Replies to post #2323 on Insmed Inc (INSM)

[Do DD]

GS,

Yup, you're right concerning the definition of p. Thanks for the refresher.

E.
Also more than 20 years out of statistics classes.

[rfoable1]

GS,
They are demonstrating that they are aware of the concerns about hypoglycemia with Increlex.

They are trying to broadcast that the 1.8% increase in hypos does not meet their statistical criteria for blaming it on Increlex. (The null hypothesis, that there is no correlation, was not rejected)

I doubt we are going to have any ability to dig up how they composed the sample population, but using 22 patients makes it a fairly low power calculation anyway. If they had the same result from all their 151 patients, it would have a lot more power.

A piece of trivia: Has anyone noticed the odd coincidence that the total number of patients on Increlex they reported matches the number of their "increased anabolic response" patent? ('151)

cheers

[schwabd]

I've spent the last 15 years selling to physicians. P values greater than 0.05 are considered not statistically significant. TRCA's 0.48 is laughable. Most physicians will tear up any study with such a terrible p value... and be reluctant to trust the results of the study.

[jellybean]

So, they followed patients for 7 days after start of treatment. Now is that start at the low dosages, or start at the higher, effective dose? Remember that all patients are titrated into full dose treatment over a period of time.

And, even though Chip denied this on the conference call, one of the big banks (I want to say Lehman) stated in a report that Tercica provided one month of treatment free of charge.