Member Info for jackdaw99

Jimzi,

I'm not sure this is about "finding" someone affected by expanding sample size but, instead, about using a sample size big enough to demonstrate the rarity of the said effect, perhaps to demonstrate that it is lower than 1 in 10,000. Since the rare effect will have a statistical distribution (Poisson probably) there are (rather arcane) statistical methods of calculating how large your sample should be before you can be, say, 99% confident that the illness is in fact rare. You are right in that, as the illness is expected to be rarer, the necessary sample size would increase, but not in the way you suggest.

They are giving analysts lots of notice. Be there or be square (=poor).

Didn't notice that one. I thought you were claiming they were changing the DNA code for "rod dystrophy", which would seem entirely sensible .... ;-)

I suspect you mean "ie" not "eg" in that second line?

And, no, this therapy is aimed at problems with the photoreceptors, not the optical system, so those of us wanting to throw away our spectacles will have to go elsewhere.

Well, it would be certainly be interesting to see the sub-group analysis for the excluded sub-group especially because I would expect to see strong interaction terms in the overall ANOVA. I don't agree the sub-group analysis is unimportant if that sub-group becomes a target patient group in the labelling. However, also interesting is the fact that, were the IDVU sub-group to have been the subject of an a-priori experimental design (ie a proper randomised, blind design,) then the results would almost certainly have been even stronger (confidence limits reduced) - for the reason that statistical significances are "harder" to achieve in post-hoc analyses compared to a-priori analyses.

Patience needed but that chart is looking interesting. When TMT and its derivatives move from the lab into the clinic and then the GP surgery, this could be priced in pounds not pennies IMO.

Well, on the face of it, those results look very good. The gloomy could focus on the presence of adverse effects but they are almost always present in a population and are most likely genotype-related. For example, I cannot use penicillin because of adverse effects but, thankfully, there are others available!

He bought some!

You might like to see it but they would be mad to do it right now....

Very interesting question. And, if the latter, what size envelope?

" several hundred thousand pages?" I find that hard to believe (words, perhaps, not pages?) but, if true, I am gobsmacked.

Exactly. Most folks don't really understand the level of statistical sophistication necessary to analyse modern statistical trials (clue: if you've never heard of RA Fisher, you haven't even begun...). Consequently there are many different, equally valid ways to analyse trial data, and the method you choose depends on what you are looking for and the relative importance of all the various results that might emerge from the trial. Therefore it is not inconceivable that what the FDA are asking for is a re-analysis with a different slant: you might say, they could do this themselves but that's not their job so they are entirely correct in throwing it back to the CRO who will, of course, have all the raw data plus anything they may have collected but not included in submission. I'm not saying this is the case, or even that it is the most likely, but what the CRL asked for could well be something like this.

Good post.

It's your imagination. See a doctor as soon as possible.

The Co never said, definitively, that no new data would be needed. Some on this board did, on their own initiative. And we still don't know whether any of this means new trials rather than a different analysis.

Of course he doesn't. He also makes exactly the same spelling mistakes as at least one other poster - or perhaps they simply skipped the same classes?

Dear Scott, No-one's listening. Go for a nice, long walk.

Not long now...

Can't agree with these conspiracy theories. A larger AB manufacturer trying to influence the FDA would be taking an enormous risk of anti-trust fines in the US, Europe, and rest of world. Just not worth it, even IF the integrity of the whole FDA mechanism could be compromised, which IMO is about as likely as those suggestions that NASA faked the moon landings. If another Co was afraid of the MTFB threat it would be much "cheaper", weighting costs with risks, to simply buy them, especially at these valuations. I can't see anyone refusing, say, 80 pence right now, and that still wouldn't be expensive for a major player. So no, for me the most likely scenario (and we will probably never know ...) is the administrators in the FDA finding every excuse to slow things down while their wages weren't being paid - we even have some circumstantial evidence for that (lovely charts BTW) posted on this board.

They don't need FDA approval before they sign agreements, which could come at any time, although they might be NDAs.