Accounting for individual variability in baseline seizure frequencies when planning randomised clinical trials remains challenging