Why medical research sucks - part 3

Please remember this is a multi part blog. It will make much more sense if you have read the past blogs related to the subject, which are below. You may have to scroll down just a bit to get to them but it will be worth it if you are just joining me.

Yesterday, among other things I pointed out that medical research rarely proves anything. The phrase "More research is needed" in one form or another seems to be there almost always.

In the most recent big medical journal I read, it was at the end of every single article except the editorials.

And even one of the editorials ended with that phrase.

Those guys are sorely in need of some Instant Einstein and Instant Endurance.

It reminded me of Seinfeld, the show about nothing. Nothing gets done that doesn't further the interests of the big drug companies.

So here are a few more reasons why medical research often sucks.

Why medical research sucks reason 3:

The wrong patients are in the study.

When a drug company does a study, it does not start out to find anything but what it wants to find.

Thus there are often selection criteria for the people they want in the study. Most commonly left out of studies are women, blacks, Hispanics, people with multiple illnesses and people over 65.

You see, the goal of the study is not to provide the doctor or the patient with real world information. It is to provide good-looking data that makes it easy to go in and sell the drug to the doc and the patients.

So for instance, if you are looking for good results with your cholesterol lowering statin drug you will want to leave out women because they will not respond well.

If you want something that prevents death from heart disease, the data is very strong in favor of fish oil the only supplement that the American Heart Association recommends in any way.

For blood pressure drugs you might want to leave out blacks and Hispanics for the same reason.

And what about people with multiple medical illnesses?

Geez, how messy is that? I mean no one out there in the real world really takes more than one pill do they?

No one out there in the real world actually has what we in medicine call "co morbidities" (multiple problems).

The fact is the large majority of people over 50 are on more than one drug. But that might mess up the pretty study so let's just not deal with reality.

Finally, we have the fastest growing group of our population, the over 65 year olds. They are left out of a lot of studies because they have a nasty habit of dying.

That looks very bad in your pretty little study but it is a fact of life, or in this case, death. If you don't study the effects of a drug in this rapidly growing population, what real world information are you providing to the doctors and patients that use this drug, many of who are bound to be older?

If you are 65 or older my research suggests you really, really need to be on fish oil!

Why medical research sucks reason number 4:

Statistical game playing. I am no statistician. As a matter of fact, when I took my managerial equivalent of the Myers Briggs personality test, I was called a forest person not a tree person.

In other words details bore me.

Thanks to Instant Einstein and fish oil I can overcome my mental handicaps.

So I do know enough about how data can be manipulated to provide results.

First off, a company may not report the negative findings. An example of this might be the Vioxx debacle. The study they used to prove their drug was good asked only the question does it work for ulcers. Answer: Yes, but it increases heart attacks.

That data was there all along and the guy heading up the study was a heart doctor!

I guess they weren't supposed to look at that part of it.

Heaven knows what lines the shredders of major drug company's offices.

One thing you are sure to find, there is all the data on how good fish oil is for so many different things.

Another favorite tool is sub group analysis.

This is exactly the opposite of the "Vioxx problem". Here statisticians don't get the results they want so they go back in and look for secondary findings.

"Ok our drugs sucked at preventing strokes but maybe it works for diabetic leg ulcers. Or maybe it sucks for depression but it works for appetite suppression."

Just invent category for the drug if it doesn't fit the one you originally intended.

I suspect Viagra was one of these and also the popular weight loss drug, Meridia, which was originally designed to be another depression drug.

There are many more complex ways that people can manipulate numbers. The final one I want to talk about today is the most dangerous because it combines so many of the previous flaws.

That is called Meta analysis. Here data is fed into a computer and analyzed to "look at everything available".

Classic example of this was the notorious "Fish Oil no good for cancer" article published by JAMA.

They took studies as far back as 1966 which had to have design flaws.

They stopped in the late 90's when Pharmaceutical Grade Fish oil was just gaining momentum. And used studies that didn't even use fish oil at all.

Conclusion: garbage in garbage out.

They don't want anyone to know this, I am sure.

They also don't want you to know that you might take the reins of your own health care and steal the money right out of the hands that feeds them: Big Pharmacy.

Take your fish oil and Monster Multi and Instant Einstein every day and help me take back the power of the American consumer.

Don't forget to join me tomorrow because I am still slapping the "health robbers" around and while you're at it, pass the blog along to your friends by clicking the envelope icon below.

All the best,

Doc