August 18, 2014

Modernizing Medicare

The Medicare Modernization Act of 2004 didn’t really modernize Medicare. It took the important step of creating coverage for prescription drugs, which are the backbone of medical treatment of older individuals. But apart from this one change, albeit a major one, Medicare was pretty much intact. The problem with the approach to healthcare enshrined in Medicare, a program that turns 40 this year, is that it assumed that most illness is acute illness that requires hospitalization. In fact, it including coverage for anything other than hospital care in Medicare was more or less an afterthought. But today, most illness is chronic illness and the sickest, costliest patients typically have multiple chronic conditions.

The good news, as summarized in a perspective in JAMA this week entitled “Optimizing health for persons with multiple chronic conditions,” is that Medicare has made a number of changes that move the program into the modern world. The article indicates that the government report, “Strategic Framework on Multiple Conditions,” released in 2010, articulated 4 goals: fostering health systems change, empowering individuals, equipping clinicians, and enhancing research. Funds from the Recovery and Reinvestment Act and the Affordable Care Act have gone a long way to supporting initiatives in each of these domains. For example, in the realm of health systems change, the Centers for Medicare and Medicaid Services (CMS) is financing integrated models of primary care; to empower individuals, government funds have supported widespread participation in Stanford’s Chronic Disease Self-Management Program; to help clinicians, many professional societies have issued guidelines that are modified for patients with multimorbidity; and to enhance research, the Patient Centered Outcomes Research Institute (PCORI) was created and is a primary source of research funding. But there’s a problem.

The problem is that the patients we really need to worry about—because they get poor quality care, they cost a great deal, and despite all the money we spend on their medical care, they don’t even get treatment that is consistent with their preferences—aren’t patients with 2 or more chronic diseases. As the authors of the JAMA article comment, we need to focus on patients who are at greatest risk of poor outcomes and high costs. The patients we need to pay most attention to are those who are old and frail. And while many people who are old and frail also have multiple chronic diseases, not all of them do; moreover, having multimorbidity is not sufficient to qualify as frail.

The trouble with frailty is that we don’t have a good way of measuring it. And the measures we do have, which include such things as how strong a person is or how quickly he can walk, aren’t anything that doctors routinely test for, let alone record in the medical record. Insisting that all doctors use electronic medical records won’t allow us to keep track of who’s frail if we don’t enter frailty (or the various measures that define it) into that record. There is now a way that primary care doctors can screen for frailty by asking patients 5 questions, and their answers, while subjective, seem to correlate quite well with more sophisticated tests for frailty. But right now, doctors don’t ask those questions (they include things such as “are you able to walk up a flight of stairs” and “are you exhausted all the time”). Unless we can determine who is frail, we are not going to be able to test the effectiveness of special programs intended to provide medical care for them.

I think a solution may be on the horizon—and it’s not shipping special calipers to measure grip strength to every primary care doctor in the country. The answer, I suspect, is to use a proxy measure for frailty. That surrogate measure is multimorbidity plus functional impairment. It turns out that the people who are hospitalized most often, who take the most medications, and who go to physicians most frequently are people who have at least two chronic diseases and who have trouble with one of their activities of daily living.  

We can refine this group even further by focusing on those who get help with a basic daily activity: 14% of the population or 42 million people have at least one chronic disease and a functional limitation, but only 14 million of them need assistance to get by, and of those, 8 million are 65 or older. And those who have 5 chronic diseases and get help (regardless of age) have average per capita health expenditures of $22,380 compared to $12,749 for those with 5 or more chronic diseases and functional limitation but no help and to $9,723 for those with 5 or more chronic conditions but no functional limitations. 

It’s time to track patients who are over 65 by the number of chronic conditions they have, whether they have functional limitations, and whether their functional limitations are sufficiently severe to require assistance. Only then will we able to determine if these measures define who is at highest risk for adverse reactions to hospitalization, for progressive disability, and for death, and who costs the health care system the most money. Only then will Medicare be able to pay for the right kind of care for all older people. Once we do that, we will have truly modernized Medicare.

August 15, 2014

Acronyms of Doom

When I was in medical school, U.S. hospitals were plagued by only one kind of “superbug” or antibiotic resistant bacteria. Methicillin-Resistant Staphylococcus Aureus, abbreviated as MRSA and pronounced “mursa,” was the Enemy and it had been around since the early 1960s. I remember the yellow precaution signs on the door of rooms housing patients infected with this organism and the ritual donning of a yellow gown and surgical gloves before entering those rooms. Staph colonizes the skin of healthy people; if it enters the body through a break in the skin it can cause a serious infection, and if the staph is resistant to what was previously the best drug for treating it, the patient can be in trouble.

Then in the late 1980s, along came another bad actor, Vancomycin-Resistant Enterococci (VRE). Enterococci normally inhabit the gastrointestinal tract; sometimes they escape and when they do, for example through fecal contamination of a wound, they can cause significant mischief. Enterococcal infections had come under control with the antibiotic vancomycin—until they developed resistance. Now, MRSA and VRE have been joined by a new threat: Carbapenem-Resistant Enterobacteriaceae (CRE). According to a study published this month, the rate of detection of this infection has jumped five-fold in 5 years. And the mortality from these infections ranges from 48% to 71%. The Centers for Disease Control and Prevention (CDC) in Atlanta took the extreme measure of classifying CRE as an urgent threat. Only 2 other organisms currently share this honor.

The newest superbug, like MRSA and VRE before it, is something that healthy people don’t normally contract. Its victims are patients in nursing homes and hospitals, especially people who are connected to a medical device such as a ventilator or a catheter (whether urinary or intravenous). Debilitated older people are at particularly high risk. The CDC offers a 4-prong strategy for attacking the problem (preventing infection in the first place, tracking resistant organisms, improving the use of today’s antibiotics, and promoting the development of new antibiotics). I suggest an additional strategy that is rarely discussed: keeping frail, old people out of the hospital altogether.

From the time that the hazards of hospitalization were first recognized 50 years ago, the main way doctors have proposed dealing with them is to try to make hospitals safer. Old people become confused in the hospital? Don’t give them sedating medicines that make them confused. Old people fall in the hospital? Use bed alarms and chair alarms to alert nurses that they are getting up. These tactics and others can be helpful, but they don’t eliminate the dangers of the hospital and some interventions, such as side-rails on hospital beds, increase rather than decrease risk. Similarly, our first impulse as we try to control superbugs such as CRE is to reach for the precaution gowns to keep nurses and doctors from spreading the germs. Because antibiotic resistant bacteria are so great a problem, we need to respond with a multi-prong strategy. So yes, educating physicians to use antibiotics judiciously (preventing the development of resistance in the first place) and encouraging pharmaceutical companies to design new effective antibiotics are important. But let’s not forget that in most cases, the patient would not have gotten the infection if he or she hadn’t been in the hospital—these are generally hospital-acquired infections, not the reason for the hospitalization. Sometimes, frail old patients can be treated satisfactorily outside the hospital. Finding an alternative to hospital care is a way to avoid a growing list of “adverse reactions to hospitalization,” including delirium (acute confusion), incontinence, falls, and all those acronyms spelling doom, MRSA, VRE, and now CRE. 

August 11, 2014

Pursuing Palliation

We’ve known for a long time that breaking a hip can be devastating for an older person. It often leads to a loss of independence and it's associated with a substantial increase in the risk of both nursing home placement and death. While the rate of hip fractures declined in the US between 1990 and 2010, it remains stubbornly high: every year, 300,000 people fall and fracture a hip. A new study suggests the consequences of a hip fracture are especially severe in people who live in nursing homes.  In light of the high rate of disability and death in the 6 months after a hip fracture, an accompanying editorial recommends that breaking a hip in the nursing home signal the need for palliative care. Absolutely—but many of the patients reported in the new study should have been getting palliative care long before their fateful fall.

The authors of the study in JAMA Internal Medicine identified 725,000 Medicare fee-for-service patients who broke a hip during a 4-year period between 2005 and 2009; just over 60,000 of them, or 8%, lived in a nursing home prior to sustaining a hip fracture. It was these 60,000 who were the basis of the study:  the vast majority were women (75%), white (92%) and demented (91%). They also tended to have multiple chronic diseases and to have difficulty with many basic daily activities, which is hardly surprising, as their illnesses and their impairments were precisely why they lived in the nursing home.

After their hip fractures, many patients went downhill quickly. Fully 36% were dead within 6 months. Among those who survived to the 6 month mark, over a quarter had become totally dependent in walking. By the time a year had elapsed, just under half of the residents who had fractured their hip had died. Only 1 in 5 of those who started out being fairly independent in walking had regained their previous level of function. Risk factors for death or disability included male sex, advanced age, white race, multimorbidity, cognitive impairment, and dependence in basic activities. Of note, patients treated non-operatively (11.8% of the sample) did particularly poorly in terms of physical functioning and survival.

These findings fit with the results of earlier, smaller studies. But I couldn’t help asking, as I read this dismal portrait of life-after-hip-fracture-in-the-nursing-home, what happens to people who live in a nursing home who don’t break a hip? How many of them decline? The study in JAMA Internal Medicine had no control group, no nursing home residents with a comparable degree of physical and mental impairment, cared for during the same time period, who happened not to sustain a hip fracture.

I turned up another study also published this year entitled “Natural course of dependency in residents of long-term care facilities: prospective follow-up study.” The study was carried out in Dutch nursing homes and it included both people who had hip fractures along with those who did not. The population was reasonably similar to the US nursing home Medicare population: 75% were women and their mean age was 84. The Dutch have a system for reporting dependency that is different from (and more sophisticated than) the American approach: they report on the “Care Dependency Scale (CDS),” a 15-item scale that grades patients from 15 (totally dependent) to 75 (almost independent in eating, dressing, walking, and dressing). What they found was that among 890 Dutch nursing home residents, just under 15% had died at 6 months and another 18% had died by one year. But the degree of dependency at baseline was strongly predictive of the outcome, with higher dependency leading to greater likelihood of death, even after correcting for gender, age, and the presence of diagnoses such as cancer or dementia. Among residents in the low CDS group (that is, the most independent people), 20% actually improved over a 12-month period and the vast majority (80%) remained unchanged. Among residents in the highest CDS group (the most dependent people), no one improved, 64% remained stable over 6 months, and 36% got worse. The middle group, not surprisingly, fell between these extremes.

How did the nursing home residents in the Dutch study (some unknown fraction of whom fell and broke a hip) compare to the nursing home residents in the American study (who were selected based on their having fallen, broken a hip, and been hospitalized)? Those with hip fracture did worse: over the course of a year, 80% deteriorated markedly in their independence, compared to only 36% of the worst-off Dutch residents; and by 6 months, 36% had died, compared to about 20% of the worst-off Dutch.

These comparisons are approximate at best, but they give some hint of what the Medicare nursing home residents might have expected if they hadn’t broken a hip. So while a hip fracture does represent a turning point, admission to a nursing home in the first place also represents a turning point, a transition to life’s final chapter. Thus while it is eminently reasonable to offer palliative care alongside of surgery to all nursing home residents who break a hip, it would be even better to offer palliative care alongside conventional medical care to everyone in the nursing home.

August 03, 2014

The Robot Will See You Now

“Disease management” is all the rage. The reason? In the US and other developed countries, most diseases are chronic rather than acute and chronic diseases can generally be treated but not cured; ergo, they need to be “managed.” Interesting that it’s the disease that’s supposed to be managed rather than the patient. In any event, despite the popularity of the concept, disease management programs haven’t proven terribly effective: one review of 35 Medicare-funded projects in 22 states involving 300,000 patients showed that most did not improve the quality of care or reduce the cost of treatment.  Nonetheless, with the continuing growth in the number of people with chronic diseases—68% of Medicare patients have at least 2 chronic diseases and 14% have six or more—disease management remains popular as a common-sense approach that seems as though it ought to work. And a study in this week’s Annals of Internal Medicine suggests that maybe it does.

The article is a “systematic review and meta-analysis,” or an attempt to get at the truth by combining findings from many different studies. As an aside, the authors initially identified nearly 3000 articles about disease management, but ended up analyzing only 18 since the overwhelming majority were not of sufficiently high quality to be worth including. These 18 clinical trials all used “nurse-managed protocols” to guide the outpatient treatment of such diseases as diabetes, high blood pressure, and elevated cholesterol. The results? Use of these protocols by nurses led to small but statistically significant falls in blood sugar levels, blood pressure, and LDL (the “bad” cholesterol). The article concludes that nurses “are in an ideal position to collaborate with other team members in the delivery of more accessible and effective chronic disease care.” In other words, nurses should take over the function of managing chronic disease.

I have no reservations about handing over large chunks of primary care medicine to nurses—in fact, I think that nurses, particularly nurse practitioners, are better suited to primary care than many physicians and should do more than just follow flow charts. If all that's needed is to stick to clearcut guidelines, then a smart machine would be better than a nurse. But I do have concerns about the mindless application of algorithms to patient care, at least for older patients with multiple chronic illnesses. In younger, less complicated patients, preferably those with only a single disease that requires managing, simply following the optimized treatment strategy is likely to be a good idea. But in older patients with “multimorbidity,” as having multiple chronic conditions is increasingly called, practicing algorithmic medicine leads to disaster.

Suppose I’m seeing an 85-year-old woman with high blood pressure and diabetes and Parkinson’s disease. Call her Janet Dover. Optimal medical management of the high blood pressure means use of a diuretic. Tight control of diabetes means keeping the average blood sugar down, even if that means occasional dips to dangerously low levels. And control of Parkinson’s disease involves using a drug such as Sinemet. But there’s a problem with giving this patient a diuretic and a hypoglycemic agent and Sinemet all together, even if each individually would be a good idea. Both diuretics and Sinemet tend to make blood pressure fall when a person stands up and low blood sugar tends to make a person unsteady on her feet. So give Mrs. Dover all 3 medicines and the next thing you know, she will stand up, be dizzy and off balance—and then she will fall and break her hip.

Good geriatric care is all about figuring out how to treat someone like Janet Dover. It’s not easy and it depends both on her particular combination of diseases and on her willingness to make certain tradeoffs. But it’s not something that can be done by following a protocol, whoever is in charge of the protocol.



July 27, 2014

Kicking the Bucket

Even though you think and behave perfectly normally, you might already have Alzheimer’s disease. In fact, you might warrant the label of  “pre-clinical Alzheimer’s disease” for a full 20 years before graduating to the full-fledged condition. That according to the latest definition of Alzheimer’s put forward by the National Institute on Aging together with Alzheimer’s Association 3 years ago.

Heralded as a radical departure from earlier clinical diagnostic criteria, the new definition is in line with how we think about other chronic conditions such as cancer or heart disease. Normal cells often mutate into cancerous cells and proliferate for a long time before there are enough of the to produce symptoms—and a cancer diagnosis. Plaques may be deposited in the lining of coronary arteries long before they narrow blood vessels enough to impair blood flow to the heart resulting in chest pain. In all these cases, the hope is that we may be able to intervene during the pre-symptomatic phase, altering what is currently the inevitable trajectory of illness. In the case of Alzheimer’s, detecting various “biomarkers” in the blood or cerebrospinal fluid would signal the need for some kind of biochemical intervention. The problem has been that we don’t as yet have any such intervention. For now, the main purpose of early diagnosis is to identify people who could enroll in research on the prevention of Alzheimer’s—assuming the diagnostic techniques are sufficiently reliable and the benefits of research outweigh the burdens of knowing what the future has in store.

Philosopher Dena Davis makes a different case for the value of early diagnosis in this month’s Journal of Medical Ethics. Knowing that a person is destined to develop dementia, she argues, is invaluable so he can kill himself.

The option of “rational suicide,” as Davis calls it, is based on wanting to avoid becoming dependent and on wanting to spare family the burden of caregiving, as well as the potential loss of an inheritance. The diabolical nature of dementia is that if you wait until you are a burden, until you can no longer do any of the things that once gave meaning to your life, then you will most likely have reached a stage where you have also lost the ability to take your own life or even to remember that you ever wanted to end your life if you developed advanced dementia. And those countries (such as Switzerland, Belgium,  and the Netherlands) and those American states (Montana, Oregon, Washington, and Vermont) where Physician Assisted Suicide (PAS) is legal all require that the patient be of sound mind at the time the request is made and that she have a prognosis of not more than 6 months. So PAS is not generally an option for people with dementia.

There is a gentler alternative to pre-emptive suicide, and that is drawing up an advance directive that authorizes the withholding of life-prolonging medical treatment in the event of dementia. But Davis discounts this option, claiming that courts and ethicists increasingly dispute the right of a person, while cognitively intact, to make any claims about the interests of his future demented self.

In fact, opponents of pre-emptive suicide such as Rebecca Dresser, whose response to Davis appears in the same issue of the ethics journal, say something a little different. They do not assert, as Davis contends, that competent Jane Doe should have no say over what happens to incompetent Jane Doe. That would make little sense as much of the point of advance directives is to plan for possible loss of competence. Rather, Dresser argues that “precedent autonomy,” in particular the choices made when competent, should be tempered by considering the present interests of demented Jane Doe. This is the dilemma of the happy dement that Dresser addressed in an earlier article on the subject. 

Suppose that today I am an anxious, driven, competitive person. I find the prospect of being unable to write articles or give talks or care for patients intolerable. But suppose that one day, I do lose my ability to write articles and give talks and care for patients. And suppose that lo and behold, I am no longer anxious or competitive. Instead, I am relaxed and cheerful. I smile at everyone indiscriminately. I am no longer inhibited and I clap my hands and tap my feet to all kinds of music. I’m no longer a picky eater and instead I say everything I’m served, no matter how bland or poorly prepared, is delicious. I don’t recognize my children but am glad to see them, just as I’m glad to see the aide who dresses me in the morning. Surely my best interest is in continuing to live as long as I can remain in my current state.

The challenge, as I think Rebecca Dresser understands and Dena Davis does not, is to balance the perspectives of Jane-Doe-ten-years-ago and Jane-Doe-today. How should we do this and who should be the arbiter? The way to proceed, I believe, is for former Jane Doe to identify the goal of care for her future self. Plausible goals are the prolongation of life, maximization of quality of life, or promotion of comfort. Death, parenthetically, is not a goal; it is a means to an end. Only rarely is it the only or the best means to achieving a legitimate goal. Most people, I suspect, would choose comfort as the goal in advanced dementia and maximizing quality of life as the goal in moderate dementia. All the physician needs to do is to translate that goal into practice. For an acute medical problem such as pneumonia, for example, comfort means using Tylenol to keep fever at bay and oxygen to ameliorate shortness of breath. Maximizing quality of life may mean using oral antibiotics—but not hospitalization in an intensive care unit with a ventilator for breathing.

We will never entirely be able to prevent people from committing suicide upon learning they are at substantial risk, if they live long enough, of developing clinical Alzheimer’s. After all, people can kill themselves when diagnosed with a disease such as metastatic cancer, even before they develop intolerable symptoms, though few do so. They could kill themselves prophylactically at age 65 or 70 to assure they will never get any awful disease, though most people do not choose this route.  Our responsibility, however, as clinicians, ethicists, and lawyers is to try to make sure that every individual can live as good a life as possible, whatever cards she has been dealt. That’s what a good society does.



July 20, 2014

Big Data, Small Data, Any Data At All

Why is it that I still remember that the formula for the volume of a sphere is 4/3π r3, which I learned in tenth grade geometry? And why is it that I never even heard of a p-value, the measure commonly used to assess whether a result is “statistically significant,” until I was in medical school? I haven’t had any occasion to compute the volume of a sphere since I took calculus in college, but I have to interpret statistical findings all the time. Something is not right here.

Understanding at least the rudiments of statistics matters—and not just to me, a physician who has to make decisions about how to treat patients by evaluating articles in the medical literature that rely on statistical methodology. Understanding basic statistics matters to everyone. You need to know some statistics to realize that it is more accurate to measure the population by using sampling techniques than by trying to count everyone. You need to know some statistics to understand why Nate Silver, with his FiveThirtyEight website, was so much more on target in his predictions about the 2012 presidential elections than anyone else. And you need to know some statistics to decide, as a patient, how to evaluate the options your physician presents you with.

Just this morning, I read an article in the first section of the NY Times “Study Discounts Testosterone-Suppressing Therapy for Early Prostate Cancer.” It turns out that millions of men with early stage prostate cancer, mainly men over the age of 65, have been treated with “Androgen Deprivation Therapy” (ADT), either by bilateral orchiectomy (surgical removal of the testes) or by drugs. A new study, published in JAMA Internal Medicine, concludes that ADT in such men does not prolong life. It does cause lots of side effects, ranging from osteoporosis to weight gain, to decreased libido, to diabetes. The article quotes one expert who was not involved in the study as saying that the findings were “eye-opening and even alarming.” According to editorial writers from the Dana Farber Cancer Institute, the treatment is a good candidate for inclusion in the “Choosing Wisely” campaign, a national effort to eliminate the use of “low value medicine;” that is, treatments that achieve little, given their cost. The article fits in nicely with a major theme of JAMA Internal Medicine, which has a section called “Less is More.” It’s a theme that resonates with me as well: I often argue on this blog that certain treatments, especially when provided to frail, older individuals, may cause more harm than good. Finding that a commonly used treatment, such as ADT in older men, doesn’t do what it promises, would not be at all surprising to me. But is it true?

I looked up the article, which isn’t actually in the print issue of the journal yet; it was published in the “online first” section, which gets important articles distributed quickly. The authors looked at data on 66,717 men age 66 or older with localized prostate cancer diagnosed between 1992 and 2009. They defined “primary ADT” as orchiectomy or the use of a drug such as a luteinizing hormone releasing agonist (a drug that stimulates the pituitary to signal the testes to make testosterone until they run out, at which point testosterone levels fall) as the sole cancer therapy given to men with localized prostate cancer within 6 months of diagnosis. The outcomes they were interested in were cancer specific mortality (that is, the death rate from prostate cancer) and overall mortality. So far so good. 

But since this was not a randomized study in which some men got ADT and others received conservative management (ie no treatment unless symptoms develop), with the selection made based on the flip of a coin, there was no reason to believe that the two groups of men would be similar to one another. In fact, they were quite different. The men who got ADT were a good bit older than those who did not (average age 79 vs 77). They were considerably sicker, with higher rates of other diseases such as heart disease or lung disease. And they were far more likely to have “high risk” prostate cancer, based on the characteristics of the cells in their tumors (47.7% vs 23%). Their PSA scores were also much higher (an average of 19.5 in the ADT group compared to 11.1 in the other men, where 4 is the typical cutoff for normal). Simply comparing the outcomes in these 2 very dissimilar groups of men would not tell the whole story. Somehow, the authors needed to try to compensate for the inherent differences between the men. The only way to do that (other than scrapping this approach entirely and randomizing men to get ADT or some other treatment), is to build a statistical model.

Build a model the study authors did. The specifics of what they actually did are too complicated to describe here. I’m not sure I fully understand what they did, but it involved a technique called “Instrumental Variable Analysis,” known as IV. Suffice it to say that when they used this approach to try to adjust for all the differences between the groups (only some of which they could specify), they concluded that the 15-year prostate cancer specific survival rate was 85.4% in both groups. And when they used a different method, the Cox multivariate model, they found the mortality rate was 2.4/100 in the ADT group and 1.1/100 in the group treated with conservative management or, after attempting to adjust for differences based on what was known about other illnesses, PSA levels, etcetera, the group treated with ADT was 1.53 times more likely to die.

What the reader needs to understand is that the results of the study depend entirely on which model you choose. If you select IV, and the authors try hard to make the case that this is an excellent choice, but which some experts think is a flawed approach, you find that ADT and conservative therapy are equivalent. If you select the more conventional approach, you find that ADT is actually worse than watchful weighting. Since neither model predicts that ADT is better than conservative management, perhaps it follows that ADT is just a bad choice for the treatment of early prostate cancer in older men. The right conclusion, I think, is that we don’t actually know what to make of ADT. If we chose yet another model, perhaps we would find that ADT is superior.


Learning about different study designs—which ones you can trust, which ones are merely suggestive and which have to be confirmed using a better, more reliable approach—is what kids should be learning in high school and college. Learning about probability and statistics is what kids should be learning, not trigonometry and solid geometry. Our math curriculum reflects seventeenth century mathematical knowledge (it typically includes elementary algebra, Euclidean geometry, and perhaps calculus, created in the fourth century BCE and the seventeenth centuries respectively). 

Today, big data is all the rage and there is a growing enthusiasm for learning how to milk large data sets for useful information. But the reality is that it’s not just big data that’s important and it’s not just important for a small cadre of people. We all need to learn how to make sense of what we read in the newspapers, of what our doctors tell us about different treatments. And to do that, we need to develop basic statistical literacy.