Related Experiment Videos
Reliability of computerised electrocardiographic reports
Insights
This study found computerised electrocardiogram (ECG) reporting to be reliable. While some interval measurements differed between computer and physician readings, overall diagnoses showed comparable variability to that between two physicians.
Area of Science:
- Cardiology
- Medical Informatics
- Health Technology Assessment
Background:
- Computerized interpretation of electrocardiograms (ECG) is increasingly used in clinical practice.
- Assessing the reliability of these automated systems against human expert interpretation is crucial for patient safety and diagnostic accuracy.
Purpose of the Study:
- To evaluate the reliability of the Marquette 12SL-SC computer program for ECG reporting.
- To compare the diagnostic accuracy and measurement consistency of computer-generated ECG reports with manual physician interpretations.
Main Methods:
- Fifty consecutive outpatient ECGs were analyzed.
- Two independent physicians reviewed each ECG, and their readings were compared with each other and with the Marquette 12SL-SC computer reports.
- Specific parameters including heart rate, PR interval, QT interval, and QRS duration were compared, along with overall diagnostic conclusions.
Main Results:
- No significant difference was found in heart rate measurements between physicians and the computer.
- PR and QT intervals showed significant differences between computer and physician readings, while QRS duration varied between physicians and the computer.
- Overall diagnostic agreement was 76% between physicians, 68% between Physician 1 and the computer, and 78% between Physician 2 and the computer.
- The computer never reported a normal ECG as abnormal, aligning with physician findings.
Conclusions:
- The Marquette 12SL-SC computer program demonstrates reasonable reliability in ECG reporting.
- Computer-physician variability in ECG interpretation is comparable to inter-physician variability, suggesting its utility as a diagnostic aid.
Abstract:
The aim of this study is to assess the reliability of computerised reporting of electrocardiograms (ECG). Fifty ECG performed consecutively at the outpatient department of the Penang Adventist Hospital on the Marquette 12SL-SC were studied. Two physicians independently reviewed the ECG and the manual readings were compared with each other and to the computer reports. There was no significant difference in the measurement of rate. The PR and QT intervals measured by the two physicians were similar but each was significantly different from the computer reading. The QRS duration assessed by Physician 1 was similar to the computer reading but each was significantly different from that of Physician 2. The overall diagnosis was the same between the two physicians in 76%, between Physician 1 and the computer in 68%, and between Physician 2 and the computer in 78%. No ECG was reported as normal by the computer and said to be abnormal by either physician. Thus, the computer programme is reasonably reliable in ECG reporting with computer-physician variability being comparable to inter-physician variability.