Everybody talks, but there is no conversation.
—DEJAN STOJANOVIĆ
Just Trust Us
Where does one go from here? Or, to phrase the question somewhat differently: What is the practical value of investigating the limits of what we know in medicine? One cannot, for instance, assume that patients and their families will become experts on the topics covered thus far, sifting through the technical literature so that specialist advice becomes unnecessary. We can’t expect people to become steeped in the data and commentary on mammography recommendations, the guidelines for cholesterol management, the limitations of Lyme testing, and so on for every conceivable medical issue that they or their loved ones might encounter. There are simply too many issues, each with its own particular oddities and refinements, for one person to assimilate. Of course, this is the point of having physicians in the room. They are a structural short-circuit for patients: trained professionals who are there precisely to relieve patients and families of the burden of mastering all the relevant studies in order to make reasoned decisions.
One of the implicit goals of this book has been to offer a new vision of medicine, one where the roles of doctor and patient are significantly redefined. At least in my model of medicine, patients should be the masters of their care, and doctors should be advisors—trusted advisors, one hopes, but advisors nonetheless. Yet doctors should not be expected to simply decree the major elements of a medical plan for a patient.** Along with many other colleagues, I conceive of the role of physician as a guide to the perplexed instead of a general commanding troops in a war against disease. Part of the reason sometimes there is a disconnect between patients and their doctors is that each party may be operating under different assumptions about which of these two different patient/doctor models is the more appropriate one.
Yet if patients should be the captains on their own ships of life, as it were, then what do we expect them to do in the kinds of situations I have outlined in this book, where often even experts disagree? Wouldn’t this justify an older, more paternalistic model in which doctors decided and patients “complied” with their judgment? That model, whose boiled-down message could be summarized as just trust us, may be the most useful because it can alleviate most of the anxieties surrounding patient decision making.
I think not. I think that the older style infantilizes patients and leaves doctors feeling as if they are engaged in some supernatural process, far beyond the talents of mere mortals. I think it reinforces the kind of hubris that we have seen over and over again in example after example in this book, where doctors place too much confidence in their judgment and in the technology they utilize. As I tried to show, the results of that system can on occasion be disastrous.
Thus, we need to imagine a new model, one based on the strengths that both the doctor and the patient possess when they encounter one another, for the clinical proficiency of the physician is only one kind of strength in medical decision making, yet we have taken it to represent the whole. I would submit that a much more powerful model is that of a conversation between the two parties, on equal footing but each with a special area of expertise without which any medical plan is incomplete.
How can such a conversation take place? And what should be expected of each party in this conversation vis-à-vis uncertainty? Let’s break down each role separately.
For doctors, the most important adjustment that is required is to acknowledge the potentially lifesaving benefits of humility. This means that doctors should nearly always be circumspect when talking about their judgment, either with respect to a diagnosis, a prognosis, or a treatment plan. It doesn’t mean not having an opinion, but it does mean that such opinions are constrained by the limitations of what is known about a particular subject, and how those constraints affect the risks of taking action versus doing nothing.
For instance, in the Introduction to the book, we looked at cancer prognosis: once upon a time, oncologists used to offer prognoses to patients. They had developed a variety of scales that predicted how many years or months a patient might have to live. These scales were built in part on complex staging information of the cancer, combined with various laboratory values that indicated how advanced a patient’s disease was. They would tailor their treatment recommendations based on how long they thought a patient with a terminal cancer would have to live.
But, as I discussed, once researchers started studying this in detail, they found that oncologists were unable to make accurate prognoses. When their ability to predict was put to the test, they failed fairly profoundly. So, instead of burying their heads in the sand and retreating from the implications of this fairly humbling revelation, oncologists made two important adjustments. First, and most important, they got out of the prognosis business. That’s not to say they stopped talking about survival odds and the data on groups of patients with a particular stage of cancer, for indeed a patient with a glioblastoma multiforme (a highly aggressive brain cancer) is much more likely to have a shorter time to live than a patient with a hormone-responsive prostate cancer caught at an early stage. But they stopped offering specific guesses for how long a given individual had to live.
Second, they created a much more simplistic scoring system to guide their recommendations for therapy. This four-point score doesn’t take into account the dozens of variables used in staging a cancer, but instead looks at patients’ general state of health and their ability to perform the kinds of daily tasks that healthy individuals do as a matter of course (like brushing their hair, bathing, and dressing themselves).** On these basic criteria alone, oncologists can make recommendations to their patients about treatment, precisely because they do not know with complete certainty which patients are likely to respond to chemotherapy or radiation and which are not. Therefore, they simply look at who is functionally strong enough to tolerate the side effects—one might use the word “ravages”—of these treatments. It is a much better and more useful model, one that incorporates uncertainty into its bones. And it rests on the premise that doctors simply cannot provide the kind of confident predictions that are often expected of them or that they may expect of themselves.
Humility is the cornerstone of this approach. It requires saying the words I don’t know without moving immediately into a defensive crouch—indeed, it requires the acknowledgment that I don’t know often reflects the status of some medical topic, even though much may be known about it. It calls for not talking at but listening to patients and families, and making a real attempt to understand their goals, especially in the setting of long-term or terminal disease. To do this, physicians have to ask questions, gently probing to make sure that both parties are on the same page, and locating themselves in roughly the same portion of the spectrum of certainty.
For patients, the conversation requires asking questions, lots of questions, about not only the benefits but the risks of treatment, and the consequences of declining to be treated for something. Sometimes those consequences are quite profound, and patients should develop a framework for understanding when a doctor’s recommendation is strong and not so strong. And if the doctor’s recommendations about everything is strong, and no explanations are offered about why this is so, it’s probably time to find another doctor. If you’re feeling bullied, you’re being bullied.
Patients should also ask about the difference between relative and absolute risks and benefits. When possible, try to have your doctor put numbers to these ideas. How many people like me need to take this drug before one life is saved? is but one question that can be asked of a physician. At present, many physicians do not automatically incorporate relative versus absolute benefits into their medical decision making, so this process may take some time and so “the” conversation may require several discussions at more than one office visit. Furthermore, patients are better served in the conversation by knowing more about statistics and probability, so that if their doctors are able to rattle off data from the latest large, randomized trial, they will have a working idea about what’s impressive and what isn’t. Needless to say, familiarizing oneself with all of these concepts takes some time. The more you read about these matters, the more adept you will become at grasping what’s really at stake in a medical recommendation. In that way you will become a more equal partner with your physician, able to formulate a plan based on your goals, rather than follow a plan handed to you by fiat.
Let’s look at two specific examples of how I think this type of conversation can play out. The first case is hypothetical, and I will talk about it from my perch as a doctor. The second case, however, really did happen, and it happened to me, but as someone on the receiving end of the medical system, as a family member of a patient.
Earlier in the book I wrote about the complicated issue concerning blood pressure management in a select, but large, group of seniors: those age sixty to eighty with blood pressures above 140 but below 150 without diabetes or kidney disease. What would I do if I were evaluating a new patient who fit this demographic and presented a pressure of 146, which was then confirmed after repeated checks at several early visits? How would I go about discussing the controversy with such a patient?
I am not a specialist in blood pressure management, and I have only a nodding familiarity with the studies that have formed the basis of the guidelines published by the JNC8 group and the dissenting minority. In this respect, I’m not much different from typical primary care physicians who must grapple with such matters for their patients, although as an infectious disease physician, I do much less blood pressure management than a primary care doctor. What this means is that I do not have a definitive, expert opinion about whether I should follow the older JNC7 guideline and recommend that the patient start treatment to take him below 140. That said, I think I could have a reasonable conversation with the patient about how uncertainty is crucial to understanding why there is disagreement, and in doing so help him make a fairly informed decision about his goals with respect to his blood pressure.
The first point I would make in such a conversation is that, despite much heated rhetoric surrounding this process, there is actually a significant overlap between the JNC8 and minority groups about most blood pressure targets in most people. “Although there was almost unanimous agreement on nearly all recommendations, a minority of the panel disagreed with the recommendations to increase the target blood pressure . . . in persons 60 years or older without diabetes or chronic kidney disease,” stated the authors of the minority report (my emphasis). In other words, the current dispute about the applicability of JNC8 applies to just one type of patient. Granted, there are about 30 million of these patients, but anyone who has diabetes or chronic kidney disease, or is over the age of eighty, need not worry about ambiguous messages. For these patients, all of the experts weighed the evidence and arrived at precisely the same conclusions. This is important because it emphasizes to the patient that the experts are hardly in overall disarray, so one just shouldn’t throw up one’s hands and say “nobody knows anything,” and walk out of the office.
If the patient in front of me was among those 30 million who might be affected by the new guidelines, our conversation would be aimed at exploring how uncertainty led to the dispute. I would explain that the difference of opinion boiled down not so much to a quarrel about the numbers, but the attitudes we have toward evidence and the goals of guidelines in the first place. For although it may be true that it is easy to memorize a target pressure of 140 or 150, nobody really believes that, all of a sudden, the totality of lifesaving benefits of blood pressure control kick in if someone drops from 141 to 139. Consequently, some of the disagreement can be explained by the intellectual and emotional response to data as much as the data itself.
I would also explain what we observed in various instances in this book: that there are big benefits to treatment if your problem is severe, but as you approach “normal” the risk-to-benefit ratio changes, such that the risks become greater and the benefits correspondingly smaller. So if my patient was skittish about taking medications, as many patients are for perfectly justifiable reasons, I would explain that one option is to hold off. But I’d also emphasize that it’s critically important to continue to receive evaluation, because if that patient’s pressure gradually moves up to the 170s or 180s, those bigger benefits kick in, and my recommendation to start treatment would become much stronger. In other words, the severity of this patient’s hypertension influences where my advice to take medication can be found on the spectrum of certainty: the higher it is, the more leftward it moves.
Some patients, however, can’t stand the idea that they are at risk for some complication of their disease, and will want to do everything they can to minimize those risks. For such patients I would explain the controversy in a slightly different way by emphasizing the potential downsides of treatment, and emphasizing what I noted in that chapter, that lowering one’s pressure too far can lead to its own set of problems. Still, if this were the kind of patient I was facing, I wouldn’t actively discourage him from trying to aim for a goal below 140, for the JNC8 guidelines merely noted that there wasn’t evidence to indicate that such a goal was unambiguously beneficial—but neither did they indicate that it put one in danger of complications.
From my vantage as a physician, this conversation depends upon a patient’s aims, concerns, willingness to live with the risks of medications, and attitudes about the disease. The conversation also depends on an honest assessment of what constitutes a small amount of risk versus a big amount of risk. These form all the essential ingredients of the process we call informed consent.
Now let me pivot to relate a different story where I was on the flip side of the equation.
Several years ago a much-loved family member went out one winter’s day to collect the mail and collapsed. A neighbor happened to see him fall to the ground and called 911. If it had not been for that chance event, he would have died in all likelihood, but because of the call and the rapid response of the emergency medical technicians, who were on the scene in a matter of minutes, he was revived, intubated, and taken to the nearest hospital. The family was summoned from various corners of the country to support one another and review the doctors’ recommendations.
One vital problem was that, unbeknownst to the EMTs, our relative had a “do not resuscitate” order in place. He was seventy-six and had various medical problems, including diabetes, heart disease, and high blood pressure. He directly expressed to his primary care doctor that he had no interest in lingering in a hospital bed for months on end if he should have some terrible medical calamity befall him. Once he was brought to the hospital, where this could be sorted out, it should have been a simple matter of “reversing” his status, taking him off the ventilator and letting him pass away (or live, if that should be the case). His primary care physician was well aware of his DNR status, and there is no ethical dilemma in extubating a patient who never wanted to be intubated in the first place. A few phone calls could have taken care of the issue, and although tragic, it would have been consistent with my relative’s clearly stated wishes.
As I made my way across the country I thought it would end there, but the team of doctors, which included an intensive care physician, a neurologist, a cardiologist, and the primary care physician, never even hinted they were considering withdrawing care. When we as a family pointed out the advanced directive, they responded that he was on a new treatment known as a “hypothermia protocol,” and that this required time to see how he would fare. The family turned to me, as the lone physician in the bunch, waiting for an explanation.
Only I hadn’t heard of a hypothermia protocol. It had been more than five years since I had last worked in an intensive care unit, and I wasn’t familiar with this new practice, which had very rapidly been adopted by the neurology community. This meant that, although I had a more sophisticated grasp of the technical matters in play, I was not much different from any nonmedical family member speaking with the doctors about the plan. I learned that the hypothermia protocol was designed to preserve the brain function of patients who were “found down” and that it involved lowering the patient’s central body temperature from the high nineties to the low eighties Fahrenheit through the use of cooling blankets and other means. The protocol lasted for seventy-two hours. By the time I had arrived, the first twenty-four hours of the protocol had already passed, and before I could even formulate questions about the process, another twenty-four had flown by. We were given status updates, but never at any point was there a conversation about whether it made sense to take this course of action given our relative’s wishes.
Eventually, I decided to call some of my own colleagues back in Massachusetts—first a few neurologists to learn more about prolonged hypothermia, and then a palliative care doctor familiar with the ethics of situations such as these. What I learned from the neurologists was a case study in uncertainty. For, far from bringing all or even most patients back to resuming anything close to their previous lives, patients who survived events like these through hypothermia protocols were assessed in terms of their ability, for example, to hold a toothbrush six months after the event. It sounded like this kind of life was exactly the fate our relative was trying to avoid. Moreover, while the trials performed on patients with these hypothermia protocols showed a reasonable relative benefit, the absolute benefit was (as we have seen many times in this book) much smaller.
Not one physician with whom our family spoke during this time ever made this clear; all of them assumed that this was something that we should simply want by dint of the fact that, well, it just worked. There was no individual or collective recognition that “worked” was a term that might need some further explanation, and that, because our relative had made explicit that he didn’t want the kind of things done to him that currently were being done to him, there absolutely should have been someone making some attempt to do so. They were behaving as if this protocol belonged at the far left of the spectrum of certainty, so clearly beneficial it hardly merited a second thought. Once I got my head around the concept and took a cursory glance at the studies on this new technology, I thought it just a hair to the left of dead center. It’s entirely reasonable that some patients and families should want such treatment without question, but it also seemed clear to me that this isn’t what our relative would want.
Eventually, after many long and painful family discussions, we became more assertive with the physicians about these matters, finally convincing them to extubate our relative, which they grudgingly did a few days after he completed the hypothermia protocol. He survived that event but passed away several days later, never having awakened after the initial event, his brain simply having taken too much of a hit during those minutes after he collapsed.
That relative was my father.
Looking back, what astonished me most about my own role in the saga of my father’s final days is how much I was intimidated by the process. After all, I know medicine; I know how to talk to doctors, I know how they think. I also knew without any doubt what my father wanted out of the medical system, and it was plain to see as I watched him lie in a bed with a tube stuck down his throat, with a machine doing his breathing for him, that this wasn’t what he wanted. But faced with a new set of variables (he had inadvertently been intubated by people who weren’t aware of his wishes, here’s this fancy new technology that we use leading to better outcomes) I hesitated. The consequences of that hesitation included a great deal of tension within our family about how to proceed. We were fortunate to resume, after a time, our loving and caring relationships with one another. Other families who find themselves in a similar bind may not be so lucky.
You might conclude from all of this that I’ve just given you another example of doctors lacking in humility about the value of their medical plan—which it most certainly is—but I am providing this as an example about what can be done when speaking to doctors because eventually I did assert myself, and, along with my family, we did ask the right questions, explained our goals, and repeatedly drew their attention to what my father wanted. I wish that they had listened better and understood us sooner, or even had been proactive in seeking our thoughts once they realized that his advanced directive was in conflict with what was taking place in his care.** Yet, with polite but firm insistence, we got them to understand (if perhaps only dimly) that this was not about their plan but his. From a patient standpoint, that is the essential point to grasp when you begin a conversation with a doctor. Your medical plan is yours: you own it, and your doctor should be your guide, not your director.
The Spectrum of Certainty, Redux
As we think about how to have a conversation with physicians and deal with uncertainty, it’s important to emphasize the altered perception of certainty among people both inside and outside medicine. I hope to have shown that our treatments and technologies are sometimes given far too much credit for their effectiveness once one carefully considers the data. This is one of the reasons why about-faces like that seen in the changing evidence on the benefits of hormone replacement therapy seem so embarrassing, and why physicians a century hence might view the rhetoric from today’s leading doctors as ridiculously and sometimes lethally smug, as we sometimes view what passed for common knowledge among physicians who came one hundred years before us.
As I’ve tried to persuade you to see, the level of certainty in many of the topics I have covered in this book resemble those of the JNC8 guidelines mentioned above, falling somewhere toward the middle part of the spectrum. The evidence supporting the proposition that screening mammograms save lives in women over fifty is mixed, but on the whole many or perhaps even most experts still favor some form of screening—although as discussed previously, the popular perception about the scientific certainty of mammography’s benefits is profoundly different from this mild statement.
In women younger than fifty, there is much, much less certainty. Does the evidence from scientific trials suggest that screening mammograms in women under fifty belong on the other side of the certainty spectrum, and move toward some level of evidence of harm? It’s hard to say at the moment, in part because it’s extremely difficult to quantify and measure harms of mammography in such a trial, and the expense in conducting such a trial is enormous at the very time when the US government (and a good many other governments) is cutting back on its support of scientific and medical research. After spending time doing the research to write this book, I now believe in general that screening mammograms for women under fifty does indeed provide net harms without a clear corresponding net benefit. It’s not far to the right, but I’m fairly confident that, barring some additional technology that will help separate false positives from true positives, it doesn’t belong to the left of the center on certainty’s spectrum.
Of course, nothing magic occurs in the biology of a woman when she turns fifty, so the precise moment when the risks of the false-positive mammogram are outweighed by the benefits of early detection may range from anywhere in the late forties to the early fifties (if they exist at all) and could be influenced by any number of factors that can really only be measured in theory. That is, one could, for instance, perform a screening mammogram trial on women aged forty-seven to fifty-three who are positive for the BRCA1 gene mutation, which predisposes women to a higher risk of breast cancer, but the logistical hurdles required to address such a narrow question make such a study impossible in reality.
(As an aside, as this book is finishing its final revisions in mid-2015, the US Preventive Services Task Force is seeking public comment on a draft form of its newest iteration of screening mammography guidelines. It does not appear to be a dramatic departure from the 2009 document for the three major items. Screening for women under fifty has retained a C grade, while every-other-year mammograms for women fifty to seventy-four continues to be given a B grade, and mammograms for women seventy-five and over are still given an I grade.**
It is hard to predict what the reception will be like, but an early indicator might be found in a press release by the American College of Radiology. Recall that the ACR had taken the USPSTF to task in 2009, and that stance has not changed. “They ignored more modern studies that have shown much greater benefit,” said Dr. Barbara Monsees, the chair of the ACR Breast Imaging Commission. “These limitations result in the misrepresentation of the real trade-offs that women and health care providers need to know about in order to make good decisions about screening. They also ignored the demonstrated views of American women on screening. Unfortunately, these recommendations will only add to confusion that is placing women at risk,” she noted. Why the ACR’s opposition doesn’t also appear to add to the confusion is not made clear by Dr. Monsees. And again the USPSTF continues to be charged with having been tone deaf to the “views of American women,” as if somehow that were one unified, monolithic view, and as if the view should forever be fixed without ever reconsidering the strength of the data.
Thus, my impression remains that even with new guidelines about to be issued, not much has changed, and this book’s consideration of a report that was issued nearly a decade ago is still very topical. For which I suspect my publisher is breathing a sigh of relief. End of aside.)
Prolonged antibiotics (that is, months and months) for so-called chronic Lyme disease occupy that part of the spectrum where we become more certain of the harms of treatment. As to what causes the symptoms that are given the label of chronic Lyme, it is a true unknown unknown—it is even unclear whether Lyme itself plays a role at all. A variety of theories have been proposed to explain chronic fatigue syndrome, many of which involve various infections and all of which are highly speculative. The simple truth at this point in time is that nobody really has any idea why this happens to people or even how to measure it. Nevertheless, we do know with much higher certainty that more antibiotics aren’t helpful and are highly likely to be harmful. Long courses of antibiotics for chronic Lyme is out at the right end of the spectrum of certainty, and we’ve known this for well over a decade.
On this very brief tour of the certainty spectrum, we should finish by coming full circle. As I’ve indicated, in medicine there are very few absolutes, few points at which we really can look a patient in the eye and say, “just trust us,” and mean every syllable. Part of this is that all treatments come with risks. The conversation must deal with the balancing of these risks and benefits. But there is one area of medicine in which the benefits are so enormous, and the risks so minimal, that it occupies a spot on the far left of the certainty spectrum and represents the single greatest triumph of modern medicine. Not altogether surprisingly, the public perception is that this medical miracle is actually a highly controversial topic. But it’s not, and the reason it’s not is due to the overwhelming scientific evidence that it saves lives and is associated with almost no harm.
I am referring to the practice of vaccination. From their origins, when the practice of protecting against smallpox through variolation (a forerunner to true vaccination) was introduced to Europe via the Ottoman Empire in the early eighteenth century, vaccines have always engendered a certain level of paranoia. Over these three centuries, some of the arguments against the practice remain more or less the same: one of the complaints that the clergy levied against smallpox prevention when Lady Wortley Montague was among the first to not only engage in the practice but publicize it as well, was that the practice defied God’s will, for if one was meant to get smallpox and die, that was fate and fate should not be circumvented. Today’s antivaxxers, with their zeal for what they fancy are “natural” remedies, sometimes organize “measles parties” so that their children can be exposed to the virus in one happy group. Whether one wants to let God sort them out or let Mother Nature take her course (two different but equally powerful kinds of narratives driving these attitudes), the hostility to vaccination in part rests on a fatalism that made only a small amount of sense at the advent of the Enlightenment, and makes considerably less sense today.
It is impossible to assess with any accuracy the total number of lives that vaccines have saved since Edward Jenner began the practice of vaccination in earnest in 1796, but that hasn’t stopped researchers from making some decent attempts at estimating the effect where the epidemiological data are more reliable. Recently, the CDC engaged in just such an exercise. In order to make reasonably solid conclusions, it focused narrowly on childhood immunizations over the twenty years between 1994 and 2013. In this brief span of the vaccination age, the estimates are mind-boggling: 322 million overall illnesses along with 21 million hospitalizations were prevented, and more than 700,000 lives were saved. The estimated cost savings was more than $1 trillion.
The biggest lifesaver came from preventing diphtheria, a disease that is almost never talked about today but was an absolute terror in the early twentieth century. The vaccines that were estimated to have saved large numbers of lives included vaccines against measles, pneumococcus, Hepatitis B,** and to a smaller extent pertussis, polio, and a relatively new vaccine against the bacteria known as Haemophilus influenzae B. The remainder of the vaccines proportionally saved a smaller number of lives but often had big impacts on morbidity. For instance, the vaccine for varicella (the virus that causes chickenpox) was estimated to have saved perhaps one thousand lives over these twenty years—a still not inconsequential sum—but prevented nearly 70 million illnesses.
That’s just over twenty years of work in just one country. Suffice it to say that the burden of disease in many other countries is much higher, and therefore the benefits of vaccines in places like Ulan Bator or Kinshasa are that much more profound. Keep in mind that these are lives saved—tragedies averted—usually with a one-time clinic visit. Moreover, most vaccines, especially those given in childhood, provide extremely high levels of protection for the recipient. Tetanus, measles, polio, and diphtheria are almost totally unheard of among those who have been vaccinated. Contrast this with the impact of statins I discussed in Chapter 6, which are clearly lifesaving drugs but only reduce the mortality from heart disease by a modest portion, require lifelong adherence, and carry a small number of risks.
It’s safe to say that, assuming we continue to develop an ever more sophisticated understanding of the human body and how to keep it healthy, doctors two hundred years hence will look back at our statins and our antidepressants, our mammograms and our CAT scans, our chemotherapy and our blood pressure management, and think that we were only really a step or two beyond butchery. But they will look at what we did with vaccines, and they will see real accomplishment in precisely the same way that we regard Newton and his laws of motion, or Harvey describing the circulation of the blood. The success of vaccination is truly the one matter on which we can bring near certainty to the discussion, and it is a heartbreak to see such an easy topic so terribly muddled in the public discourse. There are books that elegantly explain this principle in much greater detail than I have in these meager paragraphs, of which volumes such as Paul Offit’s Autism’s False Prophets, Seth Mnookin’s The Panic Virus, and Arthur Allen’s Vaccine are but a few.
Media
Exercise more. Eat less. Don’t smoke. Everything else is commentary. This should be the beginning and ending of every single health news story. It is the most important message the media could convey about health, and yet most Americans don’t understand just how important this message is because it is obscured in a dust cloud of largely irrelevant material. Of course, educated Americans are aware that there are health benefits to be had by a moderate diet and regular exercise, and certainly the news media doesn’t try to oppose that message at a conscious level. However, the media’s fascination with the latest bauble of technology and its obsession with improbable threats ends up producing the same effect.
The essentials of American health—or lack thereof—can be boiled down to these few basic points: we eat too much, and too much of that too much consists of crap; we don’t exercise anywhere near enough because we spend far too much time either sitting in front of televisions or getting to and fro in our cars; and a smaller subset of us smoke cigarettes, a practice that wreaks disproportionate havoc on the body. Of the many news stories on lifestyle research—what types of food and drink are associated with healthy outcomes, whether trans fats are healthier than saturated fats, the healing powers of the açaí berry, and so on—those that fail to place such stories in the context of those basic points aren’t really making any genuine contribution to health education. Readers of such articles invariably walk away with wildly skewed notions of what they need to do to improve their health, which accounts in part for the repeated waves of fad diets or trendy foods that dominate the health consumer market.
The conversation about uncertainty is not merely between patient and doctor, for both patients and doctors enter the clinic doors having shaped their opinions by watching the evening news, or reading online articles sent by friends or colleagues, or just soaking in popular culture where some health story is related in a movie or TV drama. It may be too tall an order to hope for all forms of media to highlight the challenges that uncertainty brings to the practice of medicine, but it seems reasonable to expect at least health reporters to understand something of it. In the news media, the state of affairs is so sorry that really the only direction to go, in terms of quality and sophistication, is up.
How can health news stories achieve this? Simply put, they can acknowledge that uncertainty happens, and, in putting it out there for all to see, they can explain where on the spectrum of certainty a given story belongs. Let’s suppose that some group of researchers has found a drug known as K29-X that can reverse multiple sclerosis in rats. They publish their results in a prestigious medical journal, and a press release is generated by the medical center where the researchers work. Some local TV stations pick up the story and devote three minutes to the promise of K29-X. Various viewers afflicted by the disease then happen to watch the feature on K29-X and soon are called by friends who have seen the pieces as well. Should they call their doctors’ offices asking for a prescription?
In reality, the better question to ask is whether K29-X has even been tested in humans—which it almost certainly hasn’t—but after such a story lands on the local 6 p.m. newscast, one can rest assured that, the following day, the administrators of local neurologists’ offices across the viewing area will be deluged with calls. I’m making the K29-X scenario up, although the phenomenon I’m describing is very real. The “HIV cure” story of the Berlin patient happened to have been a nationwide example of the same, and HIV providers spent much of those weeks in December 2010 tamping down expectations for their patients (most of whom were healthy anyway through the magic of antiretrovirals) who thought they could be free of the virus henceforth.
So the starting point here is to couch a story like this in a chain-link fence of caveats so that viewers might catch the message of “enter at your own risk.” Or believe at your own risk. K29-X was used in an experimental model for MS that was not MS itself—uncertainty—was studied by looking at animals rather than humans—uncertainty—animals that are only distantly related to humans—uncertainty—nobody has any idea whether or not K29-X is even safe in humans—uncertainty—and even if it turns out to be safe, nobody has any idea whether it’s effective in humans —uncertainty.
Is this a lot of uncertainty? Is this in that part of the spectrum of uncertainty that includes the unknown unknowns, where really just a bunch of clever scientists have been messing around with some interesting ideas, none of which may have direct applications for multiple sclerosis, and even if it does are almost definitely not going to happen anytime soon? Pretty much. Of course, local newscasters may not be interested in hearing this critical appraisal, as it may interfere with their business model. So, until this changes, the only reasonable solution in the interim may be to turn the television off—or to construct that mental chain-link fence on one’s own. There is a value in actively identifying the limitations of health news stories, and although it takes some practice, one can get pretty adept at doing so after a time.
Health news would also better serve its consumers by explaining the quality of research and the level of evidence that supports a given claim. So many of the items that purport to show odd or counterintuitive health information are based on small observational studies, which are subsequently extrapolated and inflated into sweeping statements that bear little relation to reality. The New York Times article that considered the health benefits of dark chocolate was careful to note the relatively weak level of evidence supporting the claims of the researchers, specifically noting that the research was observational in nature, akin to the kind of research that was performed on the benefits of hormone replacement therapy before a properly constructed drug trial showed that hormones didn’t lead to greater health. I see no reason why all health news can’t similarly do this, even if one is talking only about a televised “health minute.” It takes but a moment to say that the quality of a given type of research is high, medium, or low; no matter how cleverly a researcher designs an observational study, it’s still an observational study, and reporters would serve their readers and viewers well by explaining the differences among these types of research as part of every single report.** Again, readers can’t change what news organizations do, but what they can do is identify whether a useful assessment of a study’s quality is present, and, in the age of the Internet, it is possible to find legitimate sources that will discuss these studies, especially those of interest to the general community.
Another simple change media outlets could make is to attach a grade to a given health claim by noting that it is based on strong or moderate or weak evidence. Even for readers and viewers who do not wish to trouble themselves with the fine print, news media can intimate the level of confidence one should have in a given health news item. Indeed, this is the approach taken by the US Preventive Services Task Force in issuing its various recommendations, not only for screening mammography but for prevention of falls in the elderly, the use of aspirin to stave off heart attacks, and much else besides. In all of these cases, the recommendations are given grades that are easily comprehensible to anyone who has survived public school: Grade A evidence represents a recommendation toward the left, “fairly certain” end of the spectrum, while Grade B indicates less certainty, and Grade C much less certainty.†† However, all three of these grades still do recommend for something, while Grade D recommends against a practice because the evidence tilts against it on the whole, and so it is on the right side of the spectrum.
The USPSTF assigns the grade of I to its true unknowns, where the evidence is so scant that to take any stand, however meek, would be an exercise in shooting in the dark. Looking through the USPSTF website, “I” statements aren’t found at every turn, but neither are they so rare as to stand out. Incidentally, that’s not a bad thing—not only is it an honest admission of uncertainty, but it is also an identification of the areas where quality research must be done.
It’s not that hard to see the intuitive appeal of such a grading scheme. Like sportscasters who explain the meaning of some clever maneuver to an audience that may include viewers new to a sport, health news reporters could gradually ease their audience into the levels of uncertainty in medical research by attaching such grades to every article they write or every TV segment they air. The additional benefit of this approach is that all the various items of health news then become linked by analysis about the quality of evidence; news consumers can suddenly see the common theme of uncertainty in a diversity of stories, which applies equally to the latest newspaper report on a new test for Alzheimer’s disease, to a TV piece on the link between saturated fats and depression, and to an Internet webinar on new techniques for estimating cancer risks. This book has simply been an extended exercise in illustrating that point. In all of these topics, our state of knowledge is at varying levels, and making this matter explicit can only help people make better judgments and avoid the problems associated with the Just Trust Us model.
One small ray of hope has started to shine again on the media, for as of the summer of 2014, Gary Schwitzer has resurrected the HealthNewsReview and resumed his advocacy for measured reporting. One early review from the rebooted website centers on the news surrounding the latest entry in mammogram technology, so-called tomosynthesis imaging, and the glowing reviews that accompanied an important article published in the Journal of the American Medical Association. “The list of stories that resorted to sensational language—breakthrough, game-changer, best way of detection, any woman should have this—was long,” Schwitzer noted. He proceeded to itemize the ways in which caution was thrown to the wind: one story featured only one single patient anecdote, another interviewed the lead study author but sought no other expert voices, another made no mention of limitations, and so on.
But so, too, did he point out the stories that at least got part of the story right, which was that this undeniably big technological leap may not translate to a clinical benefit, and may simply exacerbate the underlying problem of false positives and the harm that may result from them, especially in women younger than fifty. Those organizations called out for varying levels of praise included USA Today, the New York Times, the Boston Globe, and the Wall Street Journal. Whether the work of Schwitzer and the many experts that comprise HealthNewsReview in its reincarnated state will be appreciated for its value is unclear, but at least they are back at it.
The End
Through efforts like the grading scheme introduced by the USPSTF or the publication of guidelines that emphasize the limitations of evidence such as those seen in the JNC8 hypertension recommendations, we’re beginning to see the conversation about uncertainty take place between specialists and laypeople. Yet these documents are still fairly technical and, although they are eminently readable, are mostly accessed by professionals rather than the general public. Still, the fact that these items are becoming ever more explicit in stressing the uncertainty attached to current medical knowledge and practice is representative of a growing awareness among medical researchers and public health authorities that we must have this conversation as medicine becomes ever more complex and intimidating.
The media has far to go in terms of enabling this conversation, although as noted above, some reporters and news organizations get it right. Among doctors, nurses, social workers, psychologists, and everyone else who can claim the title of health-care worker, the conversation is not yet happening in a systematic way, although those guidelines may offer a picture of things to come. Currently medical schools, as a rule, still encourage a model in which students are appraised by knowing right answers, choosing them among a field of wrong answers. That measures a certain type of knowledge essential to medical practice, but it consequently engenders a conception of medicine best described as overly certain, and residency and fellowship directors must spend part of their energies remolding their trainees to be aware of the pitfalls of this approach. Most nascent doctors are not judged on being able to express uncertainty to their patients, but, given the changing nature of guidelines and the many examples of uncertainty that we have seen in this book, some medical school deans may begin to incorporate those concepts at an earlier phase in physician training than we do at this time.
There is a place where a direct conversation about uncertainty is taking form. Health-care workers, alongside patients and families, are starting to wrestle openly with the challenges and opportunities that uncertainty brings to a medical problem. This place is not in a simulator at an academic center, nor is it in an experimental holistic-healing alternative medicine nationwide movement. In fact, the conversation is taking place in what at first glance might seem the least hospitable of environments, where the stakes could not be higher and the tension and anxiety experienced by patients and their families take them to the breaking point. As the Foreword described, the conversation about uncertainty—and a formal analysis of its promise and perils—is happening in intensive care units across the country.
An ICU is about the most intimidating environment possible for a layperson. The activity is frenetic, and the advanced technology is omnipresent. The pinging and beeping of every conceivable machine can paralyze family members with fear. Don’t look at the monitors, and ignore the alarms is almost always among the first things I say to family members when I am involved in the care of an ICU patient, as they anxiously glance at cycling blood pressures or react to a drop in their loved one’s heart rate by nearly fainting themselves. By definition, if you are in the ICU, you are there because you are potentially facing the end of your life, and you require the lifesaving equipment and personnel that ICUs provide. Patients are, it goes without saying, almost always quite ill. Their family members wear their tensions on their sleeve, and if they aren’t exhausted initially, they often become that way, and with each passing day, their nerves fray bit by bit. As I noted above, even I experienced it as a son when my father was critically ill. I was intimately familiar with the rhythms and noise of the place, so I felt only a fraction of the anxiety that others must normally feel, and even for me it was a brutal assault on the senses. It’s a very tough setting.
So it comes as something of a surprise that the ICU has become a testing ground for transparency about uncertainty. After all, in the midst of the chaos and tension, having frank discussions with families about what may be moderate or even high levels of uncertainty at the very moment their loved one is at their most vulnerable could seem like sheer folly. But a small cadre of professionals from various disciplines have posited that the ICU may be the optimal environment to test the theory that more openness from medical professionals—including honest assessments about uncertainty—leads patients and family members to believe they have greater control over their decisions, not less. Transparency in the ICU offers readers a living, positive proof of the potential we have to experience and improve medicine though embracing uncertainty.
It is still very much a new approach, but it is spreading across the country and gradually gaining acceptance. The origins of what is now increasingly called ICU family rounds lay in observations from the late 1990s that lack of communication from medical professionals was a much greater source of frustration to families than perceptions about competency or dedication of the staff or even the overall care of the patients themselves. There isn’t an overwhelming amount of research devoted to the topic, but what work has been done highlights a classic flaw of the modern medical machine: patients can be cared for with the finest technology and be saved from what would otherwise be almost certain death, and yet, because scant attention is paid to keeping the lines of communication open, patients and family members can nevertheless emerge from such experiences feeling sullied and violated.
Thus, the original concept of ICU family rounds was meant to overcome the problem of communication, and wasn’t thought of as a chance to hold a seminar on epistemology. Not altogether surprisingly, the early literature discussing the potential benefits of family rounds could be found in the pages of nursing journals, because nurses are on the front lines of patient care and are much more likely to be aware of the vibes of frustration emanating from families as they try to comprehend the complexities involved in an ICU stay. Among physicians, many of the initial forays that considered how to incorporate families into ICU rounds occurred in pediatric intensive care journals. But the idea has spread, and although I am an outsider to the daily practice of intensive care medicine, it seems to me to be gaining some traction.
ICU family rounds may have been conceived as a tool to improve communication, but all physicians who have done some training in intensive care medicine know that it is exceedingly rare to find a case where one can interpret the avalanche of information into a coherent whole for which there is one single treatment devoid of any downsides—so uncertainty as a topic simply can’t be avoided if one chooses to adopt this model. Uncertainty of varying levels rules the scene: Do we give blood thinners for what may be a serious heart attack and in doing so risk a life-threatening colonic bleed? Do we subject a patient with a tenuous respiratory status to a bronchoscopy if it means that we may worsen the problem and end up causing the very intubation and dependence on the respirator that we are trying to avoid? Such kinds of questions are part and parcel of the treatment of critically ill patients, and the consequence of being transparent with families, and sometimes the patients themselves, is that these questions have to be addressed openly.
In keeping with the experimentation of this new method, not all ICU family rounds are the same. Some ICUs still tend toward a more traditional model and have simply organized the daily schedule to optimize communication, performing staff rounds separately but setting aside an hour—say, in the late morning—when the nurses and physicians are available primarily to discuss the patients’ plans of care with families. Such an approach allows family members to know precisely when they can find the team and avoids one of the major hassles associated with communication. But there are more aggressive approaches, the most profound of which is to simply bring families right to the patient’s bedside and let them not only observe but to an extent participate in daily rounds.
In a medium-sized community hospital in Salem, Massachusetts, a group of intensive care nurses and physicians chose the latter strategy. Writing in a newsletter published by the Massachusetts Board of Registration in Medicine in early 2014, Dr. Barrett Kitch of the North Shore Medical Center discussed his group’s experience with this more radical form of ICU family rounds. It reads like a primer on how doctors and nurses can have the conversation about uncertainty and in doing so enable rather than exasperate patients and families. “The initial reaction to the proposal was one of skepticism,” Kitch wrote. “Patients’ families would become worried about information that shouldn’t worry them. They would become distrustful because they would hear us (physicians, nurses, respiratory therapists) second guessing ourselves. They would get lost in the details and miss the important message.”**
After convincing some of these skeptics to allow a pilot trial, the results were impressive: “[The medical staff] reported that the families were extraordinarily appreciative,” Kitch noted. And in a phrase that couldn’t be better suited to this book’s thesis, he added, “[The families] shared that transparency paradoxically seemed to increase trust despite more openness about our uncertainty, our second-guessing, or our lack of consensus as to the treatments needed for their loved ones” (my emphasis). The pilot program did overcome the more skeptical members of the staff, and this rather remarkable form of specialist/layperson interaction has been the standard of care there ever since.
Notably, rounds are not intended to be a form of participatory decision making. Despite this, they have allowed family members to feel some comfort by understanding the process, and they have enabled situations where family meetings can take place to discuss the weighty goals of care, which often involve questions of life and death. These family meetings, of course, take place in all ICUs across the world, but one gets the impression that the families in Salem might be better prepared and less mystified.
Because the ICU family rounds as practiced at North Shore Hospital is not a rolling family meeting, it is not a perfect model for how to conduct the conversation about uncertainty. The staff don’t trouble themselves to explain their jargon (Dr. Kitch wrote that “rounds were not simplified so that they could be understood by the layperson. . . . [W]e spoke in the language of the healthcare worker.”), and they make no real attempt to modify their rounding structure. Yet the fact that their form of rounding has succeeded demonstrates that families—and, by extension, patients—are able on the whole to live with uncertainty without regarding the staff as a bunch of dunderheads, and without being paralyzed by fear. They have shown that we need not pretend uncertainty doesn’t exist at the edge of life. How much more can be gained when, in other, less anxious environments, the specialists do explain in plain language the conundrums that they face in making treatment decisions?
Our age is one where there appears to be little time for reflection about what we know and what we do not, not only in medicine but in almost any aspect of our lives. The gradual process by which we have squeezed out that area where we can ponder the limits of our knowledge has come via the ever-increasing yowls of the overly certain, who as a rule have something profitable to peddle. Part of the point of having the conversation about uncertainty at all is to reclaim that space where we can contemplate our goals—for our health, our jobs, and our sense of meaning.
We are capable of achieving it.
* The philosophy underpinning this model of medicine is beyond the scope of this book (note to publisher: I have more material for another book!), but it is worth emphasizing that supporting patient autonomy doesn’t mean that one expects patients to decide on every minor adjustment to a medical plan. In discussing these meaty matters with colleagues and students, occasionally someone defending an older, more paternalistic model will get a bit too cute and wonder aloud the wisdom of letting a patient decide whether to adjust the potassium level during an inpatient hospitalization, for instance. That’s not the kind of decision making to which I am referring. But paternalism still has its defenders: Sandeep Jauhar, a physician and sometime essayist for the New York Times whom I admire, wrote about the case for what he called “hard paternalism” in an op-ed in 2014 titled “When Doctors Need to Lie,” which was a piece that I didn’t admire so much.
* Technically, five: the fifth “score” is death.
* One doctor “got it” right from the start: the palliative care physician. For me in particular, he was critically important because early on I thought I was losing my mind for even bringing up the issue of whether he should be extubated, given the subtle but unpleasant response from the other physicians, as if I was a cruel and heartless son for wishing such a thing.
* I’ll explain the I grade a little later in the chapter.
* The estimated lives saved for Hep B run into the future because Hep B causes chronic liver disease that can lead to liver cancer. Because this happens decades later, children vaccinated against Hep B (especially children of Asian and African immigrants, where the prevalence of the disease is very high) will reap the benefits later in life.
* Throughout the West African Ebola epidemic, for example, the BBC articles covering Ebola included a few bullet points on the virus—how the virus spreads, how long it takes a person once infected to become sick, what options there are for treatment, and so on. These were the same few bullet points, in every single article, so that lay readers may always have some contextual information in the background (like the fact that the current strain isn’t airborne, an important issue that needed to be addressed constantly in the early months of the outbreak). I don’t see why we couldn’t do the same for every health article with “strength of evidence” bullet points.
† Recall that the recommendation for biennial mammography in women fifty to seventy-four was given a grade of B, while the recommendation for particular women aged forty to forty-nine for any kind of mammography was given a grade of C, implying that there was a moderate to significant amount of uncertainty as to whether there was any benefit at all for any woman in that age range, but it might be indicated for women at especially high risk, such as those carrying gene mutations that are known to increase the risk of breast cancer, such as the BRCA1 mutation.
* I can’t help but be amused by this last objection because, as part of my work in an academic medical center, with multiple specialists coming and going on a daily basis, I have firsthand witnessed the more traditional model in which attending physicians hurriedly sweep into a room, make a bunch of pronouncements about a patient’s problems that could be spoken just as easily in Sumerian as in English because one could hardly tell the difference, and proceed to sweep back out with no less alacrity, leaving patients and families behind in a cloud of mystified dust. It’s certainly possible that exposing families to medical or surgical ICU rounds in its raw state might cause people to get lost in the details, but from where I stand that’s probably nothing worse than the current situation.