Construct validity of the Surgical Autonomy Program for the training of neurosurgical residents

Abstract

OBJECTIVE There is no standard way in which physicians teach or evaluate surgical residents intraoperatively, and residents are proving to not be fully competent at core surgical procedures upon graduating. The Surgical Autonomy Program (SAP) is a novel educational model that combines a modified version of the Zwisch scale with Vygotsky’s social learning theory. The objective of this study was to establish preliminary validity evidence that SAP is a reliable measure of autonomy and a useful tool for tracking competency over time. METHODS The SAP breaks each surgical case into 4 parts, or zones of proximal development (ZPDs). Residents are evaluated on a 4-tier autonomy scale (TAGS scale) for each ZPD in every surgical case. Attendings were provided with a teaching session about SAP and identified appropriate ZPDs for surgical cases under their area of expertise. All neurosurgery residents at Duke University Hospital from July 2017 to July 2021 participated in this study. Chi-square tests and ordinal logistic regression were used for the analyses. RESULTS Between 2017 and 2021, there were 4885 cases logged by 27 residents. There were 30 attendings who evaluated residents using SAP. Faculty completed evaluations on 91% of cases. The ZPD of focus directly correlated with year of residency (postgraduate year) (χ2 = 1221.1, df = 15, p < 0.001). The autonomy level increased with year of residency (χ2 = 3553.5, df = 15, p < 0.001). An ordinal regression analysis showed that for every year increase in postgraduate year, the odds of operating at a higher level of independence was 2.16 times greater (95% CI 2.11–2.21, p < 0.001). The odds of residents performing with greater autonomy was lowest for the most complex portion of the case (ZPD3) (OR 0.18, 95% CI 0.17–0.20, p < 0.001). Residents have less autonomy with increased case complexity (χ2 = 160.28, df = 6, p < 0.001). Compared with average cases, residents were more likely to operate with greater autonomy on easy cases (OR 1.44, 95% CI 1.29–1.61, p < 0.001) and less likely to do so on difficult cases (OR 0.72, 95% CI 0.67–0.77, p < 0.001). CONCLUSIONS This study demonstrates preliminary evidence supporting the construct validity of the SAP. This tool successfully tracks resident autonomy and progress over time. The authors’ smartphone application was widely used among surgical faculty and residents, supporting integration into the perioperative workflow. Wide implementation of SAP across multiple surgical centers will aid in the movement toward a competency-based residency education system.