The tyranny of statistics in medicine: a critique of unthinking adherence to an arbitrary p value