Quack, Quack--if it Sounds Like a Duck....

When Lydia Pinkham (1819-1883) began selling her patent medicine in the mid-nineteenth century, she advertised it as a panacea for all sorts of “female complaints.” Whether you suffered from “neurasthenia,” from menstrual cramps, from infertility, or from postmenopausal depression, Lydia Pinkham’s Vegetable Compound was the drug for you. It made Mrs. Pinkham and her descendants a bundle: as described by James Harvey Young in his book, Toadstool Millionaires, the remedy grossed $300,000 in the year of her death and in 1925, it earned a profit of $3.8 million. I’m not sure what that is in 2017 dollars, but it’s a lot.

Other patent medicines made similarly broad claims of effectiveness and were likewise lucrative for their developers. Warner’s "Safe Cure" stated on the label that it treated “Bright’s disease (nephritis), urinary disorders, female complaints, general debility, malaria,” and “all disorders caused by disordered kidneys and liver.”

These potions remained unregulated until the passage of the Pure Food and Drug Act in 1906, at which point the labels were required to be “truthful” about their ingredients. In particular, 11 dangerous substances including morphine, cocaine, and alcohol, which were frequently found in patent medicines (even in cough syrup for children) needed to be explicitly listed. There would be no evidence of efficacy required for another 32 years.

In the intervening years, both physicians and the public gradually began to appreciate that any drug that was touted as the treatment or even the cure for as many as ten distinct diseases was almost invariably a fraud and the disseminator of claims of its miraculous properties a quack. No one drug can plausibly have so many unrelated beneficial properties. Today, we are afflicted with the inverse of the one-drug-many-cures scam. We are bombarded with claims that a single disease can have many unconnected causes. And the disease for which such assertions are most commonly made is Alzheimer’s disease.

A few years ago, the media was all riled up by an article purporting to show that the anti-anxiety drugs, benzodiazepines,  were associated with an increased risk of Alzheimer’s disease. A year later, we learned that anti-allergy medicines and some antidepressants had also been associated with developing Alzheimer’s. Last year, the culprit was the class of anti-ulcer medications, proton pump inhibitors (drugs such as omeprazole). And now, instead of a medication, we have soda and other sugar-laden beverages allegedly leading to stroke and dementia.

Really? Do all these substances launch innocent people on the path towards dementia?

To be fair, there are disorders with multiple well-established risk factors. Coronary artery disease, for example, is associated with elevated cholesterol, diabetes, high blood pressure, and smoking. But in this case, we understand something of how the disease develops and how each of these risk factors affects that pathway. High blood pressure damages the lining of the coronary arteries, and the resulting areas of inflammation tend to trap cholesterol, producing plaques, etc. 

There are also single substances that produce diffuse toxicity: cigarette smoking has been associated with lung cancer, bladder cancer, and cancers of the head and neck (three very different types of cancer), as well as with heart disease and stroke. But again, we understand the mechanism of action that is responsible for the disparate effects. 

In the case of dementia, the connection between benzodiazepines and dementia, allergy medicine and dementia, proton pump inhibitors and dementia, and now soda and dementia, is purely statistical. Nobody has made a plausible argument for how each of these agents might work to trigger Alzheimer’s—even though we now know quite a bit about how the disease develops. No one has made a good argument because they haven't found one.

Nor is there a persuasive statistical argument. The studies based on which all these factors have been implicated in causing dementia are retrospective, non-randomized studies. The authors try to control for various “confounding factors” that might be the real explanation for the association, but they might not know what the relevant factor is or they might not be able to figure out if it was present or not. For example, maybe people who were ultimately diagnosed with Alzheimer’s disease were more likely than others to have taken benzodiazepines a few years before their diagnosis because they were already exhibiting very early symptoms, and those symptoms created anxiety.

Much exciting research is underway on Alzheimer’s disease, research that may someday result in treatment, prevention, or even cure. But fishing expeditions to come up with a commonly used drug or other substance as an explanation of this complex disease are a distraction. We don’t need people abandoning their ulcer medication or their allergy treatment out of an ill-founded fear that they are bringing dementia upon themselves. Stirring up hysteria—and potentially depriving individuals of good drugs—is a really bad idea.

-->

-->

Counting what Counts

A confession: I’m a data junkie. I don’t generally collect data (though I have carried out a few empirical studies) and I don’t analyze data statistically. But I am fascinated by descriptive data, which often provide  remarkable insights into how things were or how things are. 

When I was working on my latest book (Old and Sick in America will come out in October), and I wanted to know what nursing homes were like on the eve of the introduction of Medicare, I consulted the report, “Characteristics of Residents in Institutions for Aged or Chronically Ill: 1963,” put out by the US DHEW (as the Department of Health and Human Services was called then). I learned there were just over half a million people living in 16,370 nursing homes (just about the same number of nursing homes we have today) and that they stayed there, on average for 3 years. 

When I wanted to know what hospitals were like in the 1960s, I read “Trends in Hospital Utilization: US 1965-1986” and found that circulatory disease, which has been the number one reason for hospitalization from the 1980s to the present, was only number 6 in 1965. Over the period from 1970 to 1986, the number of catheterizations done each year in people aged 65 and older would soar from 8,000 (or 3.8/10,000 people) to 275,000 (or 85.8/10,000) and that CABG (coronary artery bypass surgery) would likewise jump from 0 to 125,000 (or 42.9/10,000). So periodically, I check whether the National Center for Health Statistics has released any “Data Briefs” about older people. This past February, the agency published “Emergency Department Visits for Injury and Illness Among Adults Aged 65 and Over, US 2012-2013.” It is a compelling reminder that the emergency department is a key source of health care for older people.

Focusing exclusively on illness (as opposed to accidents), the report finds that the elderly go to the emergency room often and the older they get, the more often they go. Each year, 29 percent of people aged 65-74 have at least one emergency visit, as do 42 percent of those aged 75-84, and 57 percent of those aged 85 and up. About a third of these older patients arrive in the emergency room by ambulance—highlighting the role of ambulances as another locus of health care for this population.

These numbers don’t tell us what actually happens to older people when they reach the ER, but other data give a few clues. In the ER, the elderly are very likely to have some kind of imaging procedure (63 percent do), with about half of those getting a plain X-ray and a quarter getting a CT scan. And fully 32 percent of those presenting to the emergency department with an illness are admitted to the hospital; 5 percent to a critical care unit. There’s much that is left out if we focus only on descriptive statistics: we don’t know whether these patients typically have a friend or family member with them; we don’t know if anyone asks about their home situation; we are in the dark about whether anyone addresses their goals of care or checks if they can walk or determines their mental status. But the numbers are a place to start.

What is abundantly clear is that with 15.5 million visits to the emergency department by older people every year, it’s high time we pay more attention to what actually goes on there. We have the opportunity to figure our whether the hospital is the right place to take care of whatever the problem is and, if not, how to shore up the home environment to make it a viable alternative. To answer these questions, we need to make certain that older patients routinely undergo a brief assessment of both their cognitive and physical functioning. We need to involve a family member if support will be needed at home. I would bet that if we did all this, we’d make far more headway in avoiding hospitalization and decreasing the rate of readmission than many of the elaborate transitional care programs operating today.

-->

Just the Facts

This week’s JAMA included an article that describes prescription drug use in the US and compares current usage to the pattern twelve years earlier. The data for older patients is striking: in 1999-2000, 84% of people took at least one prescription drug; in 2011-2012, fully 90% of people took a prescription drug. More impressive, in the earlier period, 24% of older individuals took at least 5 prescription drugs (the cutoff for “polypharmacy”), while today it’s up to 39%. Is this good news or bad news?

One of my colleagues, who is an expert on drug policy, thinks that on the whole, more is better. Treatment of high blood pressure prevents strokes; treatment of high cholesterol prevents heart disease; antidepressants prevent suicide. My colleague worries about barriers to access for these life-saving and/or quality-of -ife-enhancing medications. The finding that more people are taking such drugs is encouraging: perhaps high cost or lack of primary care are not preventing patients from getting the medications they need. Other colleagues, who are geriatricians, think that on average, less is better. Polypharmacy is dangerous in frail, older individuals. The side-effects of medications, such as dizziness causing falls or low blood sugar causing fainting, may outweigh their benefits, especially when drugs are taken in combination.

The truth is that medication use in general and polypharmacy in particular are neither inherently good nor bad. Whether a given medication should or should not be prescribed depends very much on the circumstances. What the new report does is to describe patterns, allowing and encouraging us to zero in on classes of drugs whose use has changed markedly over time.

Particularly intriguing were the few drugs whose usage fell over the 12 year period studied. Almost all drugs were used more commonly. Among patients over age 65, only two categories of drugs fell to a statistically significant extent: sex hormones (almost exclusively estrogen replacement therapy) and anti-arrhythmics (drugs that were once widely used to try to control irregular heart rhythms). Now this is interesting because it’s very hard to get doctors to stop prescribing medication and to persuade patients that the medicine they've been taking for years, a drug they are convinced is beneficial, actually is useless or even harmful. But in the case of estrogen, new studies convincingly demonstrated that what was previously received wisdom—estrogen replacement in women prevented heart disease—was wrong. And in the case of anti-arrhythmics, drugs like quinidine and procainamide, which were once used to suppress abnormal ventricular rhythms, it turned out these medications were associated with an increased risk of sudden death. These drugs have virtually disappeared, replaced by other classes of drugs entirely. What these examples demonstrate is that when new data become available that are truly compelling, physician and patient behavior can change. It’s particularly helpful to be able to offer an alternative—a beta blocker instead of quinidine—rather than to subtract one drug without adding anything in its place.

The other takeaway message for me from this study is the tremendous importance of the National Center for Health Statistics. The data presented in the JAMA article come from NHANES, the National Health and Nutrition Examination Survey. This is a representative survey of a large number of Americans (37,959 in 2011-2012), conducted every few years, that gathers information about health and diet. It is, as the JAMA authors say, “a stratified, complex, probability-based survey” that over-samples older adults, low income individuals and certain racial and ethnic groups. Just what is the National Center for Health Statistics, the source of this treasure trove of data?

The National Center for Health Statistics was established in 1960. Since 1987, it has been part of the Centers for Disease Control (CDC). Its mission is to: “to provide statistical information that will guide actions and policies to improve the health of the American people.” In addition to conducting the NHANES surveys, it has three other data collection programs. There is the National Vital Statistics System, which gathers information about births and deaths and generates information about life expectancy. Then there is the National Health Interview Survey, which interviews Americans about their heath insurance coverage, health care resource utilization, and immunization status, along with other aspects of health. And there is the National Health Care Survey that studies the organizations that provide health care, such as hospitals, hospices, and nursing homes.

The studies carried out by the National Center for Health Statistics are invaluable. No private organization can be relied upon to carry them out systematically, regularly, and reliably. In a time when the federal government is under relentless attack, it’s worth drawing attention to some of its most remarkable—and unsung—achievements.

They All Add Up

With so much attention rightfully devoted to big ticket items in medicine such as the new drug for hepatitis C that costs $1000 a pill or high tech devices such as the continuous flow left ventricular assist device, which costs on average about $200,000 to insert, not many people are talking about the little ticket items. But the reality is that spending a small amount per person on a huge number of people adds up to just as much—or maybe more—than spending an enormous amount  per person for just a few individuals. So I was very pleased to see a research letter in JAMA Internal Medicine about that lowly test, the urinalysis. 

I was pleased that the authors looked at the consequences of the rampant ordering of urine tests in people with no symptoms suggestive of either an infection in the bladder or kidneys or acute kidney dysfunction, the only circumstances in which urinalyses have been found to be meaningful. The reason, quite simply, that most urine tests are useless or, as the article suggests, actually harmful, is that the majority of older people have bacteria in their urine. What this means is that the injudicious ordering of a urine test will far more often produce a “false positive” result than a “true positive.”

It so happens that twice in one morning of seeing patients this week, I was asked to order a urine test for no good reason. To be fair, the well-meaning daughters of the patients who requested the test, quite insistently, I might add, thought it was with good reason. Their mothers were being diagnosed with dementia, a condition that had developed insidiously over a period of at least a year and probably several, and they were hoping I would identify a “reversible cause” of this otherwise progressive, ultimately fatal illness. Neither patient, however, had any symptoms to suggest a bladder infection: they did not have burning on urination, they did not have urinary frequency, they had no fever or flank pain. One lady was 96; the other was 91. Since the majority of elderly women have bacteria in their urine, I was concerned that if we got a sample from these two (no mean feat if we wanted a “clean catch” specimen, uncontaminated by bacteria from the surrounding skin and from stool), it would show bacteria. But if we did anything with the result—and what was the point of getting the test unless we were planning to treat the ladies in the vain hope that a course of antibiotics would cure their dementia—we would do little more than expose them to a risk of another problem such as clostridium difficile colitis, a potentially serious, occasionally lethal infection common in debilitated older people that typically results from killing off other bowel bacteria with antibiotics.

So what did the new study find? The authors looked at 403 consecutive adult patients admitted to the general medical service of a hospital in 2014 and 2015. They found that in this group, who somewhat surprisingly had a median age of 79, 62% had a screening urinalysis at the time of admission. Fully 84% of these individuals lacked any symptoms suggestive of a urinary tract infection. Of the asymptomatic patients who were screened with a urinalysis, 30% had a positive test. Of those with a positive test, 22% were treated with antibiotics.

Maybe this is actually reassuring: only 30% of asymptomatic patients had bacteria in their urine, not the 90% the authors quote from the literature. And only 22% of the asymptomatic patients with a positive test were given antibiotics, not everyone. 

The research letter in JAMA Internal Medicine leaves many questions unanswered. We don’t know why so many asymptomatic patients had a urine test ordered—perhaps the physicians believed that the fall or fainting episode that triggered the hospitalization was in fact caused by a bladder infection, which is conceivable, even if dementia (what my patients suffered from) is not. We don’t know what proportion of those who were needlessly treated developed complications because of the antibiotics they received. We can’t measure just how much the injudicious use of antibiotics in situations such as this contributes to the development of bacteria that are resistant to multiple antibiotics, bacteria that go on to cause real disease that is phenomenally difficult to treat. 

We do know that there are over half a million people age 65 and over hospitalized each year according to the National Hospital Discharge Survey. If over half of them have an unnecessary test, and if a third of those tests are positive, and a fifth of those positive tests lead to potentially risky treatment, that’s still a lot of bad decisions. All those small ticket items add up, and we need to pay attention to the little decisions we make every day, not just to the big decisions we make once in a while.

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.

Down for the Count

The US Census Bureau released a new report this week about our aging population. Actually, it came out with 4 separate estimates, each based on slightly different underlying assumptions about developments in the next 40 years. The age structure of the population depends on 3 factors: fertility, mortality, and immigration. It turns out we can be fairly confident about the first two; the last one is far more uncertain. What was fascinating about the report is what most news media didn’t say about it.

Many major news outlets didn’t mention the report at all. I suppose it’s not exactly newsworthy, in that it’s just an updated version of earlier projections, based on the latest available statistics. The NY Times focused on 2 pieces of data: the absolute size of the elderly population: 43.1 million in 2012 and nearly double that, or 83.7 million, in 2050; and the fact that while this may seem dramatic, the numbers are far more dramatic in other parts of the developed world. People over 65 will make up just over 20% of the US population in 2030 (the peak year of the elder explosion, when all the baby boomers will have turned 65),  but in Japan they will make up 33% of the population and in Germany 28%.  Reuters, as reported by Business Insider, also points to the absolute and relative size of the aging population in the US. In addition, it comments on the male/female ratio among old people: right now, 66.6% of Americans over 85 are women, but in 2050 the gap between the sexes will have narrowed, and 61.9% will be women. Finally, Business Insider notes that our society in general and old people in particular will be more racially and ethnically diverse in another 40 years.

What I didn’t see commented on was the changing old age dependency ratio (the population age 65 and over divided by the population from 18-64, multiplied by 100). Right now in the US this ratio is about 21, which means there are roughly 5 working age people to support each old person. In 2030, the ratio will be 35, or only about 3 working people per oldster. That means that every person will have to devote a larger fraction of his or her effort to providing for their elders. But these projections are for the “middle series,” the average of the various predictions. Immigration will play a critical role in determining the actually age structure of the US population over the next 40 years. It will dramatically impact the size of the working age population, the people who will be responsible for most of the economic output of the country—and for supporting the older generation. By 2050, the “High Series” projects 10 million more people age 18-64 than does the “Middle Series." This is because we’re likely to see only modest changes in either fertility or mortality rates in the coming years. The one area where we do have a choice—because it reflects political, not scientific factors—is immigration. So if we want to have a maximally productive economy, and parenthetically if we want to have a way to care for all those old people, we need to increase immigration. And we better start now.