Healthy Skepticism Library item: 10451

Warning: This library includes all items relevant to health product marketing that we are aware of regardless of quality. Often we do not agree with all or part of the contents.

Publication type: news

Evans M.
Why you shouldn't let the headlines scare you
The Globe and Mail (Toronto) 2007 Jun 12
http://www.theglobeandmail.com/servlet/story/RTGAM.20070612.wlevans0612/BNStory/specialScienceandHealth/home

Full text:

They say the truth is relative. In drug studies, that may not always be the case.

Every week, a newly released trial asserts a 30-per-cent reduction in this or a 40-per-cent increase in that. In recent weeks, the buzz in the medical world is about rosiglitazone, or Avandia, a fairly new and widely prescribed drug for Type 2 diabetes.

A new analysis of recent studies points to the possibility that the drug actually increases the risk of heart disease, which already claims the lives of 60 per cent of diabetics.

After the news broke, our clinic was inundated with calls from patients worried they’d have a heart attack if they didn’t come off Avandia right away. Their concern is understandable, but they may have been well served by understanding relative versus absolute risk.

Relative risk can be misleading because it measures the percentage of how much something changed, as opposed to actual numbers.

Previous research on Avandia had tipped off the medical community that there may be an increased risk of heart problems among those taking this medication. But there were no definitive answers.

When researchers can’t quite figure out what to do based on individual trials, they often conduct a “meta-analysis” combining similar studies to see if the increased numbers give more precise answers. Dr. Steven Nissen and his colleague at the Cleveland Clinic, Kathy Wolski, took on this challenge with Avandia.

A headline from the meta-analysis of similar rosiglitazone trials, published last month in the New England Journal of Medicine, said people who take the drug had a 43-per-cent increase in risk of heart problems. I’d be calling my doctor, too. But before Avandia users make out their wills, let’s think about relative risk.

Say I told you that you were at increased risk for heart disease, and you had the choice of three medications. Based on the following study results, which drug would you choose?

Of the patients taking drug A, 3 per cent had heart attacks, compared to 5 per cent among those who weren’t taking the drug. Patients treated with drug B had 40 per cent fewer heart attacks than people not on the drug. For every 50 people who took drug C, one heart attack will be prevented.

All three of these drugs are equally effective.

Work with me here. The study results for drug A reported the absolute risk reduction – 2 per cent among people who took the drug. The trial for drug B showed 40 per cent fewer heart attacks – the same relative risk reduction as drug A, which also reduced risk by 40 per cent: from five heart attacks without the drug to three with it.

And if 2 per cent benefited in the trial, this means two patients in 100, or one in 50, would benefit from the drug – the results shown for drug C.

If you are a researcher, a journalist, a drug company or a public relations agent, relative risk may be your preferred communication – it makes you or your drug sound pretty darn effective.

You aren’t the only one misled by relative risk reduction. Dr. Tom Fahey, a primary care researcher in Bristol, England, surveyed key health policy decision makers in Great Britain, posing a mammography example similar to mine. Only three of the 182 surveyed got it right.

Of course, these scenarios present only one side of the story. They don’t capture your values, other options, cost, ease, side effects or the medication’s potential for harm.

The Avandia paper was not looking at effectiveness but at harm. The relative risk for heart attack increased by 43 per cent. However, only one patient in 185 taking Avandia, typically for years, actually had a heart attack.

Is misleading people about the effect or harm of medical interventions common? Ray Moynihan, a reporter who worked with a team of researchers at the Department of Ambulatory Care and Prevention at Harvard Medical School, tried to answer this question by surveying how the media reports health stories.

The researchers analyzed common health media stories in 2000 and reported their results in the New England Journal of Medicine. Of the 207 stories found, 124 reported benefits quantitatively: 83 per cent reported relative benefits only, 2 per cent reported absolute benefits only, and 15 per cent included both.

DR. MICHAEL EVANS

Do yourself a favour. Next time you start to panic about a health headline, ask yourself what the absolute risk was, not the relative risk.

Michael Evans is an associate professor at the University of Toronto, director of the Health Media Lab at the Li Ka Shing Knowledge Institute at St. Michael’s Hospital, and staff physician at Toronto Western Hospital.