Technical skills continue to improve beyond surgical training.
BACKGROUND: There is growing focus on surgical technical competence and the means by which we are able to measure it. Ongoing studies have shown a plateau effect with increasing experience of the operator. The aim of this study was to assess the technical competence of five groups of surgeons with increasing experience and validate a new rating tool for use in surgical assessment. METHODS: Fifty surgeons performed a saphenofemoral junction ligation on a synthetic groin model. The procedure was videotaped, blinded, and reviewed independently by three assessors. Performance was assessed using a previously validated global rating scale of generic surgical skill. In addition, each procedure was rated with the procedure-specific Imperial College Evaluation of Procedure-Specific Skill (ICEPS) rating scale to establish the construct validity (ability to differentiate on the basis of skill) and inter-observer reliability. RESULTS: Both rating scales showed improved scores with ascending grades (P < .001) and demonstrated a high inter-observer reliability both for generic and procedure-specific skill (alpha = 0.97 and alpha = 0.96, respectively). Total operative scores demonstrated significant differences between surgeons in postgraduate years 1 and 2 and surgeons in years 3 and 4 and also between newly appointed and experienced consultants (P < .041). Procedure-specific performance showed a plateau effect at the registrar level. Generic skill continued to improve, and significant differences were seen between newly appointed and senior consultants (P < .026). CONCLUSION: This study shows that surgical performance continues to improve significantly beyond consultancy, and the data suggest that generic and procedural performance continue to improve, with significant improvement in the former with increasing experience. The ICEPS rating scale demonstrates construct validity and a high inter-observer reliability supporting its use in formative and summative assessment.