X hits on this document

506 views

0 shares

0 downloads

0 comments

143 / 170

12

look at this, that, you know, you have a point estimate that happens to look really good in

13

one direction, you have certain confidence around that. You really need to widen what

14

you think the real confidence intervals are around that because you haven’t really done it

15

the way you like to do it. You don’t have adjudication, you don’t have pre-specification.

16

Now that doesn’t make it no data; these are real events that were really reported, that

17

were of concern to the investigators, that were adjudicated as best we could post-hoc. It

18

just makes the estimate of it much more difficult.

2

that the absence of adjudication and pre-specification are substantial limitations. There is

3

no doubt about it, which is why I think the guidance document was written as it was. I

4

guess the way I look at this is as a clinician that sometimes practices statistics at risk.

5

You know, I guess the issue really is for us, what is the true confidence around the

6

estimate? I guess in some way that we have no way to quantify. You know, when you

7

don’t have adjudication and you don’t have pre-specification, these are things that make

8

the confidence around a statistical test questionable, and make you need to widen the

9

confidence around whatever, something that I’ll call ‘the true estimate’ in some

10

immeasurable way is.

20

like to re-emphasize the last part of this Question 2, which is, this isn’t, I suspect, the

21

only time we will be discussing this, or the FDA will be discussing this. What other

22

suggestions do we have for improvements to present and future endpoints and analysis

23

that may be applied to other diabetic programs that have already been completed or have

24

ongoing phase 3 programs at the time the final guidance was issued. A difficult question,

11

What is reality? I guess sort of that, for what it’s worth, that’s the way I

19

DR. BURMAN: Other comments on those issues? Because I would also

Scribes, LLC Toll Free 1-800-675-8846 www.scribesllc.com

1

DR. KONSTAM: Yeah, you know, I think you’re very right. I think

142

Document info
Document views506
Page views506
Page last viewedFri Dec 09 03:33:26 UTC 2016
Pages170
Paragraphs7975
Words54271

Comments