Accuracy of Surgeon Self-Reflection on Hysterectomy Quality Metrics.

Abstract

To evaluate the accuracy of gynecologic surgeons' self-reflection across hysterectomy case volume, proportion of cases performed using a minimally invasive approach (minimally invasive rate), and complication rate and to assess whether accuracy is associated with specific surgeon or practice characteristics. This was a cross-sectional cohort study of gynecologic surgeons at eight Canadian hospitals between 2016 and 2019. Surgeons estimated case volume, minimally invasive rate, and complication rate for hysterectomies for a 6-month period using an online survey. Kendall's tau-beta correlation coefficient (τ) measured association between estimated and actual performance. Differences (delta) between each surgeon's estimated and actual performance were calculated. The central tendency of differences among the cohort was represented by a median (median delta) and compared with 0 (perfect accuracy) using the Wilcoxon signed rank test. Differences in characteristics between surgeons classified as underestimators, accurate estimators, and overestimators by tertile of delta were evaluated using analysis of variance and χ2 tests. Eighty-four surgeons across eight hospitals were included. Association between estimated and actual performance was moderate for case volume (τ=0.46, P<.001) and minimally invasive rate (τ=0.52, P<.001) and weak for complication rate (τ=0.14, P=.080). Surgeons underestimated their complication rate (median delta -7.0%, 95% CI -11.0% to -3.5%, P<.001) but accurately estimated case volume (median delta 1.0, 95% CI 0.0-2.5, P=.082) and minimally invasive rate (median delta 4.0%, 95% CI -4.5% to 10.0%, P=.337). Surgeons who underestimated their complication rates had higher average complication rates (33.7%) than those who estimated accurately (12.1%, P<.001) or overestimated (7.7%, P<.001) and were more likely to be fellowship-trained (P<.001). Attending gynecologic surgeons inaccurately reflect on their complication rates, and those who most underestimate their complication rates have higher rates than their peers.