For more than thirty years, the ever-increasing number of life-prolonging, death-defying interventions and the ease with which they are deployed, especially among the elderly, has led to the well-known national complaint about too much technology and not enough personal control near the end of life. Although palliative care and hospice services have become more widely used over time, and the “less is more” call is loud among some practitioners and observers, the “do everything” ICU and acute care norm remains in place as the default, affecting millions of Americans.

The language of medicine contributes to the culture of complaint. That language, this article confirms, is stubbornly resistant to change.^{Footnote 1} Physicians, mostly historically reluctant to initiate conversation about impending death, remain evasive about it. They are not typically trained to talk about it, and they are not comfortable with prognostication. Rather, they are trained to describe the array of standard tools they can offer for pieces of an overall problem (such as sepsis) to reduce infection, return the patient to baseline, etcetera. “Treatability statements”— like We can treat your condition; Your infections are treatable; and We can offer some treatments—convey a positive message to patients and families not only about treatment effectiveness, but also about overall health and survival. Patients and families interpret such statements to mean cure, enhanced quality of life, or the absence of disease.

Although physicians shy away from outright prognostication (about time to death, remission, cure, etcetera), ironically, the plethora of diagnostic tests now available produces prognostic information. Those tests allow one to foresee the next treatments required by standard practice. The more diagnostic information is available, the more the physician needs to anticipate treatment courses and those that will follow. Because there are almost always next treatments to contemplate, physicians feel obligated to mention them and, more precisely, to offer them. Discussion of those offerings is what takes up so much time in the clinic, or at the bedside.

Generally unspoken, however, is the question of whether physicians would advise the use of any mentioned or offered treatments. Although they may talk about the big picture in the same conversation with patients (for example, end-stage cancer, or that the patient may not survive till discharge), patients and families listen to, remember, and dwell on the treatments offered. I have observed the same phenomenon of selective listening and hearing over a twenty-year period, in my own studies of physician statements and end of life treatments.^{Footnote 2}

Physicians’ topics of conversation are not heard the way they are intended. Rather, the pragmatic implications of their language conveys to patients and families not only hope for extended life and return to some kind of normal, but also the sense that recovery will occur and that impending death will be replaced by open-ended life.

For patients, discussion of treatment possibilities always implies good news. Why? They mostly do not have knowledge of the goals and limits of specific treatments, progression of the underlying disease, the medico-ethical imperative to “do something,” or specialists’ areas of expertise and mandate to treat. Furthermore, they do not know what to want in terms of specific interventions (many of which they have never before encountered), except to get better, or die and not suffer.

The authors sum up this incommensurability: “For the physicians, treatability statements emphasize what the physician can do; for patients, they emphasize how the patient will do.” Physicians discuss what’s available and what they can do right now. Patients and families listen with the goal of anticipating the future. The article takes up this conundrum with a concise analysis of the pragmatics of language. We learn, for example, that when physicians say they can treat, they simply mean that they can do something, intervene somewhere. They conceive of treatment “as a tool to accomplish particular clinical goals, defined in specific, technical terms.” We learn that specialists refer to treatment in different ways. Intensivists distinguish between treating and supporting a return to baseline, whereas oncologists distinguish between treating, that is, delaying decline and death, and curing. Patients and families are not aware of these subtleties.

Regardless of these differences in intended meaning among specialists, physicians’ words, the authors emphasize, are “underinformative.” This is key. The probability of or implications for treatment effectiveness—whether the treatment burden will outweigh the benefits, whether the underlying disease can be cured, or whether decline to death remains inevitable, for example—tend to remain unvoiced. As a result, patients and families overwhelmingly agree to whatever therapies physicians mention. My own research among older patients facing life-threatening disease found similarly. A 90-year old cancer patient spoke for many patients I interviewed when she aptly remarked to me, “If you don’t do something, it’s disease progression all the way to death,”^{Footnote 3} as though any intervention when one has advanced metastatic disease or other critical illness would inevitably thwart decline, postpone death, and improve life. Sometimes patients demand treatments even when physicians state that they do not recommend them, the assumption being that if physicians do something patients will get better and perhaps live longer. Incommensurability between physicians’ intended meaning and patients’ and families’ understandings of physicians’ words continues, and the authors illustrate how it is built into the standard language of medicine.

This article contributes new, important evidence to the large and still growing body of literature about miscommunication between doctors and patients. Its strength is the revelation of the recalcitrance and continued omnipotence of treatability statements as default medical practice when life is at stake.

The origins of the dominance and force of treatability language lie in a cluster of factors that came together beginning in the 1970s, when more therapies became available for doctors to use, and for patients to consider. For example, the mechanical ventilator became standard equipment in the ICU, where its use now routinely includes delaying the death of very old, frail people with end-stage disease. Specialization and subspecialization led to more knowledge and opinions about what can and should be treated in cases of critical illness. Biotechnology and the mushrooming drug and device industries created more diagnostic tests leading to more interventions for more specific ailments.

By the late 1980s the fact of doing something was firmly entrenched and taken for granted in physician reasoning and hospital routines, because so much could be done. That entrenchment was portrayed brilliantly by documentary filmmaker Frederick Wiseman in his film, “Near Death” (1989), in which the viewer learns how treatability discourse works, and how it dominates doctors’ thinking and interactions with patients. The six-hour documentary shows us minute-to-minute deliberation, decision-making, and treatment in the ICU of the Beth Israel Hospital in Boston, where Wiseman let the cameras roll. We, the viewers, observe the private bedside conversations between physicians and critically-ill patients about their precarious conditions, and their multiple treatment options. We are privy to the long deliberation in the ICU corridors among the doctors and nurses about how much to do, which treatments to employ and when, and what the patient and family want to do next regarding specific interventions. The conversations are long, tedious, equivocal, repetitive and without resolution. The film follows four patients and their families. In each case the physicians spend considerable time at the bedside and in the hallways in earnest discussion about what to do. They are articulate, compassionate and knowledgeable. Their talk is endless, yet in each case they never talk about death. Instead they talk extensively about stabilizing the condition, potential organ system recovery, and which other specialists to involve in the case. The talk misleads patients who voice the hope and expectation that treatments will lead to some sort of recovery. The film’s final credits report that all four patients profiled died either in the ICU or on an acute care ward. None ever left the hospital.^{Footnote 4}

One source of the problem of the entrenchment of treatability is the fact that potential and possibility are what organize how physicians, patients and families engage with medicine today. Not very long ago we died from heart attacks, cancer and almost every other disease without discussion of treatment options, because there were none or few. Patients and families did not have to manage and choose the timing of their deaths or the quality of their lives—the troubling contemporary outcomes of patient autonomy and choice, shared decision making and informed consent. Those ascendant values in medicine are given voice, health care professionals assume, in their pragmatic offering of treatment options.

The articulation of options and treatability statements in the service of patient autonomy and informed consent can be taken to an absurd extreme, as Atul Gawande describes in his book, “Being Mortal,” when writing about his father’s cancer treatment choices and the family’s deliberations on what to do. When Gawande accompanied his father to the oncologist to learn about options, “She laid out eight of nine chemotherapy options in about ten minutes. ...It was dizzying...She described a variety of different combinations of the drugs to consider as well. The only thing she did not offer to discuss was doing nothing.” Gawande concludes, “There were too many options, too many risks and benefits to consider with every possible path...But the meaning behind the options wasn’t clear at all.”^{Footnote 5}

In 1989, the same year “Near Death” was released, the first large scale study of how death occurs in the American hospital, the SUPPORT study,^{Footnote 6} began with the goal of trying to understand why so many patients were getting aggressive treatments near the end of life that neither they nor their families wanted, and why communication between doctors and patients about the nearness of death and what to do about it was almost impossible. That study, designed within the framework of bioethics, assumed that if the road to shared decision-making and shared understanding between doctors and patients could be found through investigation of the doctor-patient dyad, patients and families would be satisfied with their care, health professionals would be less frustrated, and the mounting public outcry about too much intervention near life’s end would abate. But no such things happened. There was no shared decision-making. The initial study and its intervention, by narrowing its focus to the doctor-patient dyad (and later the doctor-nurse-patient triad), had not considered the structural and institutional routines, demands, and contexts in which the talk about treatments occurs. The research had not taken into account the powerful forces that impact, and in fact determine, the language that guides medical practice.

Four institutional drivers, hidden from public view, build on one another in a chain of events that underpins and governs medical practice today, inevitably guiding physicians toward the language of treatability statements. The first driver is the vast biomedical research industry and its expanding clinical trials engine, which is churning out evidence of new, effective therapies at an unprecedented rate. The second is the approval process of Medicare and private insurance, which evaluates that evidence and then deems the new drugs, devices and procedures reimbursable. Once a therapy is reimbursed by insurance, it almost instantly becomes a standard of care. Then doctors will mention, offer and prescribe it, insured patients will have access to it, and patients and families will want it. Finally, once therapies become standard, they also become ethically necessary and thus challenging for physicians, patients and families to refuse, and as this article implies, difficult, if not impossible, for physicians to decline to mention.

Standard, ethically justified, and reimbursed treatments, combined with the technological imperative in medicine, specialist knowledge, and doctors’ mandate to treat, ensure that the language of treatability remains robust and dominant, even in an ever-aging society, even when death is near, and any associated suffering can be ameliorated by comfort measures. This institutional chain of health care drivers, rooted in American economic and political priorities, promotes the flood of treatments to consumers, and instigates all the talk about them.

Words matter deeply. The authors rightly conclude by acknowledging that the idea of shared decision-making, in the light of discordant understandings between physicians and health consumers, needs to be reconsidered because language is never interpreted literally. Miscommunication resulting from treatability statements should lead all of us, providers and consumers of treatment alike, to consider a larger question generally missing from physician training, doctor-patient conversations, and societal debate: How, in an aging society, do we want to live in relation to medicine’s tools?