Monday, October 1, 2007

How to Interpret Medical Studies


We are bombarded with news of medical breakthroughs every day. How can you know what studies are valid and important, and which ones are just fluff? Here are some ways to tell the difference:
  • How many people were in the study? The more the better
  • Who were the subjects, researchers and sponsors? The funding source of the study is important and might change the motives. Do the researchers have credentials? Are the subjects like you? A study of Tibet Nuns might not be as meaningful.
  • What was studied? The best studies look at outcomes...such as rates of heart attack or stroke. Other studies focus on test results. Outcome studies are the hardest to do, but the results are the most meaningful.
  • Meta-analysis: The researchers pool many study results to analyze information from hundreds or thousands of patients.
  • Case controlled study: Compares cases (people with disease) to controls (people without disease) to see why the disease occurs.
  • Single randomized controlled trial: patients are divided into two groups. The experimental group receives a new treatment while the control group gets an inactive treatment. The larger the difference in results, the stronger the evidence.
  • Expert opinion: Only as good as the evidence it is based on. Often it is from BOGSAT - A bunch of guys/gals sitting around talking.
  • Single Cases or Testimonials: "I lost 30 lbs in my sleep". "Loose weight while eating."
Most studies published in peer reviewed Medical journals are well done. But remember that the dramatic results on the front page of the news, may be refuted by another study a few weeks later. It is always wise to wait and see if the researchers work can be duplicated.

8 comments:

Jonathan said...

One of the most informative articles I’ve read recently on the types of medical studies is from the September 16th issue of The New York Times Magazine, “Do We Really Know What Makes Us Healthy”. The article clearly illuminates the difference between observational studies and randomized-controlled trials, and explains well why we often see conflicting medical advice over time.

The article can be read here:
http://www.nytimes.com/2007/09/16/magazine/16epidemiology-t.html

Toni Brayer MD said...

Thanks for the info and link, J.

Rey said...

Some newspaper cartoonist once did a drawing parodying medical studies. It showed three large, vertical wheels of chance. (This is a rough remembrance.) The first was to choose a noun, like nicotine, the second was to choose an action, like prevents, the third was to choose an illness, like cancer. Spin the wheels and there are hundreds of possibilities. Dr. Elizabeth Barrett-Connor uses the cartoon in some of her talks, and thinks it came from the textbook of Epidemiology by Leon Gordis. Has anyone ever found a copy of this cartoon online?

Anonymous said...

Doc,

Hate to be the cynic, but if a clinical trial you are interpreting contains a branded medication, the results lack authenticity and are embellished, so it may not be wise to rely on conclusions from such a trial.

Sunrise Global Solutions said...

http://sangambayard-c-m.com

viagra online said...

Interesting I never though that it was like that, in anycase if they are gonna made some kind of test how they should be do it?
Thanks

Anonymous said...

Hi Rey, I found the cartoon you're referring to. So true.
http://hannahhouck.blogspot.com/2007/05/trouble-with-news-and-medicine.html

sex shop market said...

It cannot have effect in reality, that is what I consider.