Cardiac abnormalities in ischemic cerebrovascular disease.

RESPONSE BY: MASATO NISHIDE, M.D. To the Editor: The study reported by Nishide et al (Stroke 14,541-545, July-August 1983) is of interest in that atrial fibrillation and/or rheumatic heart disease (generally accepted as being associated with an increased risk of systemic thrombo-embolism) were found significantly more frequently in patients presenting with "ischemic cerebrovascular disease" than in the control subjects. It is important to emphasize that the index cases in the reported casecontrol study were only those 350 patients with "ischemic cerebrovascular disease" (mainly completed stroke) who a) were admitted to the Division of Cerebrovascular Diseases of Hanwa Memorial Hospital, b) were admitted within 48 hours of the onset of symptoms and c) had a technically successful 2D echocardiogram performed. While it is impossible to overcome the inevitable problem of technically-unsuccessful echocardiography in a minority, one wonders how many patients in Japan who have a stroke are either not admitted to hospital or are admitted elsewhere than a specialist Department of Cerebrovascular Disease (at least initially after the onset of their symptoms) and whether, therefore, the patients investigated constitute a truly representative sample of all patients with ischemic cerebrovascular disease. Similarly, before deciding that the control group (350 sexand age-matched hospital in-patients) truly represents the general population of Japan one would like to know exactly how the 350 were selected "in a blind manner'' from 1000 eligible in-patients and it would also be nice to have some idea of the spectrum of disease in the control subjects. — a study which compared the prevalence (not "incidence", by the way, as stated in the paper's introductory paragraph) of cardiac abnormalities in a group of patients with cerebrovascular disease with that in a control group consisting mainly of patients with heart disease would be a curious one indeed, while a control group consisting of, for instance, mainly surgical patients, might be expected to have a falsely-low prevalence of cardiac disease, the very presence of which might have precluded admission to hospital for consideration of surgical treatment. Without knowing the answers to these basic epidemiological questions, it would be dangerous to conclude, from these authors' results alone, that a group of unselected subjects with stroke will necessarily have a higher prevalence of significant cardiac disease than the general population.