#798 X-Results

Recently there have been a lot of scandals about Big Pharma faking results. Drugs killing people and stuff, tests done by the companies suppressing negative information, you know, big business gone bad.
Some of us fear big government, some of us fear big business, I’m just afraid of big. When people, customers, and taxpayers get reduced to statistical blips, bad things happen.
In any event, there was this story about a drug that was supposed to be given to people with strokes. The idea was it would bust up potential clots and cause less disability and death. The drug was tested and found to significantly reduce disability and death in post stroke patients.
Doctors started using it. People started dying.
Guess what? The drug was, by design, never tested on severe stroke victims. So the study results naturally reflected better outcomes since it studied better beginnings. A Band-Aid is going to work a lot better on a scratch than a six-inch gaping gash.
The other thing the study did was show results using an accepted technique known as Combination End Point reporting. And interestingly, this method depends on semantics and not science.
Specifically, the tortuous reasoning one can employ with the word “and.”
Remember when I said earlier that this drug significantly reduced disability and death. Well actually it reduced disability in a lot of mild stroke patients. It only killed a couple of them. But it reduced disability in far more of them than it killed.
So the drug company could truthfully say that the combined results of the two end points—death and disability—was that when both numbers were taken together, the drug reduced disability and death.
That’s like saying that statistically, the headlights on your car reduce darkness and your battery going dead.
Combined end points leave lots of drugs on the market. Like certain cholesterol drugs that reduced cholesterol but had no effect on coronary arterial plaque build-up.
The drug company could say the drug lowered cholesterol and plaque together.
One doctor put it best after a number of people died:
“Dying with corrected cholesterol is not a successful outcome.”
America, ya gotta love it.