Skip to content

How Tiny Are Benefits From Many Tests And Pills? Researchers Paint A Picture


Mammograms are said to cut the risk of dying from breast cancer by as much as 20 percent, which sounds like an invincible argument for regular screening.

Two Maryland researchers want people to question that kind of thinking. They want patients to reexamine the usefulness of cancer exams, cholesterol tests, osteoporosis pills, MRI scans and many other routinely prescribed procedures and medicines.

And they want to convince them with statistics — but don’t worry! They promise not to use algebra or spreadsheets. Or even numbers.

Health costs continue to grow much faster than the economy’s ability to pay them. Partly as a result, scrutiny of potentially unneeded and harmful treatment has never been more intense.

Nearly three physicians in four surveyed by the American Board of Internal Medicine said unnecessary tests and procedures are a serious problem. The authoritative National Academy of Medicine estimated that 30 percent of all health spending — $750 billion — is wasted on fraud, administration and needless procedures.

But even doctors often don’t understand the tradeoffs involved in many tests and medicines, says Dr. Andrew Lazris, a Maryland internist. When they do, they have trouble explaining them to patients.

To change that, Lazris and environmental scientist Erik Rifkin are trying to popularize an intuitive, pictorial way of showing just how few people are helped — and how many are even harmed — by many common procedures.

Health is best discussed in the language of risk and probability, but the $70 billion spent on long-shot government lotteries every year suggests that Americans are a bit challenged in that regard. Behavioral psychologists have confirmed what a French writer observed in the 1600s: “Each believes easily what he fears and what he desires.”

Especially about health. Lazris and Rifkin want to give people a more realistic way of evaluating medical hopes and worries.

They ask patients to picture a hall of people getting a test, operation or prescription. Patients might be shocked at how few in the crowded room get any benefit out of the expensive care.

Their “benefit-risk characterization theater” images vividly show the odds, based on solid research. There’s a sold-out house of 1,000 playgoers or concertgoers, all getting a particular kind of exam, screen or pill.

Then the curtain falls. Everybody helped by the procedure or prescription gets up and leaves. Often it’s only a few people. Sometimes very few. Or nobody.

For breast exams, only one woman in the thousand-person theater receiving mammograms over a lifetime is saved from dying by detecting a cancer before it spreads, according to Lazris’ and Rifkin’s summary of the research.

At the same time, hundreds of women in that audience will receive test results suggesting they have cancer when they don’t — “false positives.” Sixty-four get biopsies, which generally involve cells withdrawn through a needle, for nonthreatening lumps.

Ten receive unnecessary treatment including radiation and surgery for lumps that never would have caused a problem.

The theater images show all of that as well, presenting visual demonstrations that the odds of harm, worry or inconvenience caused by the tests are often much higher than the likelihood of benefit.

Drawing conclusions from mammogram studies is contentious. Some reports show greater benefits — as many as five fewer deaths for 1,000 women. For women with a family history of breast cancer, dense breasts and others with higher risks, benefit from screening — perhaps beyond mammograms — is higher than for those with normal risk, researchers say.

But for the average woman the benefit is small by any measure.

Showing all this with theaters “seems like a great idea,” said Dr. Zackary Berger, an assistant professor at the Johns Hopkins School of Medicine who studies patient communication. “It seems pretty intuitive, and that’s the trick. You want to deliver this information in a way that people can really take in.”

Medical decision aids exist online. But doctors may not know about them, Berger said. Even if they do, showing patients requires a computer and a bunch of keystrokes. The theaters are pictures on paper.

What Lazris and Rifkin especially want to combat is the practice of discussing only relative benefits of medical procedures.

Stating that a mammogram lowers mortality risk from breast cancer by 20 percent says nothing about how likely a person is to die of that disease in the first place. Not to mention what the test might cost in pain, harm or hassle.

Cutting risk by 20 percent sounds impressive — until one realizes it might be the difference between five women in 1,000 who don’t get mammograms and die of breast cancer and four women in 1,000 who do get mammograms and die of breast cancer anyway. (Mammograms miss lots of deadly cancers, and some tumors prove fatal even with early detection.)

That’s not much change in absolute risk. The theater images capture that subtlety.

Other procedures and prescriptions show similarly small benefits.

Hip fractures prevented by bone-density pills such as Fosamax, according to Lazris and Rifkin? Roughly five per 1,000 taking the medicine.

Strokes prevented by blood thinner warfarin among patients with atrial fibrillation, a type of irregular heartbeat? Six out of 1,000 — but 12 people out of those 1,000 will suffer major bleeding episodes.

Lives saved per 1,000 men screened for prostate cancer? Zero. Nobody leaves the theater. (It’s unclear whether screening and detection change the course of the disease.) Risks of impotence from surgery or radiation for prostate abnormalities detected by screening that may be nonfatal? Six hundred men get up and walk out.

Lazris and Rifkin say they aren’t pressuring people to avoid tests.

Under the growing philosophy of “shared decision-making,” they just want patients to have a good conversation about what procedures can and can’t do. Then patients decide.

“Whenever I use this, the patients tend to not favor taking the treatment,” Lazris said. However, “there are people who look at the one out of a thousand and say, ‘That looks pretty good. That’s me sitting there. I’ll take it.’”

They published a book in late 2014 on their theater visuals and have tried the images in focus groups. They’re trying to interest insurance companies and health systems.

“When we showed people the theaters and they didn’t have to deal with numbers at all, they all responded positively,” Rifkin said.

What the world needs next is a theater showing how few people win the lottery.

Related Topics

Cost and Quality Public Health