Proctorship for CAS training: a pilot study of safety and reproducibility

The literature continues reporting a high complication rate for carotid artery stenting (CAS) during the learning-curve phase (LCP). The aim of this study was to report a simple and reproducible method designed to improve CAS results during the LCP.

Between February 2007 and December 2009, a qualified vascular surgeon ran a proctorship program for CAS. The program was divided into four practical phases: in the teaching phase (a) the first 20 CAS were performed by the proctor assisted by a trainee surgeon; in the training phase (b) for the 21st to the 50th CAS the trainee surgeon was supervised by the proctor; in the skilled phase (c), between the 51st and the 80th procedure, a trainee surgeon performed CAS while the proctor was scrubbed-in but operating only on demand; in the final phase (d), following the 81st CAS, the procedure was performed without the proctor's presence. The inclusion criterion was carotid stenosis ≥70% and patient selection was performed for the first 40 cases based on patient and lesion characteristics. The procedure for CAS was standardized.

Four trainees performed 604 CASs in two centers. The procedural success rate of CAS was 98.8% (N.=594/604) without any differences among the four trainees (P=0.902). The overall TIA, myocardial infarction, minor, major and fatal stroke rate at 30 days was respectively 1.7% (N.=10), 0.8% (N.=5), 1.2% (N.=7), 0.64% (N.=4) and 0.3% (N.=2). The effectiveness of this program was demonstrated by a significant decrease in the proctor's intervention between phase b and phase c (P<0.001) and by a similar trend in the complication rate achieved by the four trainees, in all phases and centers (P=0.075 and 0.788, respectively).

This preliminary experience of a proctorship program in the LCP, together with patient selection and standardization of the procedure and materials used, seems to be safe and reproducible. Moreover, possibly randomized, studies comparing different CAS training techniques are needed in order to validate our findings.