Do DD
(or can I call you E? :-))
It's hard to really be sure what Tercica is saying when reading the press release. When they say that
"In 22 patients with severe Primary IGFD, up to four times daily blood glucose measurements were made before and for up to seven days after initiation of Increlex treatment. The results showed no statistically significant difference in the frequency of hypoglycemia observed before versus after the start of Increlex (7.0% and 8.8%, respectively; p=0.48)."
we have no idea how many days they studied the patients prior to starting Increlex and thus no idea how many blood samples they took (population size from which to derive probability of hypoglycemia). We have a rough idea of how long they continued doing so after treatment began (up to 7 days). So after treatment, they most likely had at least a few hundred samples.
If you have a link to a site where there is more info about this study, I'd be happy to see it.
Most importantly, they don't say anything about how the blood samples were analyzed (statistically, that is). Consequently, it's hard to say just how sensitive the test was... did they do any form of multiple regression that could take into account the individual patient idiosyncracies, possible diurnal variation etc. Or did they lump all of the data from each patient into one bin to get a frequency of hypoglycemic episodes?
The P value usually refers to whether the null hypothesis (generally that there is no significant difference between treatment groups) can be rejected. A p value of 0.48 pretty much shouts out "Not significant". As for correlation, I'm used to r values being used as the variable, but even an r value of 0.48 would be pretty poor correlation, if I recall my stats courses correctly. Mind you, it's been many years.
Best wishes,
GS
Market Data 
Markets 
AI
