Anavex Life Sciences Corp (AVXL)

[boi568]

Let me try to explain my position another way.

Doing a clinical trial of a representative population, placebo-controlled and double-blinded, is simply taking a sample of the actual answer in as neutral a form as possible. You pick your subjects to faithfully represent the population intended as the medical target; you use a placebo to isolate the medical effect from any baseline biases (such as deterioration over time) and the effects of hope in the subjects; you double-blind to avoid contaminating the results, especially when there are subjective standards, to prevent human biases from influencing the outcome.

In other words, you do everything possible to take a fair sample of the whole truth. Then you let statistics take over to inform you as to how sure you can be that your sample comes close enough to the whole truth. The scientific community is not willing to accept more than a five percent chance that you may be wrong. The results will be the results, and you can rely upon them to be right at least 95% of the time, or you don't pass the trial.

If you peek, you get a smaller sample size. That makes it harder to get your 95% confidence level, since your underlying performance has to be that much better. Keep in mind that you may have a proper therapeutic, but you will have more uncertainty -- more Heisenberg -- interfering with your results. Maybe enough to invalidate your trial, randomly, even though holding off would have given you the stats to pass. There are other issues that arise, but I am not the one to explain them. Your best odds of winning a trial with a good drug is to let it play out.

A couple of other points: As time goes on and we get closer to the final readout date, the n increases and the uncertainties in an interim look decrease relative to a final readout. The later the interim look, the less problems it can create. Second, I would understand if Missling asked for a look in the context of a hostile takeover. At that point, the equities being balanced can tip in favor of the earlier readout. But absent that circumstance, I have every confidence Anavex will see the 2b/3 trial through in the form that most assures its success.