Why medical research keeps changing its mind

Robert H. Shmerling, MD

Senior Faculty Editor, Harvard Health Publishing

Did you ever wonder why medical research seems to flip-flop so often? Eggs used to be terrible for your health; now they’re not so bad. Stomach ulcers were thought to be due to stress and a “type A personality” but that’s been disproven. I was taught that every postmenopausal woman should take hormone replacement therapy to prevent heart disease and bone loss; now it’s considered way too risky. It can make you question every bit of medical news you hear.

But maybe that’s not such a bad thing. Questioning what you read or hear is reasonable. And maybe medical reversals — when new research leads to a complete turnaround regarding a widespread medical practice or treatment — are not as common as they seem. Perhaps they get more attention than they deserve and drown out the consistent and “non-reversed” medical research that’s out there. For example, it seems unlikely that the health benefits of regular exercise, smoking cessation, or maintaining a healthy weight will ever be reversed.

A new study examines medical reversals

A remarkable new study explored the phenomenon of medical reversals to determine how common they are, and to identify what types of conditions were most involved.

Researchers collected more than 3,000 randomized controlled trials; these are considered the most reliable types of research because they randomly assign otherwise similar study subjects to different treatment groups and try to account (control) for factors other than the treatments that might affect the results. For example, a trial comparing two treatments to prevent heart attacks would need to have a similar proportion of people with high cholesterol or high blood pressure, because these can affect the risk of heart attack.

In this new study, the analysis was limited to three of the top medical journals in the world: JAMA (formerly known as the Journal of the American Medical Association), The Lancet, and the New England Journal of Medicine. For each medical reversal identified, the authors searched for later studies refuting the findings and only counted those that had stood the test of time (so far!).

Here’s what they found:

  • Of 3,017 studies analyzed from the last 15 years, 396 came to conclusions that reversed prior treatments or practice recommendations. This represented about 13% of randomized controlled trials appearing in these journals and about 6% of their original research papers.
  • The most common conditions were cardiovascular disease, preventive medicine, and critical care medicine (such as care received in an intensive care unit).
  • Medications, procedures, and vitamins accounted for about two-thirds of the reversals.

Examples of medical reversals

Among the nearly 400 medical treatments or practices that were reversed during the years of this new study, here are some notable examples.

  • Wearable technology for weight loss. When fitness trackers first became widely available, recommendations to use them to help with weight loss were common. But a study in 2016 found that they were no more effective (and perhaps less so) than a standard weight-loss program that did not use an activity tracker.
  • Hormone replacement therapy (HRT). For more than 50 years, HRT was thought to prevent chronic disease, such as cardiovascular disease, in menopausal women. A number of more recent (and more powerful) studies have demonstrated that HRT provides no such benefits, and that some combinations of hormonal therapy may increase the risk of certain cancers, stroke, and blood clots. HRT may still be recommended for women with significant menopausal symptoms, such as hot flashes, but it is no longer prescribed to prevent chronic disease.
  • Surgery for a meniscal (cartilage) tear with osteoarthritis of the knee for adults ages 45 and older. This combination of problems is common among middle-aged and older adults and is frequently detected when MRI scans are performed to evaluate knee pain. While surgery was often recommended and performed to remove or repair the torn meniscus, it was uncertain whether this was truly necessary. A study in 2013 found that initial treatment with physical therapy was just as effective as immediate surgery. Guidelines soon changed to advise nonsurgical treatment as the initial approach for most middle-aged or older patients with meniscal tears and osteoarthritis of the knee.

Medical myth or medical fact?

Myth and misconception are common in matters of health and medical practice. But it’s also true that medical fact is a moving target. Things we accepted as fact years ago sometimes turn out to be wrong, as these medical reversals demonstrate. Meanwhile, certain myths could turn out to have credence if well-designed research concludes as much.

The reason this study about medical reversals is so important is that it points out how vital rigorous research is, not only for new treatments or innovative procedures, but also to evaluate older, well-established ways of doing things.

What’s next?

Hearing medical experts flip-flop on their recommendations or conclusions about medical news seems common — but is 6% of original research or 13% of all randomized controlled trials “too high”? I’d argue that it’s not. In fact, rather than casting doubt on all research, this new study about medical reversals should serve as a measure of reassurance that skepticism is alive and well in the research community, and that “low-value medical practices” will be uncovered if the right research is designed, funded, and implemented.

You can bet that there will be more head-scratching, mind-bending medical reversals in the future. Just keep in mind that most of this is simply a reflection of how researchers are continuing to clarify what works in medicine — and what doesn’t. The best they can do is to keep at it. The best we can do is to consider medical news with a critical eye and to keep an open mind.

Follow me on Twitter @RobShmerling

Commenting has been closed for this post.