DR FLYER'S MEDICAL BULLETIN
Issue No. 2April 2012
In This Issue
Health Care is a Mystery
Quick Links
  
 
 Greetings! 

For this April newsletter, I wanted to share with you my article on understanding medical news.

Sincerely,
 

Brian Flyer, MD 
 

 

MEDICAL STUDIES AND THE MEDIA

 

 

Mainstream media figured out years ago that reporting on medical studies sells. Indeed, we are all sponges for that kind of stuff. Does eating red meat shorten year life? Is coffee bad for you? Is alcohol good for you?

 

But, the information generated is often confusing, especially given the fact that studies on the same topic over time seem to lead to different conclusions. When do you believe what you hear and what can you really make of it?

 

For example, it was reported that counties in the United States with the lowest rates of kidney cancer occurred in small, Midwestern, Republican dominated areas. Most people, upon hearing this would discount politics, but attribute those results to less pollution and perhaps a healthier diet with unprocessed foods.

 

Yet, the same study also reported that counties with the highest rates of kidney cancer occurred in small, Midwestern, Republican dominated areas. Upon hearing that, most people again would discount politics, but conclude that those results might reflect a high fat diet, more smoking and alcohol, or perhaps pesticide exposure in those areas. Yet, the key word was 'small' that invalidated that study. Indeed, comparing small counties to large counties leads to sampling error. For example, in flipping coins, you are more likely to flip heads 5 times in 5 tries than 100 times in 100 tries. Sample size matters, and that study's results were meaningless.

 

Indeed, scientists design studies in an effort to reduce certain biases, but sometimes those biases still exist. There is 'selection bias', where subjects in a study don't represent the true population of interest. People who volunteer to participate in a study are often different from non-volunteers. Likewise, those who do not respond to a survey differ in important ways from those who do.

 

Furthermore, there are measurement biases. Observers in a study may expect a certain result and may err in measuring data toward that expected result. Another example is recall or memory bias. For instance, a woman giving birth to a baby with medical defects is more likely to report gestational exposures or a history of STDs than a woman giving birth to a healthy baby. Likewise, a cancer patient is more likely to recall a diet rich in red meat or radiation exposure. People recall positive events more that negative ones. In addition, certain subjects may be questioned by researchers more vigorously than others.

 

Additionally, there are intervention biases. Some 'control' group members who by design were expected to receive placebo, might inadvertently receive treatment or exposure. Also, in a study requiring many years, there are issues like maturation alone that could result in measured improvements. And certainly, there are factors related to strict adherence by study participants or withdrawal bias among drop-outs from the study.

Unfortunately, headline news stories give little details about the studies themselves, instead, highlighting only the conclusions. And because of the way our minds work, we often construct instantaneous reasons for those conclusions without ever questioning the accuracy of those studies.

 

If that's not confusing enough, there are often pre-packaged medical conclusions that get aired intentionally to promote a message; a kind of medical propaganda. I've noticed the oft stated myth that rising health care costs follow from emergency room visits by the uninsured, which indirectly raises fees for those with insurance. In reality, the uninsured, particularly at non-county hospitals, make up an insignificant percentage of emergency room visits. With the ongoing debate and pending Supreme Court decision about the role of government in health care, it is certain that we will see very frequent messages of this kind to foster one opinion or another.

 

The point is, don't accept medical headline news as fact. Consider how the study was designed. Was it a survey or observational study or was it the gold standard of scientific research, 'the double blind, randomized, prospective study'? And how big was the sample size and does it consist of people like you? Finally, there is no substitute for obtaining the advice of your physician as opposed to relying on the scare tactics or misleading studies reported by the media whose job it is to sell papers or air time, not to look after your good health.

 Brian Flyer, MD

1125 S. Beverly Dr. #700

Los Angeles, CA 90035

310-300-1122

[email protected] 

www.flyermd.com