j. ELECTROCARDIOLOGY 15 (1), 1982, 55-60 A Classification of Ventricular Ectopic Beats BY SIMON W. RABKIN, M.D., F.R.C.P.(C), F.A.C.C. AND MINORU OHMAE,* M.D. SUMMARY To develop a classification of ventricular ectopic beats based on the QRS complex, the frontal and precordial QRS vectors were examined in ECGs from 71 patients. The data were first analyzed in planar coordinates to permit comparison with some previous classifications. Next, because the QRS vector is angular data, the data were trans- formed to spherical coordinates. Multivariate statistical methods were used. Cluster analysis aided in the separation of groups. Although at least four groups were suggested using planar coordinates, only three groupings were identified on spherical coordinates. A classification function was developed that permits assignment of eetopics into these three groups. It correctly classified 97.2% of cases. Thus, this study provides a quantita- tive method that significantly defines groups of ventricular ectopic beats. Classification of ventricular ectopic beats on the basis of their configuration on the surface ECG has had a long and controversial history. 1 These classifications have had essentially a qual- itative basis: this is exemplified by the one most often used which dichotomizes ectopics into those with either a right or left bundle branch block pattern. 1-s They may be further subdivided ac- cording to the frontal plane QRS vector. 9-1~ The utility of these approaches has been ques- tioned. ~,3,4,6-s Because of the essentially qualita- tive nature of previous classifications and the po- tential importance of more appropriate classifica- tion, we undertook an examination of ventricular ectopic beats using multivariate statistical methods to develop a classification based on the QRS vector. MATERIAL AND METHODS Study Population Seventy-one patients represented a consecutive series of all those with ventricular ectopic beats (VPB) with the same form recorded in any three limb and all precordial leads in ECGs taken during an 8 month period in a hospital-based electrocardiographic laboratory. From the Department of Medicine (Section of Cardiology) and Social & Preventive Medicine, University of Manitoba. *Third Division, Department of Medicine, Kyoto University, Kyoto, Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. w 1734 solely to indicate this fact. Reprint requests to: Dr. S. W. Rabkin, Faculty of Medicine, Dept. of Medicine (Cardiology), University of British Colum- bia, 4500 Oak Street, Vancouver, BC, Canada V6H 3N1. ECGS were done on a Marquette Series 3050 ma- chine that records simultaneously leads I, II, III; then leads AVR, AVL, AVF, V1, V2, V3; and lastly V4, V5, V6. The diagnostic criteria for ventricular ectopics were that: (1) they were premature in occurrence; (2) there was no preceding premature p wave; (3) the QRS duration was ~> 0.12 sec; and (4) the QRS and ST-T morphology was different from the dominant rhythm. 1,2 Data Collection The following data were recorded for each ECG: (1) for VPB: (a) prematurity m the coupling (R-R ~) inter- val, which is the difference between the interval from the preceding R wave and the R wave of a premature ventricular ectopic (R 1) divided by the QT interval of the complex preceding the one used to calculate the R-R ~ interval; (b) the QRS duration; (c) the mean fron- tal plane QRS vector; (d) the frontal plane 0.02, 0.04, 0.06, 0.08 sec QRS vector; (e) the mean precordial plane QRS vector; and (f) the precordial plane initial 0.02, 0.04, 0.06, 0.08 sec QRS vector; and (2) for dominant rhythm: (a) the QRS duration; (b) the mean frontal plane QRS vector and (c) the mean precordial plane QRS vector. The QRS vector was determined in the frontal plane using a standard hexaxial reference system, and in pre- cordial leads considering V6 as 0~ and V2 as +90012 The magnitude of QRS in each lead at a designated time was measured and the vector was determined from tables constructed for any 2 lead combinations based on the equations: ~3 0 = cos -~ (a/d) where 0 = angle between the mean electrical axis and a given lead; d= X/a2+b 2-2abcos0 sin 0 where a- magnitude on 1 lead b- magnitude on second lead 0- angle between the leads. 55