« Extremophiles thriving in toxic waste, subject of cancer research | Main | For babies with diarrhea and pro athletes »

September 06, 2007

What makes a really bad medical story headline?

These are the attention grabbers for our jaded Web eyes and ears, meant to target our emotions, mostly fear and anxiety. But don't read the whole story, because then you're in for a big letdown. 

"Cell Phones in Hospitals: Bad Rx"
(From Time online) & "Turn that cell phone off!" (This is a video report on CNN Health, but you have to watch a commercial for Momma Mia! before you get to the good stuff.)

Turn a minor study into a reason to panic!

This study examined the possible electromagnetic interference (EMI) of cell phones with the medical monitoring equipment or mechanical ventilators used for the very ill patients in hospitals. You'd think that by answering your cell phone in a typical patient room you would wreak havoc on the ward. But most of us know that this doesn't happen.

If you read the whole Time report on the study, you'll find out the interference mostly occurred when the phone was held some 3 cm (< 1.5 inches) from the medical equipment which is mostly used in the CCU of the hospital where access is restricted anyway. Plus, the phones they were using were operating at higher power levels (at least 3x according to an electrical engineer doing similar research) than cell phones in the US.

Common sense comes into play here. Hospitals have set up guidelines were they stipulate that you keep 3 feet away from this equipment, just to be on the safe side. Besides, if you're getting that close to someone that sick, you should be more mindful of infection control, and wear a gown, mask and gloves without pulling out your grimy cell phone.

The following are headlines I made up, to serve as examples, and to make a point. And, sometimes the copy doesn't fully correct what the headline is claiming.

"New hope for diabetics"

If it were that simple to give people hope from a single scientific study. I wouldn't be too cynical in saying that sometimes a financial interest is at stake. The next group that tries to reproduce the results might find the opposite effect of the treatment or drug being studied, if these results are published at all.

Even if the therapy is further into clinical trials or approved by the FDA, this doesn't even take into account what adverse effects begin to crop up after it's tried with a very large number of patients followed over an extended period of time.

Also, calling people "diabetics" plays into the shame of being labeled as one of *those* people.

"Smokers who eat chocolate found to have fewer heart attacks."

Provide an easy solution, that just so happens to taste good too, but ignore the bigger picture. It's like when you walk into a liquor store and you see this newspaper article taped to the front door at about eye level that says, "Red wine is good for your heart!" 

But there's something else wrong here. This is the old confusion of causality versus association. Very few studies are rigorous enough to prove A causes B.

Most probably what was done was a retrospective case-control study where a group of similar people, i.e. smokers, where put into 2 different groups depending on how much chocolate they said they ate in the last 20 years, and then check their cardiac history. The study might reveal an association, correlation is a better word, but they can't say that it was the chocolate that protected the heart, nor could they suggest that if you're a smoker you'll get a benefit by making a determined effort to eat more chocolate.

The studies needed to prove causality are very difficult, if not impossible to do. You have to do a prospective, randomly-controlled, double-blinded study. Which means you'd have to select a group of smokers who are very similar to each other in lifestyle and behavior (forget about trying to control for genetics), and then randomly assign each to either a chocolate-eating group, or a control group  which eats something that looks, smells and tastes like chocolate but isn't the real deal. Then you have to follow them long enough until a significant number have reached your end point, a cardiac event.

You're better off just trying to convince your readers to stop smoking.


TrackBack URL for this entry:

Listed below are links to weblogs that reference What makes a really bad medical story headline?:


blog comments powered by Disqus