Direct and surrogate measures of the white-coat effect: methodological aspects and clinical relevance