Replies to post #149681 on Biotech Values

[Summer2762]

If I put aside all the shenanigans about dropping the crossover patients, I am actually very intrigued by slide #16.

http://www.thresholdpharm.com/pdf/supplemental_info_survival_pancreas_cancer_09-12.pdf

MOS in second line > 13m? Which is larger than 1st line MOS? How on earth did the co explain that?

PS. Yes. I know. It seems like a perfect example of meaninglessness of p value for repeated subset analysis... I am still intrigued though.