Gravar-mail: Avoidable statistical pitfalls in analyzing length of stay in intensive care units or hospitals