Computers Are of Limited Value in the Interpretation of...
Computers Are of Limited Value in the Interpretation of...
The article by my friend Charles Fisch in the November 15, 2000, issue of the Journal of the American College of Cardiology summarizes the development of the electrocardiograph machine and discusses the gradual acquisition of our knowledge about the waves that are recorded. This article should be read by all individuals who interpret electrocardiograms. I support and emphasize Dr. Fisch's admonition that the computer readout has, thus far, been a great disappointment.
The error rate made by computers is unacceptable. Fisch states: "Its role in interpreting the clinical ECG is limited." I have always stressed that the clinician's goal is to identify in the electrocardiogram the type of heart disease present in the patient. Computers do not usually do that. For example, the computer may "diagnose" right bundle branch block with left anterior superior block, but it does not give a clinical differential diagnosis as to the heart diseases that cause the electrocardiographic abnormality.
Computer software is rarely updated. Many computer programs are 25 years old. Much has happened in electrocardiography during the last decade and, regrettably, the computer never takes a "refresher course" in electrocardiography.
Finally, I favor the Grant method of interpretation rather than the memorization approach to interpretation. I believe that basic principles of electrocardiography, which include the knowledge and use of vector concepts, should be used to interpret each electrocardiogram. The computer, thus far, cannot determine the antero-posterior direction of electrical forces. Some computers record the frontal plane direction of electrical forces, but they do not do so reliably.
I agree with Dr. Fisch that the computer, thus far, "may be, in fact, an obstacle to the acquisition of ECG skills." Surely, someday there will be computer software that will be more accurate and useful for the clinician. Even then we must be careful, because too much help of that type will cause brain atrophy and stifle research.
The article by my friend Charles Fisch in the November 15, 2000, issue of the Journal of the American College of Cardiology summarizes the development of the electrocardiograph machine and discusses the gradual acquisition of our knowledge about the waves that are recorded. This article should be read by all individuals who interpret electrocardiograms. I support and emphasize Dr. Fisch's admonition that the computer readout has, thus far, been a great disappointment.
The error rate made by computers is unacceptable. Fisch states: "Its role in interpreting the clinical ECG is limited." I have always stressed that the clinician's goal is to identify in the electrocardiogram the type of heart disease present in the patient. Computers do not usually do that. For example, the computer may "diagnose" right bundle branch block with left anterior superior block, but it does not give a clinical differential diagnosis as to the heart diseases that cause the electrocardiographic abnormality.
Computer software is rarely updated. Many computer programs are 25 years old. Much has happened in electrocardiography during the last decade and, regrettably, the computer never takes a "refresher course" in electrocardiography.
Finally, I favor the Grant method of interpretation rather than the memorization approach to interpretation. I believe that basic principles of electrocardiography, which include the knowledge and use of vector concepts, should be used to interpret each electrocardiogram. The computer, thus far, cannot determine the antero-posterior direction of electrical forces. Some computers record the frontal plane direction of electrical forces, but they do not do so reliably.
I agree with Dr. Fisch that the computer, thus far, "may be, in fact, an obstacle to the acquisition of ECG skills." Surely, someday there will be computer software that will be more accurate and useful for the clinician. Even then we must be careful, because too much help of that type will cause brain atrophy and stifle research.
