Tamiflu, if you get influenza, right? That dog**** was run as a "front line, ought to do it" treatment for the last 20 years.
Uh, you better read this. The trial was halted in critically-ill patients because the treatment arm led to MORE deaths rather than fewer ones. It didn't just "not work", it actually had negative results with the primary negative result being very objective: The person DIED.
That's right -- at best it was ineffective (the CI did cross 1, so it might be, although it just barely touched 1) and the excursion on the other side was extreme.
This was a randomized trial, had a decent sample size and a well-defined and objective (you're dead!) end-point.
It wasn't close either; the odds of the treatment being harmful were close to 98% and the difference in death rates from all causes was roughly 40% higher.
The safety monitoring board responsible for medical ethics stopped the trial when this pattern emerged. Note that this obviously was always the case but gee, wasn't this all the rage when on-patent and thus extremely expensive, yet said to be "front and center" in reducing risk if you got the flu. Now we find out, TWO DECADES LATER, that it kills instead of helps.
