H. Gilbert Welch – The Health Care Blog

The Problem Is Relative

Nov 1, 2018

By H. Gilbert Welch

Numerous studies have shown that the general public has exaggerated perceptions of the health risks they face — as well as exaggerated expectations of the benefit of medical care.

Is it because they’re stupid? No. Instead, the problem relates to how various sources of health information — researchers, doctors, reporters, web designers, advertisers, etc. — frequently frame their messages: using relative change.

“Forty percent higher” and “50 percent lower” are statements of relative change. While they are easy to understand, they are also incomplete. Relative change can dramatically exaggerate the underlying effect. It’s a great way to scare people.

For example, research earlier this year found that women with migraines had a 40 percent higher chance of developing multiple sclerosis. That sounds scary.

But the researchers were careful to add some important context: Multiple sclerosis is a rare disease. In fact, for women with migraines, the chance of developing multiple sclerosis over 15 years was considerably less than 1 in 100 — only 0.47 percent. To be sure, that is about 40% higher than the analogous risk for women without migraines — 0.32 percent — but it’s a lot less scary. More importantly, it’s a much more complete piece of information.

What makes it more complete is the context of two additional numbers: the risk of developing multiple sclerosis in women with and without migraines. Epidemiologists call these “absolute risks.” You and I might call them the real numbers.

Relative change also exaggerate effects in the other direction. It’s a great way to make people believe there has been a real medical breakthrough.

A few years ago a study of a cholesterol-lowering statin drug was hailed for big reductions in heart attacks in people with so-called healthy cholesterol levels. The drug led to about a 50 percent reduction in the risk of heart attack. That sounds like a breakthrough.

But the absolute risks — the real numbers — are sure to look a little different. Why? Because in people with healthy cholesterol levels, heart attacks are rare. To get that context, get the two additional numbers: the risk of heart attack in people taking and not taking the drug.

For people taking the drug, the chance of having a heart attack over five years was less than 1 percent. To be sure, that is about 50 percent lower than the analogous risk for those not taking the drug — less than 2 percent — but it sounds a lot less like a breakthrough.

These absolute risks suggest that 100 apparently healthy people have to take the medication for five years for one to avoid a heart attack. And it’s not even clear from the research — or the federal registry of clinical trials — what kind of heart attack: the kind that patients experience (the bad kind) or the kind that is diagnosed by detecting less than a billionth of gram of a protein in the blood (the not-so-important kind). Add in all the hassle factors of being on another drug (filling scripts, blood tests, insurance forms) and the legitimate concerns about side effects, the use of relative change might now strike you as more than a little misleading.

Whatever the finding — harm or benefit — relative change exaggerates it.

Upon learning this, one of my students likened relative change to funhouse mirrors. If you are thin, there is a mirror that can make you look too thin; if you are heavy, there is mirror that can make you look too heavy.

In the case of relative change, it all happens in the same mirror. It provides a potent combo to promote medical care: exaggerated perceptions of risk and exaggerated perceptions of benefit. Can you imagine a more powerful marketing strategy?

Relative change is not the only culprit in misleading health information, but it is an important one. The good news is that more and more researchers, reporters and editors are on to this game. The bad news is that there is an awful lot of information to police and sometimes it can be hard to even find the real numbers.

That’s where a skeptical, numerate public comes in — one that knows to ask for the real numbers. And, if they can’t be found, one that knows to move on.

H. Gilbert Welch is a professor of medicine at the Dartmouth Institute for Health Policy and Clinical Practice. He is the coauthor of Overdiagnosed: Making People Sick in the Pursuit of Health. This post originally appeared on The Huffington Post.

The Testing Glut

May 3, 2012• 9

By H. Gilbert Welch

In case you missed it, a recommendation came out last month that physicians cut back on using 45 common tests and treatments. In addition, patients were advised to question doctors who recommend such things as antibiotics for mild sinusitis, CT scans for an uncomplicated headache or a repeat colonoscopy within 10 years of a normal exam.

The general idea wasn’t all that new — my colleagues and I have been questioning many of the same tests and treatments for years. What was different this time was the source of the recommendations. They came from the heart of the medical profession: the medical specialty boards and societies representing cardiologists, radiologists, gastroenterologists and other doctors. In other words, they came from the very groups that stand to benefit from doing more, not less.

Nine specialty societies contributed five recommendations each to the list (others are expected to contribute in the future). The recommendations each started with the word “don’t” — as in “don’t perform,” “don’t order,” “don’t recommend.”

Could American medicine be changing?

For years, medical organizations have been developing recommendations and guidelines focused on things doctors should do. The specialty societies have been focused on protecting the financial interests of their most profligate members and have been reluctant to acknowledge the problem of overuse. Maybe they are now owning up to the problem.

A Doctor’s Vision for Medicare

Oct 21, 2011• 22

By H. Gilbert Welch, MD

Everybody knows what the federal budget’s long-term problem is. The president knows. The Republicans in Congress know. The Democrats in Congress know. The policy community knows. You know.

It’s Medicare.

I am a physician who has been studying Medicare data throughout my professional life. But now that I’m closing in on becoming a beneficiary, I am thinking more about what I’d like my Medicare program to look like.

My Medicare would be guided by three basic principles:

It should not bankrupt our children. Let’s be clear: Medicare is rightly the central source of concern in the deficit debate. Its expenditures are totally out of control, and represent a huge income transfer to the elderly from their children. It’s a program crying out for a budget.

So let’s pick a number — more specifically, a proportion of total economic output — to cap Medicare. Now the number is 3% to 4% of GDP. We can live with that. Distribute it to geographic regions based simply on how many beneficiaries live there. Expect howls of protest: Urban areas will complain their labor costs are higher; rural areas will complain they cannot achieve the same economies of scale. And everybody will argue that their patients are sicker.

Ignore them all: Make it a block-grant program. Sure, this raises other issues, but you get the principle.

For those who view this as a tea party solution, consider this: I drive a 1999 Volvo and live in Vermont — that should tell you something.

It should not waste money on low-yield medicine. I don’t change my Volvo’s oil every 1,500 miles, even though some mechanics might argue that it would be better for its engine. Nor do I buy new tires every 10,000 miles, even though doing so would arguably make my car safer. But in Medicare (as well as the rest of U.S. medical care) such low-yield interventions are routine.