I agree. If you're suggesting that there's probably subgroups buried in the data that are less likely to benefit from TXA, or more likely to be harmed, even when given < 3 hours, and that there may be groups that have greater benefit, including at time points greater than 3 hours, then I also...