EDITOR, - The editorial on syncope by M C Petch invites comment.1 The thrust is to oversimplify the diagnosis of syncope. We agree that this may be appropriate but depends on the history from patient and observer. However, in a substantial minority there are real difficulties. In this context Petch's definition of malignant vasovagal syncope does not match either the British or the North American definition.2 The British definition hinges on the lack of warning experienced, which results in injury in attacks, and the American definition emphasises the duration of asystole on tilt testing. For these patients testing is necessary, and until the mid-1980s there was no test that could, in the laboratory, reproduce vasovagal syncope. In 1986 Kenny first showed the clinical value of tilt testing and provided the basis for the British definition.3 Since then tilt testing has spread worldwide, which reflects not …