Operative learning curve trajectory in a cohort of surgical trainees.

Abstract

Certification of completion of training in general surgery in the UK requires demonstration of competence in index operations by means of three level 4 competence consultant-validated procedure-based assessments (PBAs). The aim of this study was to evaluate the trajectory of operative learning curves related to PBA performance levels for curriculum-defined indicative operations with respect to numbers performed and training time. Logbook data from consecutive higher general surgical trainees were compared with PBA evaluations to determine the relationship between PBA performance level, operative experience, training time and indicative numbers. Learning curve gradients were calculated using the inverse trigonometric function of tan related to operative experience and training time. Eighty-four surgical trainees participated. Median caseload to achieve three level 4 competence assessments was 64 (range 18-110) for inguinal hernia, 83 (15-177) for emergency laparotomy, 87 (23-192) for laparoscopic cholecystectomy, 95 (22-209) for appendicectomy, 45 (17-111) for segmental colectomy and 16 (6-28) for Hartmann's procedure. Median learning curve gradients to achieve level 4 competence for emergency laparotomy were 15·3° and 33·7° by caseload and training time respectively, compared with 73·3° and 59·9° for Hartmann's procedure. Significant variance was observed in the gradients of all learning curves related to both the caseload between the first level 3 and the first level 4 PBA (P = 0·001), and between the first and third level 4 PBAs (P < 0·001). Significant learning curve gradient variance was observed, with discrepancies between expected indicative operative numbers and the point at which competence was judged to have been achieved. Numbers of index operations to achieve certification of completion of training warrant further examination.