Lol maybe I am a Shakespearean fool – a clever peasant trying to use my wits to outdo someone of higher social standing.
In terms of why clinical trial results don’t exist until they are legitimised?
Take BITs statistical test on that third time point. The statistical test produced a t value with a p. That’s the result simple? No.
BIT submits the results to a journal and it goes out to statistical review. The reviewer says don’t use a Welsh test use the Students t test.
Now if you change the test the result changes from what you have reported to disallowed. If you don’t make the change the reviewer is going to become suspicious about why you are being difficult.
So invariably you will make the change to the statistical test, the result changes and now you have two versions of the clinical trial results. Which version is the true one?
The answer is the version that has been accepted by the scientific community through peer review. Your initial results to disallowed simply reflected your opinion about which test to use. But the gate keeper to scientific legitimacy has a different opinion and theirs (unless you can successfully argue the case) carries more weight.
Often how companies get around this problem is either to not report actual data values for findings to stock exchange (remember how you spotted BIT didn’t report the actual p value for that test but didn’t know why they hadn’t) or label the title of announcement “Preliminary results blah blah” because they know that they are likely dealing to be dealing with a moving feast.
If you dig around deep enough and find the US QRX class action the issue of when data from QRX ‘s trials actually became “the clinical trial results” (eg when QRX analysed the data, when the FDA conducted their analyses or when the Ad Com members expressed their views about how things should be done) became quite important. SPL is another recent example where under a SPA SPL claimed successful trial results; the FDA issued a CRL and no one has the foggiest clue now what those actual trial results are.
In terms of peer review? Yes for a lot of research in run of the mill journals it is not magical. But for top tier medical journals where major clinical trials are published (and what the big court cases end up about) that peer review process is not how you think it is. It is actually mainly about the in-house speciality content sub-editors and the in-house statistical reviewers.
Changing the title of this thread? It is inflammatory to BIT and I probably wouldn’t have agreed to the couple of posts with you if I had known I was going to be adding visibility to it. Here presents a little opportunity for you - in the spirit of collegiality.
Anyway … if its one thing I laugh at on HC is posters saying they are going to take a break from posting and then 10 minutes later they are back at it. And I have just done it myself. But this time ….. I will resist … I will resist.
Good luck to all and sundry.
- Forums
- ASX - By Stock
- Statistical issues in the analysis of BITs P2 trial
Lol maybe I am a Shakespearean fool – a clever peasant trying to...
-
- There are more pages in this discussion • 1 more message in this thread...
You’re viewing a single post only. To view the entire thread just sign in or Join Now (FREE)
Featured News
Add BIT (ASX) to my watchlist
|
|||||
Last
2.6¢ |
Change
-0.001(3.70%) |
Mkt cap ! $24.36M |
Open | High | Low | Value | Volume |
2.7¢ | 2.7¢ | 2.6¢ | $4.605K | 176.6K |
Buyers (Bids)
No. | Vol. | Price($) |
---|---|---|
3 | 499991 | 2.6¢ |
Sellers (Offers)
Price($) | Vol. | No. |
---|---|---|
2.8¢ | 189421 | 3 |
View Market Depth
No. | Vol. | Price($) |
---|---|---|
3 | 499991 | 0.026 |
8 | 884000 | 0.025 |
2 | 3373333 | 0.024 |
2 | 153913 | 0.023 |
5 | 975000 | 0.022 |
Price($) | Vol. | No. |
---|---|---|
0.028 | 189421 | 3 |
0.029 | 305600 | 3 |
0.030 | 47636 | 2 |
0.031 | 159129 | 3 |
0.032 | 500000 | 1 |
Last trade - 12.24pm 20/08/2024 (20 minute delay) ? |
Featured News
BIT (ASX) Chart |