Anavex Life Sciences Corp (AVXL)

[Steady_T]

Your points are well made and accurate about the risks of an interim analysis of the trial.

If you take an interim look there is the usual 5% possibility that the results will be negative even though the effect is real. That risk is increased as you pointed out by the lower numbers involved in the analysis.

Add to that when the full trial is analyzed there is a second 5% possibility that the results will be negative even though the effect is real.

In simple terms there are now 2 opportunities to fail. There are several statistical techniques to deal with this and other risks involved with interim analysis most of which actually involve tighter alpha values.

Below are the effects of several different statistical approaches. As you can see the alpha values are changed and generally not in favor of the single end of trial approach.

Approaches for alpha adjustment

An alpha adjustment is needed to preserve the overall Type I error rate. Not surprisingly, researchers have established different methods to account for multiple analyses and ways to adjust the alpha. There isn’t any one consensus but there are a few that are commonly used.

Pocock o Same alpha for interim and final analysis o 2 analyses, a = 0.0294
Haybittle-Peto o Very strict alpha adjustment at interim, no adjustment at final o 2 analyses, a = 0.002 at interim and a = 0.05 at final
O’Brien-Fleming o Strict alpha adjustment at interim, small adjustment at final o 2 analyses, a = 0.0054 at interim and a = 0.0492 at final.

Add to that the number of tests and things that would have to be done twice instead of just once at the end of the trial. There are MRIs and more importantly CSF fluid collections. I doubt that most people would want to have an additional lumbar puncture done.