LETTERS TO THE EDITOR

In response to the letter from Baron and colleagues, we would like to emphasize some points of clinical importance. First, according to Grubb and Kosinski^ the paradigm of provocative testing (as opposed to descriptive testing) is that spontaneous and induced episodes are associated with the same prodromal signs or symptoms and that progressions of physiological parameters (e.g., blood pressure) are the same as those measured during spontaneous events. The main conclusion of our article is that patients with asystolic responses actually differ from nonasystolic patients in terms of frequency and severity of symptoms (including seizures, fractures, facial impacts, etc.). This finding reconciles the asystolic response with the clinical definition of malignant vasovagal syndrome.^ Moreover, this concordance between clinical history and tilt testing results suggests that our interpretation of the facts is reliable. Second, we feel that an artificially high proportion of asystolic responses may be obtained when patients undergo "aggressive" evaluation including > 60° tilt angle and/or arterial puncture for blood pressure monitoring. Therefore, we believe that concordance with clinical symptoms and tilt methodology are crucial determinants of foUow-