Replies to post #115064 on Biotech Values

[jq1234]

Yes, N=46 and N=37 for figure 6A, but per footnote: Censoring rates: docetaxel-naïve 85%; docetaxel-pretreated 78%, so very few patients left non-censored.

6B, N=14 and N=17, but Censoring rates: cabozantinib 79%; placebo 29%.

Overall, only placebo group is reasonablely representative of eventual result, the other 3 groups could be highly distorted at this point.

[exwannabe]

For figure 6A, which compares docetaxel treated and naive, the N is 46 for naive and about 25% of these show no change through 400 days, so it is on the order of 10 patients. A small sample size to be sure, but not 1 or 2. Were you referring to Figure 6B, the discontinuation part?

No, it was 1 patient who represented 24% of the trial arm. The other "9" were censored (I assume because of time).

Take a look at how K/M charts work (wiki will do it).