Avoiding Rubber Numbers

Of all the challenges we face when sifting through the contradictory messages on nutrition and health, one of the most menacing is the influence of industry on academia and scientific research. What are we supposed to think when the National Dairy Council proclaims that milk aids in weight loss? Most people don’t know how to evaluate whether or not a research study was well-conducted and trustworthy. How are we supposed to react when our doctors tell us a certain pharmaceutical drug offers a big risk reduction for tumor growth? Most people aren’t familiar with the distinction between absolute and relative risk reduction—but these can be very different numbers, and understanding both of them is a critical part of being an informed patient.

If you are considering a treatment, always ask your doctor what the absolute risk reduction is, meaning how much better do people do on this treatment as opposed to people who do nothing at all? Often, results from studies on the efficacy of different drugs or procedures are presented in terms of relative risk reduction—which is just a comparison between treatments. The number may stick in people’s minds, but the number by itself can be misleading, because relative risk reduction only tells part of the story.

Relative risk compares one treatment to another. Here’s a scary example: the results from a study looking at the effects of chemotherapy in breast cancer mortality for women under 50 can be described in two ways. You can either say that over 15 years, the treatment (in this case, polychemotherapy) offers a 25% relative risk reduction for mortality, or you can say it offers a 10% absolute risk reduction. And a 10% absolute risk reduction over 15 years translates to 0.66% per year, or, in real life terms: during each of the 15 years, mostly likely 1 out of 150 people fewer would die from breast cancer, even though without taking the treatment, 149 out of 150 people weren’t going to die anyway.[1]

It’s easy to see how many people might hear “25% reduction” and, without understanding the context for what that number means, leap in their minds to imagining that 1 out of 4 people are benefited by the treatment. They could even leap to thinking 1 out of 4 people fewer will die from breast cancer, when actually, that “25%” is coming from comparing this treatment to another one. There is no standard for comparison. It could mean that when taking this treatment that 3 out of 10 people are at risk for death instead of 4 out of 10. It could mean that 3 out of 100 people are at risk instead of 4 out of 100. It could mean that 3 out of 1000 people are at risk instead of 4 out of 1000. The “25%” comes from comparing the result of the two treatments—the 3 down from 4—not from comparing this treatment to no treatment, and different numbers are produced depending on what the treatment in question is compared to. If this treatment prevented 1/10 people from dying—so now 3/10 instead of 4/10 were likely to die—that might be good news! But it could just as easily mean 3 out of 1000 instead of 4 out of 1000, and this would hardly be noticeable in real life. What would be noticeable would be all the common side effects from strong drugs and treatments, like nausea, fatigue, loss of appetite, loss of sex drive, loss of quality of life…

For many people, this very low probability of seeing some benefit from a treatment may not justify all the negative side effects that could accompany it. Unfortunately most people only hear about relative risk reduction, which usually makes these treatments sound a lot more appealing—great news for big pharmaceutical companies.

The way big business influences the information that is produced and the way that the media picks up this information and spreads it far and wide without placing it in context can create a situation where the public hears information two or three degrees removed from the original source. It is up to you, as a patient and a member of the public, to demand all the information available when making decisions about your health.

The research process is designed in such a way that results can easily make treatments seem more promising than they really are. It’s standard practice in pharmaceutical drug tests to stop the study and put everyone on the same program as soon as either the control group or treatment group seems to be doing just a little bit better. In the name of not denying effective care, harsh drugs with unpleasant and often dangerous side effects may be labeled as “effective” because they show some effect in the short run. In the long run, this doesn’t necessarily translate to a positive effect on health at all, or a reduction in mortality. And big business is not keen to promote the idea that eating fresh fruits and vegetables and giving up animal foods and processed foods may have a much bigger effect on health than any drug or treatment available.

Remember to keep asking questions, and don’t let yourself be pressured into decisions without getting all the information. And remember to eat plants!

[1] Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lancet 2005; 365:1687.

Posted on Kris Carr's Crazy Sexy Wellness Blog

Of all the challenges we face when sifting through the contradictory messages on nutrition and health, one of the most menacing is the influence of industry on academia and scientific research. What are we supposed to think when the National Dairy Council proclaims that milk aids in weight loss? Most people don’t know how to evaluate whether or not a research study was well-conducted and trustworthy. How are we supposed to react when our doctors tell us a certain pharmaceutical drug offers a big risk reduction for tumor growth? Most people aren’t familiar with the distinction between absolute and relative risk reduction—but these can be very different numbers, and understanding both of them is a critical part of being an informed patient...

Go to top