Training Physicians and Health Care Providers to Accurately Read Coronary Arteriograms A Training Program

Abstract

Patterns in visual interpretation of coronary arteriograms (CAs) frequently cause incorrect assessment of percent diameter stenosis (%DS). These errors result in overestimating the results of angioplasty as well as of the number of arteries significantly affected by coronary artery (CAD) disease.

Methods. Forty-one physicians, nurses, and students participated in the standardization of 45 Kodachromes (39 arteries, 6 phantoms) and 5 photographic reproductions. Eleven of the 41 participated in a three-part training program designed to eliminate errors and improve accuracy of interpreting %DS from CAs.

Results. Improvement in reading %DS was seen in 69% of CAs with statistical (P ≤ 0.05) improvement in one third of these cases, whose narrowings ranged from 4% to 84%DS. Variability of reporting was reduced in 26% of the cases. Skewing, representing an over-estimation of "severe" disease and underestimation of "less severe" disease, was reduced with statistical improvement (P ≤ 0.05) in reported %DS noted after training. Similar improvement was seen with phantoms but not in photographic images where the arterial edges were outlined. (Continued on next page)

(Abstract continued)

Conclusion. The outcomes of clinical management, invasive and interventional (mechanical, thrombolytic) procedures, as well as research studies depend in part upon the accuracy of reading %DS from CAs. Most studies to date have been completed using extremely unreliable estimates of %DS with resultant problems in data interpretation. The use of this standardized training program has led to significant improvement in accurately assessing CAD.

CONDENSED ABSTRACT

Visual interpretation of percent diameter stenosis (%DS) from coronary arteriograms results in underestimating stenosis severity for "mild" disease and overestimating "severe" disease. These patterns in reading stenosis severity yield misinterpretation of angioplasty results as well as of the number of vessels significantly diseased. Forty-one physicians, nurses, and students participated in a training program designed to reduce errors made while reading %DS from coronary arteriograms. Following the three-part training program, participants demonstrated improvement in 69% of the arteries, with statistical (P < 0.05) improvement in one third of these. The improved interpretation resulted in more accurate assessment of the severity of coronary artery disease present, both before and after interventions. Utilization of this training program in cardiology fellowship programs would result in better interpretation of disease severity by cardiologists.