Evaluation of technician audiovisual scanning of ambulatory electrocardiograph recordings utilizing the rapid oscillographic printout technique of validation

Document Type

Journal Article

Publication Date



Clinical Cardiology








ambulatory electrocardiography; dynamic electrocardiography; holter monitoring; rapid oscillographic printout technique; technician audiovisual scanning


Summary: The accuracy of technician audiovisual scanning of ambulatory electrocardiographic recordings, coupled with physician review of demonstrative rhythm strips, was validated using the rapid oscillographic printout technique. Seventy‐five patients underwent 24‐h ambulatory electrocardiographic recordings which were routinely evaluated by audiovisual scanning for the highest hourly grades of ventricular ectopic activity (VEA). Printouts of three random hours per 24‐hours were then obtained (one hour per 8‐hour period). The total beat oscillographic printouts, written at 120 × real time, were then graded by at least two authors and compared to the results of audiovisual scanning. Hourly VEA grades were: 0 = no ventricular ectopy; 1 = ≤ 10 unifocals; 2 = > 10 unifocals; 3 = ≤ 10 multifocals; 3A = > 10 multifocals; 4 = couplets; 5 = triplets or salvos. VEA grades were further grouped into high grade (2, 3A, 4 or 5) and low grade (0, 1, or 3) categories. The comparison of rapid oscillographic total beat printout to technician audiovisual scanning coupled with physician review of ambulatory electrocardiographic data produced the following conclusions: (1) The highest hourly VEA grade was exactly identified 65% (146/225) of the time, or was closely approximated (identified within one grade of the highest VEA grade) 80% (181/225) of the time. Excluding errors that were secondary to overlooking rare electrical events, technicians successfully identified 88% (199/225) of hours for the highest VEA grade; (2) patients with high grade and low grade VEA were identified with an accuracy of 91% (51/55) and 95% (19/20), respectively; (3) technician reports which combined each of the three random hours per patient were more predictably accurate when positive for high grade VEA (98% of patients, 50/51) than for low grade VEA (79% of patients, 19/24); p < 0.05); (4) “underreading” errors in audiovisual scanning were almost six times more frequent as those of “overreading” (67 vs. 12 instances, respectively), with a failure to recognize low frequency electrocardiographic events as the most prevalent type of error. This study demonstrates the use of the rapid oscillographic printout technique for the display of ambulatory electrocardiographic data in hard copy compressed form which was successfully used to validate technician performance in the audiovisual scanning of ambulatory ECG recordings. Copyright © 1982 Wiley Periodicals, Inc.