Scottwny, it is a statistical conclusion based on trials being underpowered.
Think of it this way. The drug has a beneficial effect of some measurable size in the WHOLE patient population. A sample of certain minimum size is needed to ensure there is a reasonable probability the differences is present also in the sample (trial) population. The probability that the difference will show up in a given trial population is called power. A trial with 80 % power will miss the difference 20 % of times. If one wants to have 90 % probability that the existing differences is captured within the trial population the size of the sample must be increased significantly. And that is costly.
Few early p2 (p2a) trials have even 80 % power. For P3 trials 80 % is considered the bare minimum and hence 'bad form' - most P3 trials have at least 90 % power.
News 
Market Data 
Discover 
