Training physicians and health care providers to accurately read coronary arteriograms. A training program.
UNLABELLED: 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 < or = 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 overestimation of "severe" disease and underestimation of "less severe" disease, was reduced with statistical improvement (P < or = 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. 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 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.