Skewing the Statistics

It’s important when reading statistics that there is a true understanding of what the numbers really mean. Here is an article that effectively demonstrates this. It discusses a recent report in which people use loaded language to make benefits appear much greater or risks much larger than they actually appear.

The researchers accuse drug companies of knowingly using misleading statistics to promote their products because a “big number” can lead to a big headline and lots of sales. And a well-meaning institution might also seek a big number to warn of possible dangerous side effects of a drug, even if that number implies a much greater risk than actually exists…

“Many doctors, patients, journalists, and politicians alike do not understand what health statistics mean,” according to the study, authored by two medical professors at Dartmouth Medical School, Steven Woloshin and Lisa M. Schwartz, and their collaborators, psychologists Gerd Gigerenzer, Wolfgang Gaissmaier and Elke Kurz-Milcke of the Max Planck Institute for Human Development in Berlin.

They go on to talk about one case in which doctors were warned that third-generation contraceptives had double the risk of developing blood clots in the leg. While I agree that women should be told of the risks, and not just assume that any medication is risk-free, the “double the risk” was difficult to understand without knowing the absolute risk. The second-generation Pills had a 1/7000 risk of developing a blood clot, while the third-generation Pills had a 2/7000 or 1/3,500 risk, or still quite small. This was in Great Britain, and led to women going off of their birth control pills, and some blamed 13,000 abortions (hardly risk-free) the following year on the hysteria.

And the report cited studies that showed that many doctors don’t have a strong grasp on these statistics — that when doctors were given tests with questions on risk assessment and statistical statements (such as the above), many answered incorrectly because they misinterpreted the data:

In several studies, physicians were asked if they really understood what the widely known 25 percent risk reduction by mammography screening really means. If 1,000 women are tested, how many fewer will die of breast cancer?

The answers were all over the map, ranging from one out of 1,000 to 750 out of 1,000. Fortunately, most gave the correct answer, the researchers say, which is one. One more survivor will be added to the list, and that’s 25 percent of the total number of survivors. But even after the answer was revealed, one physician insisted it should be 250.

Hearing “double the risk” sounds a lot worse than “one in 3,500 risk”. I wonder how many women would refuse to go on birth control pills if the risks were said in such absolute terms, rather than in relative terms.



One Response

  1. […] That’s an exaggeration which reminds me of the recent post I wrote about Skewing the Statistics. Here are the facts: amniocentesis is recognized as having a 1 in 200-400 risk of miscarriage; and […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: