Persuasive health information could be misleading

NEW YORK (Reuters Health) - The most persuasive health information won't always serve your best interests, U.S. doctors said Wednesday.

In a new report, they describe how both patients and physicians making hypothetical treatment decisions are more easily swayed by impressive-sounding numbers than useful ones.

In principle, that could lead to unnecessary healthcare spending and overtreatment of patients, according to one of the report's authors, Dr. Elie A. Akl of the University at Buffalo in New York.

For example, he said, a study from November showed smokers screened for lung cancer with a CT scan saw 20 percent fewer deaths over the study than those not screened.

The finding got a lot of people cheering, but the excitement fizzled out when it became clear that the absolute risk only dropped by 0.33 percent -- from 1.65 percent to 1.32 percent.

In other words, about 300 people would need to be scanned to stave off one death. Add to that the risk of false positives and the anxiety of going through the procedure, and the equation ends up looking less appealing.

"You can understand why people make different decisions based on these numbers, but they are actually all derived from the same statistics," said Akl, whose findings are published by the Cochrane Collaboration, an international organization that evaluates medical research.

In Akl's example, the 20 percent figure is called a "relative risk reduction." It doesn't say anything about the actual, or "absolute," risk of dying during the study, however. That would be the two smaller numbers just over one percent.

The 300 figure - derived from the absolute risk reduction -- is called the "number need to treat."

Compared to relative risk, "the absolute risk reduction is probably more informative," Akl told Reuters Health. "It gives the whole picture."

The new report summarizes 35 earlier studies that tested how health professionals and consumers respond to different ways to present risks -- how well they understand them, how big they perceive them to be and how persuasive they think they are.

"The main result is that the effect of an intervention is falsely perceived to be larger when it is presented as a relative risk reduction," Akl said.

"Also, people who receive information in the format of relative risks are more willing to start taking a medication or go through a procedure compared to when they get the absolute risks."

Both consumers and health professionals tended to be more impressed with the relative numbers. However, the studies only looked at hypothetical health decisions, not real-life choices.

Akl added that drug companies often use relative risks to promote their products.

"They know it is more persuasive," he said.

SOURCE: bit.ly/heNAUM The Cochrane Library, online March 16, 2011.