Medical Science Then And Now

Via Reason, The Washington Post has this story on the 50th anniversary of the first clinical trial in the United States:

Fifty years ago this summer, the preliminary results of that trial were published. They changed medical history. Almost overnight, physicians stopped automatically giving supplemental oxygen to preemies, ending the epidemic of retrolental fibroplasia (RLF), as the disease was called then. (It is now known as retinopathy of prematurity.)

But the study's results did something else equally important and historic. They convinced many American physicians of the usefulness of randomized controlled trials, which had been "invented" less than 10 years earlier in Britain. Not least, the study taught doctors they couldn't assume that what seemed like a good idea -- extra oxygen -- would necessarily lead to a good outcome.

Read the whole thing - especially how further trials showed the definite harms of oxygen therapy to premature infants, and still more trials that elucidated its true benefits. Now, just as then, we realize medical science ain't an exact science. As an illustration, Balko points to a recent study that estimates the mortality of obesity(per year in the US) is 25,000 (previous studies put the figure at 350,000-400,000):

Being overweight is nowhere near as big a killer as the government thought, ranking No. 7 instead of No. 2 among the nation's leading preventable causes of death, according to a startling new calculation from the CDC.

The Centers for Disease Control and Prevention estimated Tuesday that packing on too many pounds accounts for 25,814 deaths a year in the United States. As recently as January, the CDC came up with an estimate 14 times higher: 365,000 deaths.

The new analysis found that obesity — being extremely overweight — is indisputably lethal. But like several recent smaller studies, it found that people who are modestly overweight actually have a lower risk of death than those of normal weight.

Share this

Yeah, the CDC, like any

Yeah, the CDC, like any govt. agency, wants to have a reason to exist. So it has a long history of extending the definition of "disease" and exaggerating the pervasiveness and danger of "diseases", so that it has a reason to have more funding. Same w/ the WHO. That's why we get these reports of "killer epidemics in the making" (like SARS) that end up killing 20...100..1000

If the study you're

If the study you're referring to is the same one I read about yesterday, I think it was in the newspaper, then you should know that several peer scientists who reviewed that paper/study said the statistics and overall methodology presented there were both deeply flawed.

Right. And several

Right. And several reputable peer scientists said it was a very good study. And several reputable peer scientists said the first study was good. And several reputable peer scientists said the fist study was flawed. What's the point?

My simple point was that medical science is not yet precise.

I guess I missed your point

I guess I missed your point then. But now that I think I get it, it's worth noting that the improper use of statistics in quantitative analysis is by no means unique to biomedical researchers. *If* the methodology used in this study was flawed, the conclusions reached could be outright wrong, not merely "imprecise."

Or, the methodology used

Or, the methodology used could be flawed and the conclusions still could be dead-on right. Or totally off, or anywhere in between.

Note that both studies involve the CDC. When two studies purported to seek the same number find a disparity by a factor of ten, and without a gold standard for comparison, imprecise is probably the best description.