The use of fluoroscopy to construct learning curves for coronary angiography

The aim of this study was to assess learning curves for coronary angiography using registry data from the Swedish Coronary Angiography and Angioplasty Registry (SCAAR).

ACC and ESC guidelines for cardiologists in training recommend 200-300 coronary angiographies as primary operator. Whether this is safe or sufficient to reach an adequate proficiency level is not known. The development of learning curves and factors that can determine progress is not yet clearly stated.

We extracted data from SCAAR 2005-2009 and identified 20 novel and 21 expert operators in coronary angiography during this observation period. Metrics possibly influenced by proficiency level were compared between the two groups. Learning curves were then identified with the expert's performance as reference defined as interquartile range (IQR).

Data from a total of 24,000 coronary angiographies were examined. Beginners used similar volumes of contrast but had longer fluoroscopy time compared with experts. Fluoroscopy time appeared to be a metric that demonstrated a clear learning curve and beginners reached experts IQR in 50 % (median) of the procedures after ∼140 coronary angiographies. The risk of complications was independently associated with fluoroscopy time.

Fluoroscopy time seems to be the best metric to determine coronary angiography performance level and might therefore be a good proficiency measure during training. On the basis of our results we recommend the trainee to perform at least 150 coronary angiographies during supervision before proceeding with unsupervised procedures. It is not clear if the suggested number of procedures will result in decreased number of complications but the data suggests that fluoroscopy time might be a surrogate marker for complications.