Ann: HSV-2 Interim Results Announcement-AHZ.AX, page-154

Hi Daniel03 ....

A reduction of X% in your treatment group is good or bad depending what the reduction was in your placebo group.

In order to determine this you run a statistical model ... post viral shedding as your dependent variable by group controlling for baseline.

But go back a step.

I am assuming in my post that a statistical model was in fact run.

This is because of the use of the term “error bar” which implies statistical inference; it is the standard error on which P values are calculated from. Which is what they don’t want anyone to do.

The problem is the 58% reduction figure is being used purportedly to illustrate a treatment effect. It needs a confidence interval or error bar to judge its precision.

Why is this important to the price of eggs?

The precision of a treatment estimate in early phase trials is predictive of the reproducibility of results in later stages trials.

The precision of the estimate is a function of sample size and variability. Variability has many sources, including measurement error and heterogeneity in the sample of patients recruited.

Generally people think lower variability is a good thing because it suggests you are measuring things reliably and/or you seem to know how to select a homogenous group of treatment responders.

So the very one thing that everyone thinks you need to assess trial results is the one thing they won’t tell you. So obviously the estimate is pretty imprecise.

Which leads to the other oddity. What to do in the face of imprecision of an early phase trial result?

AHZ have opted for the phone a friend method. Independent experts have apparently looked at the results and decided it is worth continuing with.

But just like on the TV show this method is highly dependent on the friend you ring.

In this case we don’t know who these friends are. And at the end of the day companies can’t really outsource trial development risk to other (unnamed) people. Its their call; not someone else’s.

Quite a bit of ingenuity goes into micro-cap trial early phase (1/2) result announcements. They always seem so positive. Mice don’t lie sort of thing.

But around the traps this has many punters pondering. How come after a decade of positive trial results the share price of their favourite biotech is a fraction of what it was when they started?

Answer - ASX P1/2 trial result announcements are principally for investor puffery.