Not so. The statistician was at Amarex. The statistical analysis plan didn't provide for the possible sub analyses, so the study took a statistical hit when the signal was seen and a decision was made to perform to he sub analyses. They were pretty close, so they might have skated by if they'd had a more robust SAP - one also allowing for meta analyses so data could be pooled across studies to look for signals
News 
Market Data 
Discover 
