In order to determine the incidence and significance of the various auscultatory findings in hypertension, a clinical and phonocardiographic study was carried out in 100 patients. The patients were observed over 18 months to two years, and the necropsy findings in those who died have been correlated with the clinical signs.
METHODSThe patients had average casual blood pressure readings of over 180 mm. Hg; 69 per cent were out-patients and 31 per cent in-patients. Following the classification of Keith et al. (1939), 20 per cent were Grade I, 41 per cent Grade II, 16 per cent Grade III, and 23 per cent Grade IV. Eighty per cent showed both electrocardiographic and radiological evidence of left ventricular hypertrophy, 18 per cent showed radiological but no electrocardiographic abnormality, and 2 per cent showed left ventricular hypertrophy on the electrocardiogram but not on the X-ray. Thus the study represents the findings in moderately severe hypertension.The clinical examination was carried out by each of us within the same hour and the findings were recorded independently and subsequently compared. The only physical sign on which we frequently differed was concerning the atrial sound: this problem is discussed fully in the text. Where there was disagreement over other signs a re-examination or phonocardiogram decided the issue.Phonocardiograms were performed with an apparatus with frequency response specifications similar to those recommended by Leatham (1952) for clinical work. Standard lead Il was employed throughout for the simultaneous electrocardiogram. In cases where summation confused the interpretation of added sounds, it was usually possible to determine their origin, either by slowing the heart with carotid sinus pressure or by serial phonocardiography after clinical improvement.
RESULTSThe Atrial Sound. Audible sounds occurring 0 07 second or later after the beginning of the P wave, but preceding the beginning of the QRS complex of the simultaneous electrocardiogram, were accepted as atrial sounds. Care was taken to exclude third heart sounds which may be constantly presystolic in timing if they occur relatively late and there is a moderate tachycardia. The interpretation of a phonocardiogram in terms of audibility of any vibration may be difficult, but with any given recording system it is usually possible to distinguish between the comparatively higher-frequency audible atrial sound and the low-frequency presystolic vibrations that may occur normally and precede the main audible component of the atrial sound when this is present in a pathological heart. The true incidence of atrial sounds is difficult to assess. Clinically it may be impossible to distinguish an atrial sound from a "split" first heart sound in which the initial component is low pitched (Kincaid-Smith and Barlow, 1959 a). We have shown that with clinical improvement the