General Surgical Resident Operative Autonomy vs Patient Outcomes: Are we Compromising Training without Net Benefit to Hospitals or Patients?

Abstract

Resident operative autonomy has been steadily declining. The reasons are multifactorial and include concerns related to patient safety and operating room efficiency. Simultaneously, faculty have expressed that residents are less prepared for independent practice. We sought to understand the effect of decreasing resident autonomy on patient outcomes and operative duration. Retrospective study utilizing the Veterans Affairs Surgical Quality Improvement Program (VASQIP) database. Operative cases within the VASQIP database from July 1, 2004-September 30, 2019 were analyzed. All adult patients who underwent a surgical procedure from July 1, 2004 to September 30, 2019 were analyzed. The subpopulation of patients that underwent a surgical procedure in General Surgery or Peripheral Vascular Surgery were identified based on the code of the specialty surgeon. Within these subgroups, the most frequent cases by current procedural terminology (CPT) code were selected for study inclusion. The principle CPT code of all cases was further coded by level of supervision: attending primary surgeon (AP); attending and resident (AR), or resident primary with the attending supervising but not scrubbed (RP). Baseline demographics, operative variables, and outcomes were compared between groups. The VASQIP database included 698,391 total general/vascular surgery cases. 38,483 (6%) of them were RP cases. Analysis revealed that the top 5 RP cases account for 73% of total RP volume-these include: 1) Hernias (55% total; 33% open inguinal, 13% umbilical, 5% open ventral/incisional, and 4% laparoscopic) 2) cholecystectomy (18%), 3) Amputations (17% total; 10% above knee, 7% below knee), 4) Appendectomy (7%) and 5) Open colectomy (3%). The percentage of cases at teaching hospitals that were RP cases significantly decreased from 15% in 2004 to 5% in 2019 (p < 0.001). RP cases were generally sicker as demonstrated by higher ASA classifications and more likely to be emergent cases. Operative times were also increased with resident involvement, but RP cases were faster than AR cases on average. After adjusting for baseline demographics, case type, and year of procedure, mortality was no different between groups. Complications were higher in the AR group but not in the RP group. The rate of resident autonomy in routine general surgery cases has decreased by two-thirds over the 15-year study period. Cases performed by residents without an attending surgeon scrubbed were performed faster than cases performed by a resident and attending together and there was no increase in patient morbidity or mortality when residents performed cases independently. The erosion of resident autonomy is not justified based upon operative time or patient outcomes. Efforts to increase surgical resident operative autonomy are needed.