Electrocardiographic patterns typical of hypopotassemia and compatible with hypopotassemia were defined on the basis of the number of electrocardiographic signs of hypopotassemia present in 2 leads (generally leads II and V3). In 50 hypopotassemic patients a good correlation was found between the electrocardiographic pattern and plasma potassium concentration. Appearance of the electrocardiographic signs of hypopotassemia was not prevented by disturbance of other plasma electrolytes or by blood pH. LLECTROCARDIOGRAMS of patientswith low concentrations of extracellular potassium frequently show characteristic abnormalities that disappear after administration of potassium salts.1' 2 Regression of the electrocardiographic pattern of hypopotassemia during administration of potassium is characterized by a gradual increase of T-wave amplitude, decrease of U-wave amplitude, and diminution of S-T segment depression in the standard limb and precordial leads without any change in Q-T or other components of the Q-U interval.' This sequence of regression suggests that the converse, the evolution of the hypopotassemia pattern, consists of a progressive decrease of T-wave amplitude, increase of U-wave amplitude, and S-T segment depression in the standard limb and precordial leads. Accordingly, a schematic construction of 5 patterns, representing 5 stages in the evolution of the electrocardiogram in hypopotassemia, has been made.' At that time no During the past fewl years we have gained an impression that in a majority of patients with hypopotassemia, electrocardiograms can be correlated with the concentration of extracellular potassium. However, since others3-7 have failed to find a correlation between the electrocardiogram and serum potassium level, it was thought that a new investigation might be helpful. We made an attempt to develop quantitative criteria for objective evaluation of the electrocardiographic changes ill these patients. We also considered certain factors other than the potassium concentration that, may influence the electrocardiogram of patients with hypopotassemia.This paper presents the results of the correlation between the electrocardiogram and plasma potassium level in 50 patients with hypopotassemia. METHODS ANXI) MATERIALFifty adult patients with plasma potassium levels below 3.5 mEq./L. were selected for the study.

The ratio of the amplitude of the QRS complex to the amplitude of the U wave varies in different leads of the same electrocardiogram. In the majority of electrocardiograms, the U wave has the same polarity as the T wave and the ratio of the U wave and T wave amplitude is relatively constant in all leads. The tallest positive U wave is usually observed in the area of leads V 2 to V 4 . The deepest negative U wave is usually observed in the area of leads V 5 to V 6 . Secondary changes of the T wave are very frequently accompanied by similar changes of the U wave. T-wave changes caused by myocardial infarction and digitalis are usually not accompanied by changes of the polarity of the U wave. The electrocardiograms of 297 cases of hypertension were divided into 6 groups on the basis of the relationship between the polarity of the T wave and the U wave. Both waves were positive in all precordial leads in 48.1 per cent of the cases. Negative U waves were found in 21.8 per cent of the cases and these were predominantly in the leads with negative T waves. A negative T wave in the left precordial leads was accompanied most frequently by a negative, less frequently by an isoelectric, and least frequently by a positive, U wave. An inverted U wave in the presence of an upright T wave was found in only 2.8 per cent of the cases. A change from a negative to a positive or isoelectric U wave was observed after slowing of the heart rate, a drop in blood pressure, and nitroglycerin administration.

NEGATIVE U waves are usually preceded by negative T waves. It also seems that only secondarily inverted T waves are followed by inverted U waves. It was found by us, however, that the U wave was inverted in less than one half of the electrocardiograms that showed presumably secondary inversion of the T wave. The presence of an upright U wave in leads where the T wave was already inverted could be explained by the existence of a time lag between the inversion of the T wave and the inversion of the U wave in the evolution of a "left ventricular strain pattern." An isolectric U wave would probably signify a transitional phase preceding actual U wave inversion. Our previous clinical observations tended to support such a concept. If this assumption was correct and the occurrence of an inverted U wave was a late finding in 'left ventricular strain patterns," then an inverted U wave should be associated with clinical findings of more advanced disease of the left ventricle. METHODS AND MATERIALSBecause the majority of negative U waves are found in electrocardiograms of patients with hypertension, a group of 287 patients with elevated blood pressures who had electrocardiograms in which the U-wave amplitude and polarity could be accurately determined were selected from the files of the Division of Cardiology of the Philadelphia General Hospital. The method of selection was described in the previous communication.' The interpretations of the electrocardiograms were performed without knowledge of the clinical findings. The latter were obtained later from the hospital records.The electrocardiographic diagnosis ( "left ventricular hypertrophy" was based on the presence of left axis deviation with the sum of the deepest S wave in the right precordial leads plus the tallest R wave in the left precordial leads exceeding 35 mm. without T-wave inversion. "Left ventricular strain" was defined as the left ventricular hypertrophy pattern plus T-wave inversion in the left precordial leads.In analysing the clinical findings, the degree of congestive heart failure was determined on the basis of the physical examination performed at the time the electrocardiogram was taken. Renal function was determined by urinalyses and blood urea nitrogen values, (available in all cases), phenolsulfonphthalein tests, Fishberg concentration tests, intravenous pyelography, blood creatinine levels, and urea clearance tests (obtained in many cases). Heart size was determined by teleroentgenogram or orthodiagram. The fundoscopic examinations were performed by a staff ophthalmologist.In evaluating the mortality (table 2), all deaths occurring within 6 months of the time the electrocardiogram was taken were recorded and the cause of death noted. RESULTSThe sex, color, and age distribution of the cases studied are presented in table 3. The distribution by sex and color was about the same for both the negative and positive U-wave groups. The age range and distribution were also similar. It should be noted, however, that the Negro women, who made up the largest ...

No abstract

No abstract