Avid Bioservices Inc (CDMO)

[tradero]

LOOKINs Projection. Part II: the MOS model approach

We are taking for granted that doses – Placebo and Bavi - are given alternatively. So at any moment in time the distribution of the doses should be: 50% Placebo and 50% Bavi.

Let’s take a look at an example for this distribution of “eventings”


When designing these distribution tables, It is easy to organize the data in such a way that the Median of each situation corresponds to the specified label (mind you, I said Median, not mean or average).

The question is to what extent the variations of the distribution of these data (spread or concentrated) impacts the final result. To find out, let’s compare it to other different distributions …
First to another variation of the same “spreadout” kind


and finally to the extreme case of concentrated data.


The result of the comparison goes as follows:



We can see that at most (this corresponds to comparing it with the “concentrated “ distribution (specially at the low end) the difference is about 3 weeks. But the difference between the other two “spreadout” distributions is about 1 week, which can give us an idea of the error margin of the final calculations.
Though these distributions are guesswork I find this error margin acceptable, or in other words just to keep it in mind.

If we pay some attention to these results for this particular case of MOS (Placebo 7 months, Bavi 11 months), we see that 1st lookin should take place within 2015. Now, mngt expects 1st lookin to take place later than that. This may tell us that the health of patients, or rather that the performance of Bavi is much better. So the 7 MOS for the Placebo arm or the 11 MOS for the Bavi arm should be increased. These other MOS distributions for both Placebo and Bavi are analyzed In Part III.

All these results have been obtained using a tool I have developed with Visual Basic in an ExcelSpreadSheet. Once the MOS distribution cases are designed the results come out automatically.

For results and conclusion please see PartIII