Should Your Doctor Lie to You?

            The nation is riveted by President Trump's illness: whether we hate Trump or love him, we want to know how he is faring with Covid-19. We want to understand what this disease looks like in an elderly man with at least one chronic health condition. Unfortunately, what we have been told by the physicians involved in Trump's care has been marred by commissions, distortions, and downright lies.

            There is a long history of presidents wishing to mislead the public about their health and of their physicians colluding in the deception—Woodrow Wilson’s stroke was concealed, Franklin Roosevelt’s high blood pressure and heart problems were downplayed, and John F. Kennedy’s Addison’s disease and chronic back pain were not fully disclosed. But however objectionable we may find this public lack of transparency, President Trump’s personal physician has claimed a different reason for being less than forthcoming. He asserted that he had understated the seriousness of his patient’s condition because he “didn’t want to give any information that might steer the illness in another direction.” That is, Dr. Sean Conley didn’t want his patient to know that his low oxygen levels and high fever were worrisome, so he lied about his condition. Telling the truth, he was asserting, could harm his patient. But is that true?

            Truth-telling in medicine has been the subject of extensive ethical analysis and of clinical study. The bottom line is that while doctors used to routinely lie to their patients in the belief that they were protecting them, for the last 50 years the standard of care has been to keep patients informed to whatever extent they wish and, based on their accurate understanding of their situation, to engage them in decision-making about treatment.

            The change in practice occurred in the sixties and seventies: in 1961, when a questionnaire was administered to oncologists asking them if they told their patients that they had cancer, fully 90 percent of them said they did not. When the study was repeated in 1979, 97 percent of them said they would tell patients their diagnosis. The earlier view was based on the paternalistic belief that physicians always knew what was best for their patients and on the conviction that if patients knew they were seriously ill they would become depressed and possibly even suicidal. Between 1961, when the first study was conducted, and 1979, when the second study was carried out, western biomedical ethics came into its own as a field.

            Physicians and medical ethicists increasingly recognized that there was often no single optimal course of treatment: several different possible approaches might be possible, each with its own likelihood of benefit and each with its own risks; which approach was “right” for a given patient depended on that person’s preferences and values. One person with a particular type of cancer might wish to undergo treatment with chemotherapy that had a high probability of resulting in serious side effects in exchange for a small chance of life-prolongation; another individual with the same disease might opt for a different treatment that was less likely to cause severe side effects but that offered a smaller chance of life-prolongation. Whenever the choice of treatment depended on values as well as technical expertise, the patient had to be included in the decision-making along with the physician. The principle of beneficence, or doing good, and the principle of non-maleficence, or not doing harm, co-existed with the principle of autonomy, or the right of patient self-determination. 

            Choosing the right treatment for a particular patient, in many cases, required that the patient know the truth about his diagnosis. Without knowing the facts, he couldn’t possibly participate in a conversation with his physician about treatment options. Moreover, growing evidence indicated that when patients are engaged in their own health care, they do not become morbidly depressed or overtly suicidal; on the contrary, health outcomes improve. 

            The regrettable example set by the president’s personal physician notwithstanding, you should expect honesty from your doctor. Yes, you should expect that your doctor will have the communications skills necessary to impart bad news sensitively. Trust is at the core of the doctor-patient relationship, and trust cannot be built on a lie.

The Dignity Dilemma

            As soon as I walked into Mrs. L’s room, I knew something was wrong. She was sitting quietly in her wheelchair, her head bowed; her back was towards me, towards the door, towards the world of the nursing home which, for the past year, had been her world and where I was a volunteer. I walked over to her and greeted her, gently resting my hand on her shoulder, and saw her face was contorted into a grimace. I asked her what was the matter and she said, her voice almost inaudible, that she had to go to the bathroom.

            That seemed like an easy enough problem to solve. I went to the nurse’s station, a few steps from Mrs. L’s room, and found two nursing assistants chatting so I let them know that Mrs. L. needed help going to the bathroom. “Her roommate is in the bathroom; she has to wait,” I was told. I returned to Mrs. L’s room, gave her an update, and tried to distract her. I asked her about her plans for Mother’s Day, I commented on the trees that were flowering outside. Ten minutes passed. Fifteen minutes passed. The roommate was still in their shared bathroom. Mrs. L. began to moan softly.

            I went back to the nurse’s station. This time I found a nurse and asked her if she was in charge.  She hedged, saying she was taking care of some of the residents. As it turned out, Mrs. L. was one of “hers.” I explained the situation, indicating that Mrs. L. had been waiting at least fifteen minutes—I wasn’t sure how long she had been waiting before my arrival. “Well,” the nurse explained, “she has to wait. Her bathroom has a raised toilet seat adjusted for her, so she has to use her own bathroom.”

            No doubt a raised toilet seat is helpful. But nursing home residents do sometimes leave the nursing home. They visit their relatives and use their bathrooms. They go to restaurants and use the public bathrooms. A raised toilet seat is desirable but not essential. 

            I asked about alternatives, given that Mrs. L’s bathroom of choice was occupied. There were, after all, many other bathrooms on the same floor: there was a staff bathroom, there was a public bathroom and, if necessary, there were the other residents’ bathrooms. Mrs. L’s section of the nursing home must have had at least ten bathrooms. And her “unit” was one of four on the same floor, each of which had another ten bathrooms.

            No, the nurse explained. For “infection control” reasons, Mrs. L could not use any of those other 39 bathrooms.

            I was incredulous. Mrs. L. was some sort of Typhoid Mary? A danger to the public health? 

            There are, to be sure, situations in which hospital patients or nursing home residents harbor highly infectious agents and are on “contact precautions.” But Mrs. L. had no such infection. She was not subject to such measures.

            I tried one last time. Surely, I indicated, there must be some sort of backup plan if the desired bathroom was unavailable. No, I was informed. The nursing home had no backup plan. Mrs. L. would have to wait.

            Really? How long a wait was acceptable? Half an hour? An hour? What if the plumbing was out of order? 

            I waited with Mrs. L. After twenty-five minutes (plus whatever time she had waited before my arrival), the roommate emerged from the bathroom. I informed the nursing assistants, who were still sitting at the nurse’s station, that the bathroom was now available, and Mrs. L. was taken to the bathroom. 

            To Mrs. L., this treatment felt like punishment. It felt like an arbitrary exercise of control by the powerful over the powerless. To me, it sounded like adherence to rules intended to promote the health and safety of the residents but designed for the convenience of staff, rules that were enforced without creativity, flexibility, or compassion.

            Treating frail, vulnerable older nursing home residents with dignity is not a new challenge Since Mary Mendelson wrote the muckraking “Tender Loving Greed” in 1974 and Robert Butler received a Pulitzer Prize for “Why Survive? Being Old in America,” published in 1976, nursing homes have been criticized for failing to treat residents with dignity and respect. But we have come a long way since the 1970s. The Nursing Home Reform Act (OBRA ’87) promulgated a “Bill of Rights” for nursing home residents that guaranteed the right to privacy and to being treated with dignity. Dr. Bill Thomas, a geriatric physician, introduced the Eden Alternative in 1991: he brought plants and pets into nursing homes to convert them from sterile, cold, institutions to something more closely resembling a home. A few years later, a consortium of nursing homes  calling itself the Pioneer Network was established to transform nursing homes through “culture change.” Their goal, as the network’s website puts it, is to help “care providers to transition away from a medical, institutional model of elder care to one that is life affirming, satisfying, humane and meaningful.” It goes on to say that the “Pioneer Network advocates for a culture of aging in which individual voices are heard and individual choices are respected.”

            Implementing this model is not easy and some of the early ideas about how to go about transforming care were problematic. One idea, for example, was the “universal worker.” In a person’s home, responsibilities for caretaking are often shared by family members. Everyone can bring Mom a magazine or get her a cup of tea or escort her to the bathroom: there is no clearly demarcated division of labor. The hope was that replicating this model in the nursing home would allow greater flexibility—and shorter waits for residents—but nursing homes aren’t homes and a division of labor is often necessary because not all employees have the same skills. Translating the model into practice is especially challenging when residents have cognitive impairment and cannot make autonomous choices. 

            Nursing homes also face organizational challenges and regulatory challenges. I understand that. The Centers for Medicare and Medicaid Services, which dictates the terms of the state surveys that assess quality of care in nursing homes, also understands that it’s not easy to address quality of life as well as safety. But the resident interviews that are now a mandated part of quality assessment include questions about dignity and respect.

            Mahatma Ghandi is widely quoted as saying that “a nation’s greatness is measured by how it treats its weakest members.” I would elaborate on this to say that a nursing home’s quality is measured by whether it treats its most dependent, vulnerable residents with respect and dignity. By this standard, many nursing homes today, including some that pride themselves on their care, have a long way to go. They would do well to consult some of the training guides assembled by the Pioneer Network and to follow the model of British nursing homes, that have launched campaigns to systematically promote dignity--including bathroom use--in nursing homes.

Are there Truly No Atheists in Foxholes?

A nagging concern about advance care planning is the possibility that patients may change their minds when faced with imminent death. That’s not a problem if the person is still able to make decisions and really can change his mind. But for the majority of older people whose thinking is impaired at precisely the time that a potentially life or death decision about their medical care must be made, it is very much a problem. By etching their wishes in stone, is an advance directive too inflexible? Does it make it hard for surrogate decision-makers to change their minds on the patient's behalf?

We know that patients do change their minds over time as they adjust to their new (and presumably more  limited) level of functioning. The most extreme case that’s been described comes from younger patients:  individuals with quadriplegia after an accident are initially despondent; after they adapt to their situation, they still aren’t happy with their status, they would still much rather not have quadriplegia, but many if not most of them come to recognize that they are still human beings with emotions, feelings, and abilities. Whatever suicidal tendencies they had at the onset of their disability have typically abated. 

Older people also change their minds over time about what they are willing to live with, but a substantial proportion shift from wanting their medical care to focus on prolonging life at any cost to wanting their medical care to focus on maintaining their level of functioning. And older patients who were recently hospitalized for a serious illness such as heart failure or a flare of COPD, when asked about what matters most to them, reportedly say they would not want life-prolonging treatment if it left them with severe cognitive impairment. But what about the very moment that they are in the hospital, acutely ill, and confronting their own mortality? Do they also say they would rather die than be left with dementia or profound disability? A research letter in JAMA Internal Medicine finds that many of them do.

It’s a small study (only 180 patients) conducted at a single academic medical center (in Philadelphia). They were all, by definition, able to engage in a discussion (actually a structured interview) about their preferences. They were all seriously ill, suffering from advanced cancer, severe heart failure or advanced lung disease. They were asked about 10 different possible forms of functional debility: bowel and bladder incontinence, relying on a breathing machine, unable to get out of bed, constant confusion, dependence on a feeding tube, needing constant care, living in a nursing home, being homebound, in constant, moderate pain, or confined to a wheelchair. While they differed in their ratings of various health states compared to death, large numbers rated each of the possible conditions as worse than death, and the first six were viewed as the same or worse than death by over half of all patients.

Now these patients were not actually confused and many of them weren’t incontinent, on a breathing machine, unable to get out of bed, etc., even temporarily. If they were, they might, like the young person with quadriplegia, adapt. But they were in the hospital with a serious enough illness that they could die. Presumably they knew that (though this is not clear from the article). And what is striking is that none of them had any limitations on life-sustaining treatment documented in their medical record. It’s unlikely anybody had spoken to them about which of the treatments they were getting or they might get could in fact leave them with exactly the condition they indicated in the interview they wanted most to avoid.

The possibility that there really are no atheists in foxholes has to be taken seriously. Maybe most people will change their minds about what treatments are tolerable to them if they are sufficiently close to death. The best we can do is to ask people about preferences who do understand something both about the state they may find themselves in and the nature of the treatment being proposed. That means we should do the asking when people already are old, sick, and debilitated, not when they are young and healthy.

Home Alone

Last week, I made the politically incorrect suggestion that patients don’t always know what it is that they don’t know. Continuing my contrarian streak, I’m going to suggest this week that home isn’t always what it’s cracked up to be, at least not for frail older people.

A recent study provides valuable information about the homebound population in the US. Examining cross-sectional data from the National Health and Aging Trends Study (collected in 2011), the authors were able to identify a nationally representative sample of just over 7500 non-institutionalized Medicare beneficiaries who were 65 or older. What they found is that 5.6% of the elderly, community-dwelling population were completely or mostly homebound. Extrapolating to the entire country, that comes to nearly 2 million people, where “completely homebound” refers to someone who never went out in the last month and “mostly homebound” refers to someone who went out rarely (once a week or less) in the last month. Another 5.3 million older people are “semi-homebound,” which means they go out at least sometimes (twice a week) but not by themselves (somewhat over one million people) or only with difficulty (the other 4 million people).

Interestingly, the study observes that the semi-homebound individuals are very similar to the completely homebound individuals in terms of disease burden and functional capacity. This suggests that social support is as important as medical conditions in determining the course of a person’s life. Environmental and psychological factors may also play a role. It is entirely reasonable to conclude--as the authors do--that providing adaptations such as stair lifts or grab bars and assistive devices (walkers and canes) could transform a person from being homebound to being semi-homebound or even from being semi-homebound to non-homebound. Other services such as home-based primary care could also make a difference.

The authors of the new study are also right to conclude that the information they have uncovered could “inform improvements in clinical and social services” for frail older people.  But it is equally reasonable to ask whether at least some of those dependent, disabled individuals might not be better off in a well-designed facility that offered social stimulation as well as physical assistance and medical care. I made a similar argument years ago in my paper, “Long-term care options for the frail elderly.” I made a similar argument just a few years ago in the paper “No place like the hospital,” where I suggested that the widely quoted claim that 80% of people want to die at home but only one-third do reflects what people say they want when they aren’t dying. When they actually get sick and are facing imminent death, at least some of the people who were previously so sure they wanted to die at home discover that they feel more secure and better cared for when they have round-the-clock nursing care and some of the technological interventions of the hospital. When hospitals were terrible at pain management and resolutely focused on maximally invasive medical treatment, they weren’t good places to die. As the availability of palliative care in the hospital has increased and the quality of pain management has improved, hospitals aren’t always such an awful place to spend one’s final days. 

The same is true for nursing homes. When the quality of nursing homes was abysmal, when residents sat tied up in chairs, lined up along dark, dingy corridors, and when they developed pressure ulcers because they were neglected by poorly trained staff, of course nursing homes were to be avoided. As standards were raised and resident-centered care took root in nursing homes, the quality of care improved. Sometimes nursing home care is superior to home care.

It’s still true that nobody wants to live in a nursing home. It’s also true that the majority of people don’t want to die in a hospital. But what many of those people really mean when they say they don’t want to live in a nursing home is that they don’t want to be very disabled and dependent. What they really mean when they say they don’t want to die in a hospital is often that they don’t want to die. 

It's all too easy--and too commonplace--to confuse an undesirable state of being (disabled, dying) with a site of care (the nursing home, the hospital). But preventing disability and death is one thing; figuring out where to go when frailty develops or death is imminent is something else entirely.

The Deciders

Shared decision-making has become something of a sacred cow in medicine, even though few physicians actually practice it. There is certainly evidence that patient participation in discussions about their health care and patient engagement in self-care lead to better outcomes, as well as to greater patient satisfaction. Promoting patient autonomy requires that patients play a role in shaping their fate. So some kind of patient involvement is decidedly a good thing. But shared decision-making, I argue in an article just published in the Journal of Medical Ethics, needs to be re-engineered for it to work in practice.

To be sure, not everyone means exactly the same thing by shared decision-making. But most of the definitions look like this one, offered by a leading proponent and expounder of the model: shared decision-making is an approach in which the patient receives information about available treatment options (including their risks and benefits), the clinician and the patient consider each one in light of the patient’s situation, goals, and preferences, and the two parties jointly select the best option.

The focus of all this deliberation is the selection of a treatment. Which treatment to provide is what the doctor needs to know. And the reason for involving patients in the decision-making is that, unlike in selecting which antibiotic to use in the treatment of pneumonia, which is strictly a technical decision, deciding whether to use chemotherapy or radiation for the treatment of cancer, or whether to use fourth line chemotherapy or hospice for treatment of cancer when it progresses to a very advanced stage, depends on the patient’s values. But what I suggest in my essay is that respecting patient autonomy requires eliciting those values. Once the doctor understands the patient’s goals, once he or she knows what is most important to the patient, that information constitutes data that goes into the decision about treatment along with other data involving outcomes, side effects, and probabilities. To ask the patient to draw conclusions about which treatment is best, rather than to have the doctor make a recommendation based on a whole raft of information that includes the patient’s input about goals and values, makes no more sense than providing a patient with data about antibiotics and then expecting a patient to select which antibiotic he should take for his pneumonia.

I argue that despite several decades of work seeking to overcome the barriers to shared decision-making—barriers such as cognitive biases, innumeracy, and health illiteracy—and despite evidence that sophisticated decision aids can help patients, most doctors and patients don’t like the conventional approach to shared decision-making and don’t use it. Even medical ethicists who believe strongly in honoring patient autonomy and who have traditionally advocated shared decision-making balk when they themselves or their family members develop cancer and the physician tries to implement shared decision-making. 

The approach I advocate doesn't go back to the older paternalistic model in which physicians decree and patients obey; rather, it reformulates the way shared decision-making takes place by suggesting that what needs to be shared is the process of determining the patient’s goals of care, not the process of deciding how to translate those goals, along with other highly technical information, into a treatment decision.