Using Cumulative Summation to Draw Otolaryngology Trainees’ Learning Curves in Tonsillectomy

Abstract

Objectives:

Assess the applicability of cumulative summation (CUSUM) in 1) showing trainees progress, 2) achieving competency, and 3) highlighting concerns in tonsillectomy surgery.

Methods:

We followed 9 otolaryngology junior trainees in their first 6-12 months in the specialty between 2011-2013. Tonsillectomy is the most common operation that introduces trainees to operative otolaryngology. Two outcome measures were used: operative time and post-operative bleeding. These were then compared to experienced surgeons (>100 tonsillectomies). For time CUSUM, the mean and standard deviation of the experts’ performance was used as the factor to draw the curve applying the formula (C_n=0, C_n-1+X_n-k) on consecutive procedures. k=0.2 to allow for 20% incidence of overrunning. For bleeding CUSUM, any post-operative bleed was regarded a negative outcome and k=0.05 to allow for 5% incidence.

Results:

Trainees performed 14-35 tonsillectomies each (ongoing). The average time for experts over 150 procedures was 23min (SD:11). Trainees’ operative time was higher (38min, SD:16, P<0.01). Xn=0 if the time was within 1SD of the experts (34min) and 1 if higher. Trainees’ time CUSUM plateaued after a variable number of cases (20-30) while that of experts remained low. For bleeding CUSUM, Xn=1 if reported and 0 if not. Bleeding was rare and did not reflect the same pattern.

Conclusions:

CUSUM using time can be used to monitor performance and draw learning curves for tonsillectomy. Bleeding is a rare complication and hence needs larger numbers to show improvement. The flexibility of the CUSUM makes it adaptable to any outcome. It allows early detection of poor performance to instigate intervention.