July 16, 2018

Us or Them?

Nearly 30 years have passed since Congress passed the Patient Self-Determination Act, enthroning advance care planning as an important part of care in the last phase of life. While the frequency with which patients designate someone to serve as their proxy in the event of incapacity has increased dramatically and the use of various planning instruments such as living wills and instructional directives has also risen, in most settings no more than one-third of patients with serious illness actually have conversations with their physicians about their preferences. 

Assuming that it’s a good idea for patients with serious illnesses to speak with their physicians about their goals of care, whose responsibility is it to raise the question? Does it make more sense to concentrate on educating physicians to communicate well about goals or to focus on empowering patients? 

Over the years, different initiatives have targeted one population or the other. The Robert Wood Johnson’s “Project on Death in America,” a multi-hundred-million-dollar project launched in 1994, focused to a large extent on prospective patients. The AMA’s program, “Education on Palliative and End-of-Life Care” (EPEC), also introduced in the 1990s, is targeted entirely to physicians (in fact, the “P” in the title used to refer to physicians: when first launched, the program was called “Educating Physicians on End-of-Life Care”.) More recently, a series of videos by Angelo Volandes of ACP Decisions aims to show patients and their families what various medical treatments entail, while the “Serious Illness Conversation Guide” from Susan Block, Atul Gawande and others at Ariadne Labs offers a checklist of questions to help physicians structure their discussions.

So, who needs more attention, the doctors or the patients? A new study in Health Affairs may offer some clues. I should point out at the outset that the Health Affairs article does not seek to weigh in on the question of whether to concentrate on physicians or on patients when designing interventions to promote advance care planning. The aim of the study, as indicated by its ponderous title, “Factors Contributing to Geographic Variation in End-of-Life Expenditures for Cancer Patients,” is to understand why some regions of the US spend so much more on care near the end of life than other reasons, with no discernible difference in outcomes. 

There are methodological problems with this study, as with all studies that start with death and work backwards—it is possible that although the people who died had similar outcomes (perhaps not surprising, as they all died), other people who weren’t considered in the study because they didn’t die were more likely to do well if more money was spent on them. Nonetheless, we have two groups of people with advanced lung or colorectal cancer on whom very different amounts of money were spent—in the lowest quintile, the average outlay was $10,131 and in the highest quintile, the average expenditure was $19,318). Was the decision to spend more coming from patient pressure or was it something that physicians were pushing for?

What the authors found was that in the high spending areas, physicians were less knowledgeable about treating dying patients (by their self-report), less comfortable providing care to patients near the end of life, and had less favorable attitudes toward hospice than their counterparts in the low spending areas. The high spending areas also tended to have more specialty physicians but fewer primary care physicians, as well as fewer hospices (per capita) than elsewhere. But—and here’s the key—patients’ beliefs about what they wanted were no different in high and low-spending areas.

All the statistically significant findings were in the domain of physicians, not patients:

Now, this finding does not directly translate into the question of who drives decision-making near the end of life in general and advance care planning discussions in particular. But it strongly suggests that physicians play an outsized role in shaping what happens to patients. Many patients don’t have pre-determined preferences; their values do not unambiguously determine what kind of medical treatment they should get. Or, if they do have some idea of what would be best for them, they are nonetheless strongly influenced by the views of their physicians. If “shared decision-making” is to work, both partners need to be informed and on board—and the physicians are particularly in need of some attention.

July 08, 2018

Where We Die

Honoring patient preferences, which is shorthand for providing an approach to medical treatment consistent with what patients say they want, has become a fundamental tenet of American medicine. And one preference that the vast majority of Americans share, according to multiple studies, is the wish to die at home. 

Whether dying at home is actually what patients want when they are faced with impending death, rather than an abstract preference expressed when they are healthy, is another matter—and I’ve previously argued that as hospital-based palliative care improves and home-based palliative care places an ever-growing burden on families, hospitals have become more attractive as a site for dying. But what patients tend to agree on is that they don’t want to suffer as they are dying.

To the extent that hospital care entails interventions such as ventilators or ICU treatment or chemotherapy, patients are reluctant to subject themselves to this type of care, especially if its likelihood of prolonging life is remote. To the extent that fewer hospital deaths and more home deaths is a marker for less suffering at the end of life, change in this direction is very desirable. A new study by Teno and colleagues shows we are continuing to make progress in this direction.

A few years ago, Teno et al performed a similar study comparing the experience of patients in 2000 to the comparable experience in 2009. What they found then was a marked decrease in the percentage of elderly Medicare fee- for-service decedents (health-policy-speak for people over age 65 with conventional Medicare who died) who expired in the hospital (32.6 percent vs 24.6 percent). Over the same period, however, they found ICU use increased among decedents in the last month of life (from 24.3 percent to 29.2 percent) as did the percentage of dying patients who underwent a transition of care (nursing home to hospital, for example) in the last 3 days of life (10.3 percent to 14.2 percent). The current study updates these findings by extending the period of analysis to 2015 and by adding data from older patients enrolled in Medicare Advantage programs, who now account for 30 percent of the Medicare population.

What they discovered today was that the proportion of hospital deaths among the fee-for-service group has continued to fall, going from 32.6 percent in 2000 to 24.6 percent in 2009 to 19.8 percent in 2015. ICU use in the last 30 days of life, which had risen between 2000 and 2009, remained stable at the 2009 level in 2015. Transitions to another site of care in the last 3 days of life, which had also risen between 2000 and 2009, went back down in 2015 to the same level as in 2000. And the chance of being enrolled in hospice at the time of death rose from 21.6 percent in 2000 to 50.4 percent in 2015. When the investigators looked at a sample of Medicare Advantage patients, they found these individuals had the same experience in 2015 as their fee-for-service counterparts.

What does all this mean? I suspect what it means is that when we know with a high degree of certainty that someone is going to die in the very near future, we tend to focus on comfort care. If physicians, patients, and families recognize that death is imminent, hospitalization is relatively unlikely, as is ICU care or transfers from home or nursing home to another site of care. However, physicians often cannot be so certain that death is likely to occur in the next few weeks or months. As long as the usual strategy is to pursue maximally aggressive care until death is virtually sure to occur in the immediate future, and then to abruptly transition to care focused exclusively on comfort, the picture we see today is likely to continue.

There is another approach. That approach involves opting for a goal that is neither comfort only nor life-prolongation at any cost. Instead, maximization of function is paramount; treatment aimed at prolonging life is also acceptable, provided it will not affect quality of life in a major way. So, too, is comfort a goal, but only to the extent that it does not conflict with maintaining function. For people who are frail, extremely old, or both this alternative strategy translates into fewer hospitalizations, fewer ICU stays, and fewer transitions of care in the final stage of life, whether that period is measured in weeks, months, or even years. 

Medical treatment does not have to be all or none; there is something in between. It’s quite possible that many people would opt for this type of care—if only they knew it existed.

June 28, 2018

MedPAC Speaks

I once called MedPAC the most influential organization in the medical arena that most Americans have never heard of. That was in a blog post from April, 2009 called “Follow the Money.” I continue to find MedPAC—the non-partisan, independent congressional agency that advises on issues related to the Medicare program—an excellent source of information and analysis. Its 17 members, representing academia, public health, medicine, and industry, make recommendations to Congress for revisions to the Medicare program twice a year. The March report is called “Medicare Payment Policy” and the June report is “Medicare and the Health Care Delivery System.” This month’s report contains more than the average share of interesting material. 
There’s a section specifically mandated by Congress that analyzes the Hospital Readmissions Reduction Program (it found, on average, that the program led to a 3.1 percent decrease in readmissions for five major medical problems, although it couldn’t be sure that the program caused the drop). Another section addresses how to improve Medicare’s approach to paying for medical devices (among other tactics, limit the ability of physician-owned distributors through greater regulation and accountability).  There’s a chapter on Accountable Care Organizations (possibly effective in controlling costs and improving quality, but with various modifications proposed). But what intrigued me most was the final chapter on “Medicare coverage policy and use of low-value care.”
MedPAC defines “low-value care” as services (including treatments, medications, and tests) that are of “little or no benefit” and where the risk of harms exceed the potential benefits. That, of course, is a value-laden definition since it assumes “benefit” can be clearly established and quantified, ditto for “harms,” and that we know the probability of either benefit or harm. Nonetheless, a remarkable number of treatments, medications, and tests are widely agreed to fall into this category. They comprise the list of no-no’s compiled by specialty professional societies for the Choosing Wisely campaign which, in its first five years of operation, came up with 525 recommendations. 
For frail elders, incidentally, the relevant professional organization inventoried by Choosing Wisely inveighed against gastrostomy tubes in people with dementia, obtaining urine cultures for people without urinary symptoms, use of lipid lowering agents in people with a limited life expectancy, screening for breast, colorectal, or prostate cancer in individuals with a life expectancy of less than ten years, hospitalization without taking into account goals of care, and aggressive blood pressure regimens.
The first observation MedPAC makes about low-value care is just how widespread it is among Medicare fee-for-service beneficiaries. Using two measures developed by Aaron Schwartz, Bruce Landon, and colleagues, the commission estimated that between 25 and 42 percent of patients over age 67 received at least one value service per year, costing Medicare between $1.9 and $8.5 billion. Imaging tests, cancer screens, and other diagnostic and preventive tests accounted for the lion’s share by volume, with cardiovascular tests and imaging the major offenders in terms of cost.
MedPAC concludes by considering six “tools” that could improve the situation. The commissioners raise the possibility of expanding the use of pre-authorization, a much-despised approach used by prescription drug plans and hospitals to limit the use of expensive medications. This tactic would create administrative obstacles to the use of various tests and procedures—and would further the bureaucratization of medicine deplored by many physicians. 
They also propose implementing “clinician decision support” programs and provider education, claiming that inappropriate antibiotic prescribing has been reduced in this way. No one could disapprove of trying to make clinicians wiser and more knowledgeable—the question is how to achieve this. Many of the usual “decision support” and “education” programs have been of dubious benefit. 
Another option is increased cost sharing. This is a favorite approach of commercial insurance programs but mercifully has rarely been used by Medicare—it places the burden for deciding whether a test is worthwhile squarely on the shoulders of patients, often without providing them with the information necessary to decide. It encourages wealthy people to squander their resources on low-value tests since the cost-sharing involved is trivial for them, and it may distort the decision-making of the less affluent by leading to their accepting cheap but useless tests and rejecting expensive but useful ones. 
Then there’s the strategy of revisiting Medicare’s “national coverage decisions” periodically—for the small number of tests in which the choice about coverage is made at the national level. The problems with this approach are that it fails to address the vast majority of coverage decisions that are made at the local level and that the reality is that revisiting coverage decisions has historically led to Medicare’s paying for more procedures, not fewer.
That leaves only two options to MedPAC. The first is introducing new payment models in which providers are held accountable for both cost and quality. This is the rationale behind Accountable Care Organizations and it may be working, according to MedPAC. It’s not entirely clear since the organizations that were able to lower the use of low-value care by and large choseto belong to an ACO so there may be something about the organization, not the ACO structure. The second option is to introduce cost effectiveness into determinations of Medicare payment.
MedPAC doesn’t explicitly endorse balancing cost and benefit in determining whether Medicare will pay for a service and if so, how much it will pay. It is at pains to review the long and sad history of attempts to introduce cost into Medicare’s decision-making. In fact, MedPAC includes a 14-page “primer” on Medicare’s decision-making process, with several highlighted sections on the failed efforts in 1989 and again in 2000 to use some form of cost-effectiveness. Both attempts were effectively derailed by the device industry that lobbied heavily against a strategy that they knew would result in restrictions on their ability to sell any device that is “safe and effective” (the FDA standard) at whatever price they wish.
But why, if including cost in coverage decisions is out of the question, does MedPAC go on and on explaining what cost effectiveness analysis is, reviewing the ethical objections to the use of QALYs (Quality-Adjusted Life Years), and detailing the history of efforts to define “reasonable and necessary,” the phrase used in the original 1965 Medicare legislation to describe what Medicare must cover? Why all the verbiage devoted to cost-effectiveness analysis in the commercial sector and by the non-profit Institute for Clinical and Economic Review? 
I have to assume that the reason is that the MedPAC commissioners recognize that while better physician education and new improved ACOs may be of some use, the most effective strategy is for Medicare to refuse to pay for “low-value care” when that is a euphemism for “useless intervention.” It is true, as the report indicates, that there is no explicit statutory provision including cost as a component of Medicare’s coverage decisions. But neither is there any statutory prohibition. It is time to try once again to exploit the loophole, perhaps using the language of “value” that is currently in vogue—after all, who could argue against choosing treatment that is valuable?

June 25, 2018

Reforming Medicare: Enhancement or Evisceration?

Recent reports indicate that Congress will try to slash Medicare in order to balance the budget—making older people and disabled people shoulder the cost of its enormous tax cuts. The proposed plan, according to the Washington Post, would extract $537 billion dollars from the Medicare program over the next decade. At the same time, the budget passed by Congress and signed into law by the president in February created CHRONIC (Creating High Quality Results and Outcomes Necessary to Improve Chronic Care Act) which, the NY Times suggests, is a hidden jewel buried in the voluminous budget bill. Which is it? Is Medicare headed for enhancement or for evisceration?

Thus far, the cuts are theoretical (it’s not clear that the House Budget Committee will get very far with its recommendations) whereas the reforms are real—or will be when they go into effect in 2020. CHRONIC is to be lauded for accomplishing several important goals. 

First, the act recognizes that good outcomes among people with chronic conditions are contingent on what are not strictly medical services—as well as access to physicians, hospitals, pharmaceuticals, and medical devices. Wheelchair ramps and grab bars, as well as other products that lie outside the traditional definition of “durable medical equipment,” can in the future be paid for by Medicare—at least by Medicare Advantage plans that opt to offer them. This strategy does not go as far as the National Health Service did in the UK with its “personal health budgets” that allowed patients or their representatives to decide in concert with their physicians how to spend their share of the health care pie. The result—and the program, despite some vocal protests, has been so successful that it was recently expanded—is that patients with early dementia can choose, for example, to spend NHS money on creating a garden that will keep them engaged, potentially obviating the need either for medication to control symptoms of agitation or for institutionalization in a nursing home. Nor does the US strategy apply to traditional Medicare: in an effort to make Medicare Advantage plans, which currently have 19 million members, even more attractive (furthering the Republican goal of privatizing Medicare), only MA plans will be allowed to reimburse for these new supportive services. 

Second, CHRONIC permanently authorizes Medicare Special Needs Plans (SNPs) that cater to the highest risk Medicare beneficiaries including those living in institutions. These are special types of Medicare Advantage plans that offer enhanced integration and coordination of care, a critical feature for this complex population. 

Finally, CHRONIC extends its support of non-traditional forms of care, of which the allowance for grab bars was one example, to telemedicine (particularly relevant for homebound patients and in rural communities) and to home care (expanding the Independence at Home Program 50 percent from 10,000 enrollees to 15,000). 

So, what’s not to like? Two cautionary notes. First cautionary note: CHRONIC focuses overwhelmingly on Medicare Advantage plans (which currently cover 32 percent of Medicare beneficiaries), not on traditional Medicare (which covers the other 68 percent). This is no surprise, as the Republican Congress, which is interested in privatizing Medicare, sees shifting to the MA model as a route to achieving this goal. In principle, I don’t have any problem with expanding the number of MA plans (currently there are 3300, according to MedPAC, the Medicare advisory council) as they offer great potential for the coordination of care so essential to frail elders, but it will be essential to maintain the regulatory oversight of CMS if these plans are to be guaranteed to provide quality care. Moreover, we need to begin collecting detailed data on the utilization and outcomes of MA members. Right now, almost all of the voluminous data gathering by the federal government exclusively deals with fee-for-service enrollees so no granular analysis of the performance of MA plans is possible. 

Second cautionary note: while some of the provisions of CHRONIC appear to address programs, in fact the legislation is often grounded in how the programs are to be reimbursed. And the underlying philosophy is that the way forward lies with “value-based” care. I’ve blogged about this before, most recently in my post “V is for Value.” My concern about this approach is that it assumes that better and less costly medical care can be obtained simply through tweaking reimbursement. It’s the triumph of the economists’ view of health care as an industry subject to manipulation like other industries. The trouble with this insistence that VBP is key to all our problems, aside from the fact that so far value-based reimbursement systems such as pay-for-performance have not succeeded, is that it discounts the role of culture, advertising, and popular expectations. It is these factors, and not just payments to physicians, for example, that shape the enthusiasm for technology manifested by patients, corporations, and physicians alike.

CHRONIC is an admirable piece of legislation—for what it includes. What should concern us, however, is what it leaves out.

June 20, 2018

Much Ado About CPR

Two short essays in this week’s New England Journal of Medicine present differing perspectives on how to treat a desperately ill woman w in the ICU of an American hospital, a woman who is unable to make her own decisions and has no advance directive. One author responded that the right course of action was to “institute a DNR order” and the other that it was to “continue full resuscitative measures,” setting up a quasi-debate. I say a quasi-debate because the two authors, as per the editor’s instructions, simply laid out their own arguments without responding to the other’s point of view. I know, because I am one of the two.

I asserted that the physician should declare the patient DNR—a woman with no written advance directive and no designated health care surrogate who was dying of multiple organ failure triggered by sepsis and unresponsive to maximal medical therapy. Attempted CPR in this setting, I argued, was as close to futile as medical treatment ever is; moreover, the patient had clearly stated to a friend and neighbor that she would not want aggressive measures when facing overwhelming odds. I stand by my arguments, which I won’t repeat here, but I regret that the discussion focused on the wrong question.

The wrong question is whether the patient should have a DNR status; the right question asks about the general treatment strategy rather than a specific medical intervention and about the process for deciding, not just the outcome of the process. It moves the discussion beyond considerations of futility—a debated that raged in the 1980s and ultimately led to the recognition that futility is meaningless without reference to what treatment is intended to accomplish. Maintaining a patient who is in a persistent vegetative state on artificial nutrition is futile if the goal is to restore her to full functioning as a thinking human being. But it’s entirely appropriate if the goal is to sustain life in the sense of a heart that beats and lungs that respire. Discussing how to make medical decisions for patients who have lost capacity and how to think about treatment for those who are dying allows us to discuss much more important decisions than whether or not to attempt CPR.

Ever since 1984, when a New York hospital was found guilty of putting purple dots on patients’ charts signaling that they were not to be resuscitated—without their knowledge or input or that of their next of kin—we’ve been obsessed with DNR orders. I suspect that more ink has been spilled on whether, when and why to write a DNR order than on any other medical decision. The reason, presumably, is that it seems irreversible: if a patient sustains a cardiac arrest and CPR isn’t performed, the patient will surely die. Other seemingly momentous decisions often prove not quite so momentous—even Karen Ann Quinlan, the first patient to bring the possibility of not intervening medically to public attention, lingered for nine years after her “life-sustaining” ventilator had been disconnected. A choice not to resuscitate is far more unambiguous—although CPR, it should be pointed out, is hardly a guarantee of life.

Cardiac arrest may have special cachet because it seems to divide the living from the dead. But in fact, for older patients who are critically ill, it is often merely the last step in what is often a relentless progression of markers on the way to death. From the perspective of human suffering and of doing good (and avoiding harm), other steps along the way are often far more significant. After all, a person whose heart has stopped and who is not breathing is no longer able to experience anything, neither the existential angst from awareness of imminent death nor the physical discomfort of aggressive treatment. The nurses and doctors who attempt CPR in a dying patient may feel guilty of assault; they often regard intubation, chest compressions, and electrical shocks as an undignified way to end life, but they are the ones experiencing malaise, not the patient.

Subjecting a patient who is in the final phase of life to dialysis for four hours a day, three times a week, or to major surgery to repair a damaged heart valve, or to ventilator care for advanced lung disease, by contrast, may cause pain and suffering to the person receiving treatment. Both the potential benefit of treatment (longer life) and its potential burdens (ranging from delirium and functional decline to a shorter life) are difficult to assess quantitatively. These decisions, which are difficult to make with a competent patient and correspondingly more fraught with an incapacitated patient, are the ones we should be discussing, not CPR.

Ironically, despite the now 42-year history of DNR orders, there is still confusion about whether DNR indicates that CPR will not be performed in the event of a cardiopulmonary arrest (as it is defined by the American College of Cardiology and other authorities) or something more. Indeed, the opposing piece in the New England Journal of Medicine talks about a time-limited trial of “continued resuscitative measures,” which presumably refers to ongoing ventilator treatment for respiratory failure, dialysis for kidney failure, and fluids and antibiotics for sepsis, rather than for a few minutes of attempted CPR. This uncertainty about what DNR means reflects another problem with the obsession with “getting the DNR order,” and that is the widespread belief that treatment decisions must be all-or-nothing: if you don’t want CPR, then surely you don’t want any potentially life-prolonging medical interventions. The reality is that many people don’t want burdensome treatments that are extremely unlikely to be beneficial, but that doesn’t mean they want to focus exclusively on comfort. 

There’s much more to discuss—I haven’t even touched on the process for making medical decisions. That includes not only who the surrogate should be in the event of decisional incapacity and what standard that surrogate should use for making decisions (substituted judgment is the usual standard, with best interests the back-up if no information about the patient’s preferences is available), but also how the discussion should unfold. Typically, physicians are so eager to reach a decision about what to do that they jump to this step in the process without first clarifying the patient’s underlying health state (in the NEJM vignette, the crucial bit of information is that the patient was dying), as though preferences are independent of any broader context. But for now, let me leave you with the thought that both medical decision-making and advance care planning involve much more than checking off items on a menu, and that if we do engage in intervention-specific planning, CPR should be low on the list of what we discuss.

June 07, 2018

The Sting of the Jellyfish

Several years ago, the public health community launched an initiative to persuade pharmacies to stop selling cigarettes. Responding to steady pressure, CVS announced in 2014 that it was pulling all tobacco products from the shelves of its approximately 7700 stores. The result? Overall sales of cigarettes dropped, with fewer teens starting to smoke and more people quitting. The other chains, such as Walgreen’s and Rite Aid, have not yet followed suit, but the precedent has been set: retailers dedicated to the promotion of health should not be in the business of peddling death—and cigarettes are the leading cause of preventable death in the U.S., accounting for over half a million deaths per year.

In the same vein, we might ask why news organizations that are dedicated to ferreting out the truth continue to allow ads that peddle falsehoods. And that’s exactly what CNN, Fox News and MSNBC are doing when they advertise Prevagen, a protein that its manufacturer, Quincy Bioscience (a company that seems to be devoted almost exclusively to making and selling Prevagen) touts as “clinically shown to improve memory,” exhorting aging individuals to “try Prevagen for yourself today” in order to “support healthier brain function, a sharper mind and clearer thinking.”

These claims are untrue. The FDA issued a warning letter to the company in 2012, arguing that the company’s unverified health claims indicate that it is selling a drug and not a “supplement” (as the company calls Prevagen) and thus must go through the rigorous approval process for new drugs. Today, the manufacturer continues to call Prevagen a “supplement” because it is a synthetic version of a protein found in jellyfish and continues to make the same claims about its effectiveness, though with the disclaimer in small print that “these statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat cure or prevent any disease.” 

In 2016, the Federal Trade Commission’s Bureau of Consumer Protection and the NY State Attorney General filed a lawsuit arguing that, contrary to its manufacturer's claims, Prevagen has not been shown to improve memory in 90 days; it has not been shown to reduce memory problems associated with aging; and it has not been shown to produce other cognitive benefits.That lawsuit, incidentally, was dismissed because the judge failed to understand that the methodology used in the clinical trial on which the company’s assertions are based is irredeemably flawed. With the added support of AARP, the case has been appealed to the Second Circuit Court of Appeals. 

Why anyone should be motivated to take Prevagen based on its activity in jellyfish, which are not known for their cognitive abilities, is a mystery to me. 

Why anyone should think that the active ingredient, Apoaequorin, could possibly do anything when it is destroyed in the stomach by the enzyme pepsin and therefore doesn’t even make it into the circulation is another puzzle. And why anyone should believe that Apoaequorin could affect brain function (if it somehow managed to elude stomach enzymes and find its way into the blood stream) when it is 40 times too large to cross the blood-brain barrier is also unknown. But what is clear is that Quincy Bioscience has been successful in earning $165 million between 2007 and 2015. It is willing to spend big bucks on advertising on national television to promote its product because it knows the strategy works.

Followers of CNN, MSNBC, or Fox trust the news they hear from commentators on those stations. Because they trust the station, they tend to trust its advertisers. News programs have an ethical responsibility to vet the ads they run; they don’t have to like the products, they don’t have to want the products themselves, but they should be committed to fact-checking their veracity.

A separate question is why FDA regulation of “supplements” is confined to assuring that they are accurately labeled—that is, that the ingredients are correctly listed on the bottle. Manufacturers of dietary supplements do not have to prove them safe nor do they have to make claims about efficacy that are accurate or truthful. A related question is why products such as Prevagen can bypass the FDA's regulatory process for new drugs by calling themselves “supplements,” a term primarily applied to vitamins and minerals but increasingly used for all sorts of “herbs and botanicals” as well as other substances found in nature, or their synthetic analogs. 

In today’s anti-regulatory climate, we cannot realistically expect increased government protection for vulnerable consumers. We can expect—and demand—ethical behavior from any organization dedicated to promulgating the truth.

May 17, 2018

Can We Talk?

This week, the Massachusetts Coalition for Serious Illness Care released the results of a survey of state residents about advance care planning. The coalition wanted to know whether people were talking about their preferences for care near the end of life and if so, with whom—family? Physicians? The concern was also with whether people who had suffered a serious illness over the preceding year were more likely to have had such conversations and whether factors such as age, ethnicity, or gender influenced behavior. And regardless of whether people had discussed their views, had they designated a health care proxy, a substitute decision-maker in the event of incapacity?

On the assumption that more talk is better, it’s difficult to spin the results as encouraging. Massachusetts residents do not talk much to their doctors about such issues, even if they have had a serious illness and even if they are over age 65—although these two groups do have appreciably higher talk rates than their healthier or younger counterparts. On average, they are not very different from people in the rest of the country and the rates of conversation have not appreciably changed since the last such survey was administered in 2016.

In the population as a whole (the survey sampled adults over the age of 18), only 13 percent had had a conversation with a health care provider about their end of life preferences. Among those with a serious illness or over age 65, this rose to 27 percent.

Those surveyed were more likely to have chosen who would make decisions for them than to talk about what their surrogate should decide, with 63 percent of those over 65 having completed a health care proxy form. 

Perhaps what’s most interesting is that among those who reported they had had a discussion of their wishes, 59 percent said they had initiated the conversation, not the physician or other member of the health care team. Moreover, 34 percent of those surveyed said they did not want to have such a discussion with their doctor—they felt confident that their family members understood their wishes.

Before we get too carried away by the numbers, we should note that only about 210 of the participants were over age 65 (according to the 2010 Massachusetts census, 14 percent of the population was over 65; there were 1500 people sampled in the survey). That’s not many people on whom to base sweeping conclusions and this is the group for whom advance care planning is most essential.

Which leads me to my conclusion: advance care planning doesn’t make much sense in young people or healthy people. The likelihood that such a plan will have to be implemented is just too remote (Karen Ann Quinlan, Nancy Beth Cruzan, and Terri Schiavo notwithstanding—they are the exceptions, not the rule) and the chance that their perspective on what they would want if they were in an unimaginable future state reflects what they would actually want if such a situation came to pass seems slim. Efforts to think about future wishes make most sense for people who already have the condition that is apt to kill them in the future. Once they have started on a particular path, they are better positioned to think about the next steps on that journey.

Geriatricians learned a long time ago that “targeting” interventions to the appropriate population is crucial if they are to be beneficial (and cost-effective). Proponents of advance care should likewise concentrate their efforts on the people who stand to gain the most from discussions of the goals of care—and who are in a position to imagine their future. Even in this arena, we have a great deal of work to do. But spreading our efforts too thinly won't help.

May 07, 2018

To Fix or Not To Fix

Over the past ten years, people with advanced dementia have gradually come to be recognized as dying. Death may not come in a day or a month, but it will usually arrive within six months and almost always within a year. In recognition of this reality, growing numbers of physicians recommend a palliative approach to medical care for persons with advanced dementia, and increasing numbers of family members accept comfort care for their relatives. But what, exactly, is the best way to make someone with advanced dementia comfortable?

One common problem in this population, an event that is decidedly uncomfortable, is falling, especially when it results in a hip fracture. In nursing homes, which many family members assume is a perfectly safe environment, falls and fractures among individuals with advanced dementia in fact occur with some regularity. Individuals with advanced dementia by definition have poor judgment, limited awareness of the risks to their own safety, and may suffer from visual problems or gait abnormalities on top of their cognitive impairment. Nursing home staff members cannot provide constant one on one supervision. At the same time, walking around may be one of the few remaining sources of enjoyment for the person with dementia. If he or she does fall and does break a hip, and if the person’s family has indicated that comfort is the most important goal, how should that goal be translated into practice?

While we know that over one-third of nursing home residents with advanced dementia and a hip fracture die within six months of their fracture, what we don’t know is how operative treatment (hip fracture repair) compares to non-operative treatment (bed rest and pain medication) in terms of either comfort or mortality. Berry and colleagues attempt to provide some insight into this question in an article appearing in JAMA Internal Medicine

Because it’s not currently feasible to randomize patients to the two strategies, the authors emphasize, they resorted to administrative databases. They use information from the Minimum Data Set, an extensive database mandated for use in long term care facilities, together with Medicare claims data to define a population of nursing home residents with advanced dementia and a hip fracture and to look at what happened to them. So, what did they find?

Over a two-year period, they were able to identify 3083 nursing home residents with both advanced dementia and a hip fracture. Fully 85 percent of them underwent surgical repair; the remaining 15 percent were treated non-operatively. Those who underwent surgery were considerably less likely than those treated conservatively to die; early on, there was a 12 percent mortality rate in the surgical group compared to a 31 percent rate in the comparison group. The median survival among those with surgery was 1.4 years; the median survival in the other group was 5 months. 

Perhaps more interesting than survival, since family members tend to regard comfort rather than life-prolongation as the primary goal of care, is differences in pain control. This is especially important since one of the main reasons given for advocating surgery is to control pain. What Berry et al found was that at six months, 29 percent of the post-surgical group were in pain vs 31 percent of the non-operative group. While this difference was statistically significant, it’s actual clinical importance is less clear. Moreover, we really want to know how much pain was experienced in the immediate post-fracture period, probably the first 30 days, which the methodology used cannot tell us. The investigators also found that only 11 percent of those individuals who had been able to walk before the fracture and who underwent surgery were still able to walk six months after the surgery. Retaining the ability to walk is the other major reason for contemplating surgical repair of a hip fracture.

One problem with the study is that we don’t have any information about why surgery was not done in 15 percent of people. Maybe their physicians or their family members perceived them as closer to the end of life than their counterparts who underwent surgery, even if the two groups are indistinguishable in terms of measurable risk factors, and maybe they were correct in their assessment. Or, what is at least as likely, the families who advocated a non-surgical approach were also likely to favor withholding antibiotics or other treatments (with the exception of pain medication) in the ensuing months. It may well be that it was the decision to minimize all kinds of interventions, not just surgery, that led to a shorter life. 

A second problem is that we don’t know anything about the quality of life in the immediate post-fracture period. We know about pain, restraint use and pressure ulcers at six months—but not at one month, which is the time frame over which the treatment decision is most apt to matter. For all we know, people who don’t have surgery have better, more aggressive pain management in the weeks after the fall because physicians understand they are in pain, whereas they are notoriously poor at pain management after an operation, especially with patients who cannot articulate that they are in pain.

Berry et al deserve kudos for asking the right question—what constitutes good palliative care in advanced dementia? They've made a good start, but much more research is necessary to answer the question.

April 19, 2018

Medicare Part E?

Congress first created the Medicare hospice benefit in 1982. During the first full year after the law went into effect, only 200 Medicare patients enrolled in hospice, accounting for less than one-half percent of the deaths that year. By 1986, seven percent of Medicare patients who died were enrolled in hospice at the time of death; in 1998, it was 19 percent and in 2013 it reached 47 percent. A new report from the National Hospice and Palliative Care Organization shows this percentage has remained fairly stable: in 2017, it was 48 percent.

The length of time that Medicare patients are actually enrolled in hospice remains woefully short, particularly for some diagnoses: while the average length of stay is 71 days, reflected a few very long stay patients, the median length of stay is only 24 days. In fact, fully 28 percent of individuals got between one and seven days of care, far less than most clinicians regard as optimal. Patients enrolled in hospice because of a cancer diagnosis tended to have the shortest lengths of stay in the program. Nonetheless, the historical shift towards hospice as a reputable and popular approach to end of life care demonstrates that we can introduce a new Medicare benefit package and expect it to catch on. The same happened with Medicare Part D, the drug benefit (though this was an add-on rather than a substitution), with enrollment going from 52 percent in 2006 to 71 percent in 2016. Medicare Part C, the partially privatized HMO-style benefit now known as Medicare Advantage, has also taken off: in 2003, only 13 percent of Medicare beneficiaries opted for such a plan but by 2017, 33 percent did. 

The time is ripe for adding a new alternative benefit package to Medicare’s offering: whether we call it Medicaring, the name preferred by Joanne Lynn, Director of the Elder Care and Advanced Illness division of the Altarum Institute,
or Intermediate Care, my term, or something else, it would provide enhanced home care services for frail elders and others with advanced illness in exchange for limitations on access to intensive, expensive, burdensome, and seldom beneficial treatments such as bone marrow transplantation or open-heart surgery. 

Americans tend to dislike or at least be very skeptical about any kind of limits to care. But often the problem is that we focus on what people cannot have rather than emphasizing what they can have—and what they often need and want. In fact, I think one reason that patients only enroll in hospice very late in the course of their illness is that hospice is presented as the withdrawal of care, the discontinuation of care, rather than as the substitution of one approach to care for another. Many patients may be unwilling to forgo treatment altogether, which is how hospice is frequently presented, but are eager to choose a less invasive, less burdensome form of care: infusions of very toxic chemotherapy drugs in the physician’s office every week or two may be rejected if there is an oral medication they can take at home that has far fewer side effects. But if the only alternative is certain death, chemotherapy looks far more attractive.

Important and interesting work has been done on developing and implementing programs to care for people with advanced illness or frailty, such as the Coalition to Transform Advanced Illness Care's 2017 reportThe way forward is now to create a Medicare benefit to accommodate this approach.

April 02, 2018

Vavavoom--the Vital VA

The Veterans Health Administration, as its website reminds us, is the largest integrated health care system in the US, where “integrated” is a contemporary buzzword in health care. It denotes coordination across many sites and implies a focus on populations, not just individuals, both of which are generally viewed as essential features of a good health care system—and which are widely lacking in the US. The VA, as it is known, provides care to 9 million people at 170 medical centers and over 1,000 outpatient sites. It is also the only form of socialized medicine in the United States: the federal government does not just provide health insurance (as is the case with Medicare); it employs the physicians, nurses, administrators, and others who actually provide medical care. Not surprisingly, the VA is under threat from the current administration, despite its popularity with veterans and the organizations that represent them.

The goal of the current administration is to privatize the VA, and arguably the real reason the Secretary of Veterans Affairs was summarily fired is that he wasn’t moving rapidly enough in that direction. At issue may not be money—Trump’s proposed 2019 budget for the VA would be nearly $200 billion, or a 12 percent increase over the current year. Rather, the main concern is ideological—getting government out of health care. My question is: what would that mean for older people, many of whom receive VA benefits?

In many respects, the VA provides excellent care for older veterans. It has a long history of innovation in the geriatric realm: it has for years had a network of Geriatric Research and Clinical Education Centers (GRECC) that carry out research and provide clinical care to older people. Its Geriatric Evaluation and Management Units (GEMU) that offer inpatient geriatric assessment have been shown to slow the rate of functional decline compared to standard care. The VA has a model Extended Care program that provides room, board, and medical care to veterans needing long term care. And in recent years, it has developed an innovative home visit program for its neediest members. Are all these programs a fluke?

To understand, we need a little perspective. The VA health system was created by Congress in the immediate aftermath of World War II. It had several missions including: improving the health and function of US veterans, providing education and training to medical students and residents, and conducting research. But for many years, it was the poor stepchild of the federal government. The Secretary of Veterans Affairs did not become a cabinet level position until 1989. The system became known for inefficiency and mediocrity. But then in 1994, then Under Secretary for Health in the VA John Kizer initiated a radical transformation of the VA system. He created a new system of regional integrated service networks. He developed and implemented a system of quality indicators to drive improved performance.

The transformation worked. A 2007 study published in the Milbank Quarterly looked at performance over the 1998-2005 period and found that the VA outperformed traditional Medicare, Medicaid, and the private sector on a variety of indices.  

A RAND study published in 2016 examined the period from 2005 to 2015 and again found that the VA performance was comparable to or superior to standard American health care in most domains.

But there were cracks in the system. By 2014, those cracks split wide open, and a scandal over long wait times—and a cover-up to conceal the problem—erupted. As Kizer himself, the mover and shaker of the earlier reforms, argued in a New England Journal of Medicine essay in 2014 written with health policy expert Ashish Jha, the reforms have become unfocused and the bureaucracy bloated. The original idea of using quality indicators had mushroomed until there were so many indicators that both administrators and physicians were paralyzed; the earlier regionalization and centralization of the VA system, while effective managerially, had led to excess insularity of the entire system from new developments in the private sector (many of which may have taken their inspiration from the VA in the first place!) 

The solution they proposed was to fix those problems, not to outsource VA care to the private sector. Firing Secretary Shulkin, who by all accounts was a very competent and visionary leader, and replacing him with the president’s personal physician, Ronny Jackson who has no experience managing a large organization (let alone one as complex and needy of reform as the VA), is a classic case of throwing out the baby with the bathwater.