Gravar-mail: Incorrect statistical method in parallel-groups RCT led to unsubstantiated conclusions