Like chas has told us, a small variable value change in modeling can change the result greatly, from experiences of following many cancer PIII trials results, the delaying events are very often from placebo arm performing much better than assumed caused by better SOC, and trails fail.