Evaluation of the efficacy of hand-held computer screens for cardiologists’ interpretations of 12-lead electrocardiograms

• Published in: The American heart journal Vol. 138; no. 4; pp. 765 - 770
• Main Authors: Pettis, Karlton S., Savona, Michael R., Leibrandt, Paul N., Maynard, Charles, Lawson, William T., Gates, Kathy B., Wagner, Galen S.
• Format: Journal Article
• Language: English
• Published: New York, NY Mosby, Inc 01-10-1999; Elsevier
• Subjects: Biological and medical sciences; Cardiology. Vascular system; Coronary heart disease; Data Display; Electrocardiography - instrumentation; Electrocardiography - methods; Electrocardiography - statistics & numerical data; Heart; Heart Diseases - diagnosis; Heart Diseases - epidemiology; Humans; Medical sciences; Microcomputers; Observer Variation; Telemedicine; Human; Infarct; Acute; Screen; Physician; Cardiovascular disease; Coronary heart disease; Myocardial disease; Electrodiagnosis; Electrocardiography; Myocardium; Computer; Diagnosis; Cardiology; Medical specialty
• ISSN: 0002-8703; 1097-6744
• DOI: 10.1016/S0002-8703(99)70194-8

Background Newly designed computer-based applications and the development of wireless technology have allowed the transmission of 12-lead electrocardiogram (ECG) waveforms from remote locations to the hand-held computers of cardiologists. If these computer ECGs can be reliably interpreted, then the time to treatment for cardiac patients may be reduced.
Methods and Results Twenty classic examples of cardiac abnormalities were chosen to test the efficacy of the hand-held computer’s liquid crystal display (LCD) screen in the interpretation of 12-lead ECGs. Ten cardiologists interpreted these 20 ECGs on the hand-held computers and then twice later on traditional printed paper. The control intraobserver agreement between the sets of paper-displayed ECGs was measured against the agreement between each of the paper sets and the LCD-displayed set of ECGs. Eighty-nine percent (178/200) of the ECGs were interpreted identically by the participants between the 2 paper sets. When comparing the interpretations of the LCD-displayed ECGs with those of each of the paper sets of ECGs, 88.0% (176/200) and 87.5% (175/200) of identical diagnoses were noted. These differences of 1.0% and 1.5% in intraobserver agreement between paper-to-paper and each of the 2 paper-to-LCD comparisons were not significant (
P = .75 and
P = .88, respectively).
Conclusions The strong intraobserver agreement shows that cardiologists make the same diagnoses when viewing LCD-displayed ECGs as they do when viewing paper-displayed ECGs. A study to measure the intraobserver agreement of the decision regarding administration of reperfusion therapy after interpretation of ECGs of patients with acute chest pain is now underway. (Am Heart J 1999;138:765-70.)