are in line with the rest of society on this."

The issue of assisted death for non–terminally ill patients is even more controversial. Just over a quarter of respondents (28%) say yes to assisted death for those with incurable suffering. One respondent referred disparagingly to a recently passed law in Belgium allowing non–terminally ill patients to legally obtain assisted-suicide.

Approximately another quarter (27%) of respondents say it may be ethical to help suffering patients end their lives, depending on the circumstances. They say the situation would depend on the patient's psychological condition, competence, and whether all pain control options had been exhausted.

"Experience in Europe has shown that unhappy sufferers seek assisted suicide even with illnesses that are not life-threatening! Blind persons have sought assisted suicide! So have depressed people. The slope is not just slippery, it's dangerous!" writes an orthopedist.

"The point of hospice is to relieve pain and suffering in the face of incurable disease," writes an emergency physician. "What other wheel are you trying to invent?"

"Absolutely not—that would be murder," says an internist.

"Assisted suicide should be legal. A physician's highest calling is to relieve pain and suffering, not to prolong life or to cure illness," writes an anesthesiologist.

In recent years, the vaccination/antivaccination controversy has flared up, and more physicians have had to deal with patients who refuse vaccines. The United States has seen outbreaks of measles and whooping cough, raising concerns that herd immunity is being compromised by parents who do not vaccinate their children.

"No. I feel that the risk/reward of vaccines indicates that they should be done, and if you choose the side of risk, I would prefer not to treat you," says an internist.

"They have the right to make their own decision about their health," says a family physician.

"The message has gotten across that vaccination is a public duty and not just a personal choice," says NYU's Caplan. But Caplan questions the ethics of dismissing such patients. He notes that doctors are missing an opportunity to educate vaccination holdouts. "Besides, if you fired everyone who didn't do what you wanted, you'd have a very small practice."

Pediatricians are the most likely to say that they would turn away families that reject vaccinations. Nearly half (49%) of pediatricians say they would or might refuse to treat such a family, compared with 40% of family physicians and 36% of internists.

"As an MD, I can educate and attempt to guide my patients, but ultimately it is their decision on what recommendations to take," writes an internist.

"Their poor decisions are bad for their kids, bad for society, and bad for their doctors. Also, I have better things to do with my time than try to talk people out of their delusions," says a pediatrician.

Several pediatricians note that the question involves the nature of the physician-patient relationship. "If a family doesn't trust my advice on this important topic, then we are not in the right relationship together," writes one.

How do you get experience on a new procedure without practicing it? This conundrum evokes markedly different views from physicians. Half say they would tell their patients and the other half say they might not. Many doctors say it would depend on how complicated or risky the procedure was or what level of backup they had.

"If I were not experienced, I would not do the procedure!" writes a pediatrician.

"I would absolutely tell the patient. Patients deserve to know the experience of their physicians. But I would explain the other similar experiences, skills, and training that qualify me for the new procedure."

"I may not say specifically that I have never done it before, but I may say that someone needs to be here to observe, to make sure everything is done properly. I would not deny my inexperience if asked, but I would not volunteer that information."

When patients have been diagnosed with a terminal illness, many doctors feel that "addiction is a moot point." Still, doctors are cautious, noting that they would "document, document, document." Many say they would refer the patient to a pain management specialist or palliative care.

"A terminally ill patient's pain should be treated adequately—PERIOD. There is no excuse for not sufficiently treating a terminally ill patient's pain," writes an internist.

"As long as I justify it and can defend my behavior, I would help my patient," says an oncologist.

"I couldn't care less whether a terminally ill patient with intractable pain becomes addicted to controlled substances. Relief of pain and suffering is paramount, and regulators be damned!" adds a neurologist.

"Yes," writes an emergency physician. "Becoming addicted doesn't matter for these patients, but if I have my license revoked then I can't help anyone."

In light of the opioid epidemic and growing scrutiny from medical boards, physicians are concerned about finding the right way to deal with pain management. Nearly half (47%) of physicians surveyed say they would or might undertreat a patient's pain for fear of regulatory repercussions or concerns that the patient might become addicted.

Many doctors note that they are worried less about the DEA or sanctions but are troubled about the danger of addiction.

"Pain medication addiction is a growing problem in the country. I believe that it is our responsibility to be part of the solution," writes a cardiac surgeon.

"The patient's potential to become addicted always factors heavily in my treatment of their pain," says a family physician.

"I don't think it's the end of the world if somebody still has a little bit of pain, to reduce the risk of overtreating with narcotics and creating addiction," writes an emergency physician.

Desperate patients often want to try anything that they believe might save their lives. On the other hand, it's been publicized that Steve Jobs ultimately felt that he could have lived longer had he started standard medical treatments earlier.

Nearly half of physician respondents say that terminally ill patients should be allowed to try any treatment they want, but they must be prepared to foot the cost. Others fear that patients will put themselves at physical and financial risk by seeking non-evidence-based treatments.

"A competent patient can do whatever they want. But society should not pay for it," writes an anesthesiologist.

"Allowing treatment and expecting insurers or hospices to pay for any treatment a patient may request are two different issues. I may advise against certain unproven or potentially dangerous treatments, but generally I do not stand in the way," says a family physician.

"It's ridiculous that the patient should be exploited by false hopes for medicines that don't work," writes an internist.

It's not a surprise that many patients want to hang on to life as long as possible or that their family members want to keep their loved ones alive. These desires can make ethical decisions even harder for physicians.

"I'm in geriatrics; I see this all the time. Why give futile care? The patient does not have quality of life," says one physician.

"Futile care is inappropriate and does not meet professional responsibility standards for proper ethical conduct," writes a maternal-fetal medicine specialist.

Many doctors maintain that they would provide treatment for a limited time so that family members can say goodbye or so that the patient can achieve a near-term goal. As one emergency physician notes, "If it is so that a family can say goodbye, then it is worth it."

"I wouldn't recommend it but might not be able to refuse the family's request," writes a neuropsychiatrist.

Oncologists, who recognize the toxicity of the therapies they provide and frequently face this situation, are increasingly determined not to provide therapy that they consider to be futile. More than half (56%) of oncologists say they would not provide futile treatment, a significant increase from 2014 (47%). Only 22% said they would continue treatment.

"That is doing harm," writes one radiation oncologist. Another notes, "I am very realistic with patients and not afraid to talk about death." A third notes that the definition of "futile" will vary from patient to patient. "Prolonging survival by weeks to months is meaningful for some, while minimizing the financial burden to the family is meaningful to others."

Physicians practicing emergency medicine, by contrast, are far more likely to soldier on—more than 3 in 10 (31%) say they would recommend such treatment, up from 27% in 2014.

Even though most doctors say they would not provide futile treatment, when asked whether patients are being withdrawn from life support too soon, many echo a sentiment provided by an emergency care physician. "Actually, quite the opposite. We continue futile care far too often."

"The vast majority of patients are kept on life support long after it is clear that there will be no benefit," writes a neurologist.

"Actually, in my experience there is a lot of moral anguish among our ICU RNs because aggressive care is carried on excessively," comments a family physician.

But some physicians disagree. "Sometimes I feel that the fight is given up too quickly," says an emergency physician.

"This varies. There are hospitals that have leaned toward withdrawing patients after only a few weeks on life support, when other places have maintained them for months at a time before considering removing them," says a cardiologist.

In 2014, 28% of doctors said they would abide by a family's wishes and stop treating a patient, even if they felt that the patient had a chance of recovering. This year, 34% say they would do so, with many of the respondents saying that the question assumes that the patient is not competent and that "there is no living will."

"There's a shift toward individual choice becoming a stronger value in American society," says Art Caplan, bioethicist. "There's a feeling of, 'You listen to me or you listen to my family, but don't try to tell me what to do just because you are an expert.'"

"As long as the family understands the situation, I would have to follow their wishes. This is one sphere in which a physician can attempt to intrude but is only able to stand on the outside, hoping that the decision-maker acts with all knowledge of the situation," says an emergency physician.

Others say they would comply with the family's wishes "as long as the advanced directive is clear on who can make medical decisions on the patient's behalf" or "as long as the family has power of attorney and there are no directives that indicate otherwise."

But some physicians are more cautious. "It depends on the motive of the family. If I felt that there really was some underlying devious motive, then I'd continue treatment and get appropriate people involved, but I'd also need to consider the patient," says a neurologist.

Physicians' thinking on this subject has changed somewhat since 2010.

As with the general population, abortion remains a divisive issue among doctors. A May 2015 Gallup poll showed that 55% of Americans feel that abortion should be illegal.

"A physician's personal beliefs have no place in the workplace, and under no circumstances should they ever present a barrier to a patient accessing needed appropriate medical care. No exceptions," says a family physician.

More than half of ob/gyns say they would perform the procedure, despite their personal beliefs. One notes, "I am an advocate for the patient's desires, not my own."

An internist notes, "Abortion is the same as allowing capital punishment. It is equally abhorrent and against the Hippocratic Oath."

"I would not do something that's against my moral code," says a family physician.

More than half (53%) of doctors say they would or might devote limited resources to a younger patient even if it meant withholding care from an older one, down from 62% in 2014. Those who would do so note that age can be a factor in triaging patients to determine the more effective use of resources.

"Limited resources, such as organs, should be given to those with the highest likelihood of having favorable outcomes and long-term life," writes one pulmonologist. Says another, "I would devote resources where they would be most likely to be effective."

"Age per se would not be my criterion for allocating resources. I would care more about the patient's level of function, their prognosis, etc.," says a neurologist.

In society as a whole, "there is a strong view of rationing that favors the young over the old, but in medicine that sentiment is less strong," says Caplan.

Doctors know that honesty is essential to patient trust and that patients need to know their diagnoses so they can put their affairs in order and make end-of-life decisions.

That said, they are sensitive to how and when they deliver the news as well as the impact it might have on the patient's continued treatment. Those who say, "It depends" often note that the situation would depend on the patient's age and mental status and on how much the patient wanted to be told.

"All patients should have the opportunity to understand their conditions," says an orthopedist.

"If the patient were depressed at the time and the diagnosis would be overwhelming, I'd delay announcing the new diagnosis until the patient was less depressed and more likely to be able to manage the news," says a pulmonologist.

"This would be an inexcusable breach of their patient autonomy, informed consent, and any semblance of trust," says an emergency physician.

Three-quarters (75%) of physicians say they would not withhold information from a competent patient.

"I work for the patient and must always respect the patient's rights," writes an oncologist, voicing a frequently expressed sentiment.

Still, on rare occasions—particularly when dealing with children or due to cultural sensitivities—doctors may withhold information.

"If they convince me that the patient's culture and emotional stability are such that this information would harm him/her," writes a pulmonologist.

"Rarely, in specific cultures, that request has been granted," says a surgeon.

"Depending on their cultural values and if the patient agrees that all information should go through the family," says a critical care doctor.

Most doctors say that there is no wiggle room on informed consent. The vast majority of respondents (78%, up from 76% in 2014) say they would not sugarcoat or downplay risks even if they felt strongly that a patient would benefit from a particular procedure or treatment.

"All physicians do this; they just might not be aware of it. This is called framing, and depending on how I frame a treatment, I can virtually ensure that a patient would or would not choose to undergo it," an emergency physician says.

"Informed consent is just that: It is consent of the patient based on correct and adequate information provided," says an ob/gyn.

"Risk versus benefit is still the patient's call," says a urologist.

This question stipulates that the family has given consent. Just over half of the survey respondents felt that a "brain-dead" patient's organs should be harvested—even in the absence of a donor card, as long as the family gives consent. "As long as power of attorney rules are applied and the POA understands the patient's prior wishes," says a cardiologist.

"Next of kin should make the decision that they think the patient would want," writes a pediatrician.

"I would rely on the family members to know what the patient would desire," says an ob/gyn.

Some physicians latched on to the relevance of the donor card and provided markedly different views:

"A dead person has rights that were implied by lack of donor card," counters another. "Autonomy does not expire." Others noted that harvesting a patient's organs might violate his or her religious convictions.

Free-market economics should not apply to organs, say the vast majority (77%) of doctors who participated in our survey. For some, the question is simply one of fairness: Poor people would be unable to afford organs and would more likely feel pressure to sell them. For others, it represents a potential regulatory nightmare: "How many organs could you sell before your own health was at risk?" and "What mental health evaluations for these people would be put in place?" they ask. "Harvesting or distributing organs should be closely regulated."

"It would disproportionately affect the poor and might lead to 'forgetting' important medical history," says a wound care specialist.

"A market for organs would be worrisome in terms of protecting patients from exploitation and pressure. Organs should be donated," writes a radiation oncologist.

However, a minority of physicians feel that consenting and competent adults should be allowed to make decisions for themselves. "A patient should be able to sell a kidney or half of their liver, a lung, etc., if they would like to," writes an orthopedist. "It should be allowed as long as it's done safely and not in a back alley," comments another.