Biotech Values

[iwfal]

When Mice Mislead - an interesting article in Science that is pretty damning of the rigor of mouse experiments:

1) not only do many mouse experiments not follow ITT, they don't follow PP. Animals excluded - no explanation required.

2) no randomization - just whatever animals climb into your hand. (and no blinding either)

3) interesting stats on the apparent efficacy as seen in non-randomized non-blinded animal trials vs those randomized. (guess which did a lot better)

4) generally company animal trials are more rigorous - presumably because they want to minimize the potential wasted $ in human trials. (I'd hazard the guess that small biotech is better than academic, but not as rigorous as BP)

5) as usual in such cases one of the big defenses trotted out is that 'outsiders will think our data was worthless'. Well, yeah. Not a particularly good defense.

Finally it should be noted that some of the big publishers (e.g. Nature) are just recently trying to enforce standards. But even cynical I was surprised at the size of the problem.