The Use of a Computerized ECG Interpretation System in an Epidemiologic Study^[*])^[**])

ECG tracings of 10,059 men aged 40 and over first participating in 1963 in the Israel Ischemic Heart Disease Project were interpreted by computer at the Instrumentation Field Station in Washington. The computer, using criteria developed by the Field Station, reported a high prevalence of ischemic heart disease. Computer interpretation of a sample group of ECGs was compared with that of a group of cardiologists using the same criteria. It was concluded that a computer diagnosis of normal could be accepted, but those readings denoted as abnormal by the computer had to be reinterpreted by cardiologists using uniform criteria that were modified as a result of their combined clinical experience. The final analysis of these tracings together with a sample of computer normals showed a much lower prevalence of ischemic heart disease. Differences between magnetic and standard ECG tracings are described and the advantages of computer interpretation indicated.

Die EKG-Kurven von 10 059’ über 40 Jahre alten und seit 1963 am Israel Ischemic Heart Disease Project teilnehmenden Männern wurden in der Instrumentation Field Station in Washington vom Computer interpretiert. Bei Verwendung der in dieser Untersuchungsstelle entwickelten Kriterien stufte die Maschine einen hohen Anteil in die Gruppe ischämischer Herzkrankheiten ein. In einer Stichprobe der EKG’s wurde — bei Verwendung derselben Kriterien — die Computer-Interpretation mit derjenigen einer Gruppe erfahrener Kardiologen verglichen. Dabei ergab sich, daß die Computerdiagnose »Normales EKG« ohne weiteres akzeptiert werden konnte, vom Computer als pathologisch erklärte Kurven aber von den Kardiologen reinterpretiert werden mußten, wobei einheitliche, aber auf Grund der gemeinsamen klinischen Erfahrung modifizierte Kriterien anzuwenden waren. Die endgültige Analyse dieser Kurven zusammen mit einer Stichprobe von Computer-»Normalen« ergab eine sehr viel niedrigere Prävalenzrate ischämischer Herzkrankheiten. Unterschiede zwischen magnetischen und Standard-EKG-Aufzeichnungen werden beschrieben und die Vorteile der Computer-Interpretation erörtert.

Key-Words

Computer ECG Analysis - Ischaemic Heart Disease - Epidemiology - Inter-observer Variation.

Schlüssel-Wörter

Computer-EKG-Analyse - Ischämische Herzkrankheiten - Epidemiologie - BeobachterVariation.

^* Paper presented at the Seventh International Biometric Conference, Hannover, August 17, 1970.

^** Collaborative Project of the National Heart Institutes, Bethesda/Md. and Heart Institutes, Government Hospital, Tel-Hashomer and Rambam Government Hospital, Haifa and Hadassah Medical Organization, Jerusalem. — Supported by PL 480 Counterpart Funds, Research Agreement No. 375106.

• REFERENCES

• 1 Blackburn H, Keys A, Simonson E, Rautaiiarju P, Punsar S. The electrocardiogram in population studies; a classification system. Circulation 21 1960; 1160-1175.
• 2 Caceres C. A. Electrocardiographic analysis by a computer system. Arch, intern. Med 111 1963; 196-202.
• 3 Caceres C. A, Steinberg C. A, Abraham S, Carbery W. J, McBride M. M, Tolles W. E, Rikli A. E. Computer extractions of electrocardiographic parameters. Circulation 25 1962; 356-362.
• 4 Dobrow R. J, Fieldman A, Clason W. P. C, Gorman P. A, Reinfrank R. F, Caceres C. A. Transmission of electrocardiograms from a community hospital for remote computer analysis. Amer. J. Cardiol 23 1968; 687-698.
• 5 Edeiken J. Elevation of RS-T segment, apparent or real, in right precordial leads as probable normal variant. Amer. Heart J. 48 1954; 331-339.
• 6 Epstein F. H. The epidemiology of coronary heart disease. A review. J. chron. Dis 18 1965; 735-774.
• 7 Goldberger E. Recent advances in the use of augmented unipolar extremity leads. Med. Clin. N. Amer. 34 1950; 857-867.
• 8 Goldman M. J. RS-T segment elevation in midand left precordial leads as normal variant. Amer. Heart J. 46 1953; 817-820.
• 9 Groen J. J, Medalie J. H, Neufeld H. N, Riss E, Bachrach C. A, Mount F. W, Smitii H. An epidemiologic investigation of hypertension and ischemic heart disease within a defined segment of the adult male population of Israel. Israel J. med. Sei. 4 1968; 177-194.
• 10 Hiss R. G, Lamb L. E, Allen M. F. Electrocardiographic findings in 67. 375 asymptomatic subjects. X. Normal Values. Amer. J. Cardiol 06 1960; 200-231.
• 11 Holzmann M. Klinische Elektrocardiographie. G. Thieme; Stuttgart: 1955
• 12 Korth C. Klinische Elektrokardiographie. Urban & Schwarzenberg; München: 1952
• 13 Lepesciikin E. Modern Electrocardiography. Williams and Wilkins Co; Baltimore: 1951
• 14 Medalie J. H, Neufeld H. N, Riss E, Groen J. J, Kaiin H. A, Bachrach C. A. Variations in prevalence of ischemic heart disease in defined segments of the male population of Israel. Israel J. med. Sei. 4 1968; 775-788.
• 15 Pipberger H. V, Stallman F. W, Yano K, Draper H. W. Digital computer analysis of the normal and abnormal electrocardiogram. Progr. cardiovasc. Dis 05 1963; 378-392.
• 16 Sokolow M, Friedlander R. D. The normal unipolar precordial and limb lead electrocardiogram. Amer. Heart J 38 1949; 665-687.
• 17 Wasserburger R. H, Alt W. J, Lloyd C. The normal RS-T segment elevation variant. Amer. J. Cardiol 08 1961; 184-192.
• 18 Wiener L, Rios J. C, Massumi R. A. T-wave inversion with elevated RS-T segment simulating myocardial injury. Amer. Heart J 67 1964; 684-688.
• 19 World Health Organization: Hypertension and Coronary Heart Disease: Classification and Criteria for Epidemiologic Studies. Technical Report Series, No. 168. Geneva: 1959