Abstract
Background:
Teaching hospitals that train both general surgery residents and fellows in complex general surgical oncology have become more common. This study investigates whether participation of a senior resident versus a fellow has an impact on outcomes of patients undergoing complex cancer surgery.
Methods:
Patients who underwent esophagectomy, gastrectomy, hepatectomy, or pancreatectomy between 2007-2012 with assistance from a senior resident (post-graduate years 4-5) or a fellow (post-graduate years 6-8) were identified from the ACS NSQIP. Based on age, sex, body mass index, ASA classification, diagnosis of diabetes mellitus, and smoking status, propensity-scores were created for odds of undergoing the operation assisted by a fellow. Patients were matched 1:1 based on propensity score. Postoperative outcomes including risk of major complication were compared after matching.
Results:
In total, 6,934 esophagectomies, 13,152 gastrectomies, 4,927 hepatectomies, and 8,040 pancreatectomies were performed with assistance of a senior resident or fellow. After matching, overall rates of major complications were equivalent across all four anatomic locations between cases performed with the participation of a senior resident versus a surgical fellow: esophagectomy (37.0%% vs 31.6%, p=0.10), gastrectomy (22.6% vs 22.3%, p=0.93), hepatectomy (15.8% v 16.0%, p=0.91), and pancreatectomy (23.9% vs 25.2%, p=0.48). Operative time was shorter for gastrectomy (212 vs. 232 min; p=0.004) involving a resident compared to a fellow, but comparable for esophagectomy (330 vs. 336 min; p=0.41), hepatectomy (217 vs. 219 min; p=0.85), and pancreatectomy (320 vs. 330 min; p=0.43).
Conclusions:
Senior resident participation in complex cancer operations does not appear to negatively impact operative time or postoperative outcomes. Future research is needed to further assess this domain of surgical practice and education, particularly with regard to case selection and operative complexity.
Keywords: surgical oncology, surgical education, surgical residency, surgical fellowship
Introduction
Optimal management of intra-abdominal and intra-thoracic tumors, including gastric, esophageal, pancreatic and hepatic malignancies, requires multidisciplinary treatment and extensive integration of evolving knowledge across medical and surgical specialties. Surgical fellowship is an opportunity for trainees to develop the technical prowess necessary for advanced cancer resections, build a holistic understanding of multidisciplinary treatment modalities, and gain necessary experience in delivering multifaceted care to oncologic patients.1 Historically hepato-pancreato-biliary (HPB), transplant and thoracic fellowships have been integral in training surgeons to treat oncologic disease. Unofficial fellowships in complex surgical oncology have been documented as early as 1947, under the oversight of the James Ewing Society that is now known as the Society of Surgical Oncology (SSO).2,3 Formal recognition of Complex General Surgical Oncology (CGSO) as a subspecialty was granted by the American Board of Surgery and American Board of Medical Specialties (ABMS) in 2011.4 To date, there are 36 Accreditation Council for Graduate Medical Education (ACGME) accredited programs.5
As the number of fellowship programs and positions continue to increase, there has been variable concern that co-existing fellowship programs may degrade resident operative experiences and clinical training, especially as operative training in complex HPB among general surgery residents is limited at baseline.6 Principally, this concern is derived from the possibility that fellow participation will be prioritized given potential positive effects for patient safety and perioperative outcomes. Multiple studies have investigated the impact of resident and fellow participation, versus attending only involvement, on perioperative outcomes, with the results largely demonstrating that resident and fellow participation is safe and does not adversely affect clinical outcomes.7–14 Data on the differential impact of residents versus fellows on perioperative outcomes is less robust; however, in minimally invasive, breast and endocrine surgery, trainee level, specifically senior resident versus fellow, did not translate to adverse outcomes.9,13–15
In complex surgical oncology, an analysis of data from the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) assessing the impact of resident participation demonstrated that resident participation was associated with significantly higher 30-day postoperative complications, but lower 30-day mortality and improved rescue rate after complications.16 While these findings may relate to differences in the complexity and acuity of patients at academic and non-academic institutions, the data expose a potential detriment of resident participation in complex surgical oncology. However, the differential impact of resident versus fellow participation on outcomes of complex surgical oncology is not known. Utilizing the ACS-NSQIP database, this study sought to investigate the impact of resident versus fellow participation on outcomes following complex surgical oncology cases.
Materials and Methods
Study Design
This is a retrospective, nationally validated cohort study, derived from the ACS NSQIP database, of patients undergoing complex surgical oncology operations, including esophagectomy, gastrectomy, hepatectomy, or pancreatectomy between 2007-2012 with assistance from a senior resident or fellow. Inclusion criteria was based on procedural code, as outlined below. Using propensity matching, patients were matched based on the odds of undergoing surgery assisted by a surgical fellow, rather than senior resident, based on age, sex, BMI, ASA class, diabetes mellitus, and current smoking status. Surgical outcomes, principally the occurrence of a major postoperative complication, were compared between the two groups. Approval for this study was obtained through an agreement with ACS-NSQIP. The study was also reviewed and deemed exempt by Institutional Review Board at our institution (H-42952).
Dataset
Data was derived from the ACS-NSQIP database, which is a nationally validated, outcomes-based program designed to measure and improve the quality of surgical care in the United States (US). As of 2022, roughly 700 hospitals participate in ACS-NSQIP, and contribute hospital-level data detailing all aspects of surgical care and related patient outcomes for all inpatient and outpatient surgical procedures. At each participating institution, clinically trained personnel abstract data from medical records, including preoperative risk factors, intraoperative variables and 30-day morbidity and mortality outcomes.
Patient selection
Current Procedural Terminology (CPT) codes were used to identify patients who underwent complex surgical oncologic operations, involving the esophagus, liver, stomach and/or pancreas. With regard to the esophagus, eligible operations included total esophagectomy (CPT 43107, 43108, 43112, 43113) or partial esophagectomy (CPT 43116, 43117, 43118, 43121, 43122, 43123, 43124). Operations involving the liver included partial or total hepatectomy (CPT 47120, 47122, 47125, 47130). Gastric operations included total gastrectomy (CPT 43620, 43621, 43622) and partial gastrectomy (CPT 43631, 43632, 43633, 43634, 43635). Pancreatic resections included distal pancreatectomy (CPT 48140, 48146) and proximal pancreatectomy (CPT 48150, 48152, 48153, 48154).
Among patients undergoing one of the above eligible operations, all non-emergent cases performed with the participation of a general surgery senior resident (PGY-4-5) or fellow (PGY-6-8) were included. Resident participation was defined using the “PGY” variable in the ACS-NSQIP Participant Use File, which defines the current clinical post-graduate year of the participating resident. The study was restricted to patients in the ACS-NSQIP database between 2007 and 2012, as the NSQIP program discontinued collecting data on resident involvement after 2012. Patients who underwent emergent operations, as defined within the ACS NSQIP database, were excluded (n=1,778), as well as patients over 89 or with any of the following variables missing: resident involvement, age, height, weight, sex, diabetes, American Society of Anesthesiologist (ASA) class, diabetes, bleeding disorder or operative time (n=11,659).
Variable definition
Patient demographics included age, sex, body mass index (BMI), ASA class, vascular comorbidities, cardiac comorbidities, diabetes mellitus, current smoking status, bleeding disorders, immunosuppressive therapy, malnourishment, receipt of neoadjuvant chemotherapy, and emergency operation. Vascular comorbidities included history of hypertension, amputation or revascularization procedure, rest pain or gangrene. Cardiac comorbidities included congestive heart failure within 30 days, myocardial infarction within 6 months or history of angina within 1 month of surgery, and prior percutaneous coronary intervention or cardiac surgery, as previously described.17 Malnourishment was defined as greater than 10% loss of body weight in the 6-month period preceding the index operation.
Study outcomes
Consistent with established NSQIP outcomes, the principal outcome of this study was the occurrence of a major postoperative complication, defined as deep incisional surgical site infections, organ space surgical site infection, pneumonia, pulmonary embolism, progressive renal insufficiency, acute renal failure, stroke or cardiovascular accident, myocardial infarction, sepsis or septic shock, cardiac arrest requiring cardiopulmonary resuscitation, wound disruption, unplanned intubation, coma >24 hours, ventilator requirement >48 hours, any unplanned reoperation, or mortality within 30 days from operation. Secondary outcomes included minor complications, operative time, reoperation and 30-day mortality. Minor complications included superficial surgical site infections, urinary tract infections, and deep venous thrombosis or thrombophlebitis, as outlined by the ACS NSQIP database. Additional details regarding the variables and outcome definitions are available via the ACS NSQIP Participant Use File.18
Statistical analysis
Categorical variables are presented as frequencies, and continuous variables are presented as median and interquartile range (IQR). Univariable analyses were conducted using Chi-square test for categorical variables and either student T-tests or Kruskal-Wallis tests for continuous variables. To mediate the potential impact of case selection and biased trainee participation in more or less complex operations, propensity scores were calculated for the odds of undergoing a complex surgical oncologic operation assisted by a surgical fellow, rather than senior resident, based on age, sex, BMI, ASA class, diabetes mellitus, and current smoking status. Patients were matched 1:1 based on their propensity scores. Outcomes were compared after matching using student T-test or Kruskal-Wallis test. Statistical analysis was performed using SAS, version 9.4 (SAS Institute). P-values <0.05 were considered significant.
Results
Considering cases involving the participation of a senior resident or surgical fellow, a total of 15,897 eligible patients were identified during the study period, including 1169 patients who underwent esophagectomy, 2470 who underwent partial or total gastrectomy, 4700 who underwent a hepatectomy, and 7558 who underwent pancreatomy (Table 1). Senior residents participated in 60.4% of esophagectomies, 86.8% of gastrectomies, 75.2% of hepatectomies, and 86.0% of pancreatectomies, with the remainder of operations being assisted by a surgical fellow.
Table 1:
Patients Undergoing Eligible Operations Involving a Senior Resident or Surgical Fellow
| Procedure | Total Eligible Patients | Senior Resident (PGY 4-5) Participation |
Surgical Fellow (PGY 6-8) Participation |
|---|---|---|---|
| Esophagectomy | 1169 | 706 (60.4) | 463 (39.6) |
| Gastrectomy | 2470 | 2143 (86.8) | 327 (13.2) |
| Hepatectomy | 4700 | 3535 (75.2) | 1165 (24.8) |
| Pancreatectomy | 7558 | 6503 (86.0) | 1055 (14.0) |
| Complex Surgical Oncology | 15897 | 12887 | 3010 |
Unmatched Results
Baseline demographics and outcomes before matching are presented in Table 2a (esophagectomy and gastrectomy) and Table 2b (hepatectomy and pancreatectomy). Prior to propensity matching, overall rates of major complications were equivalent for all four anatomic locations between cases performed with assistance of a fellow or a senior resident. Median operative time was shorter with resident participation for gastrectomy (208 vs 232 min, p<0.001) and pancreatectomy (318 vs 330 min, p=0.02). For hepatectomy, prior to matching, the 30-day mortality rate (1.9% vs 3.0%, p=0.03) and the rate of minor complications (9.2% vs 11.3%, p=0.03) was lower for cases involving a resident. Similarly, for pancreatectomy cases, reoperation rate (5.6% vs 7.6%, p<0.009) and rate of minor complications (14.5% vs 18.5%, p<0.001) was lower for cases involving a senior resident.
Table 2a:
Baseline characteristics and surgical outcomes for esophagectomy and gastrectomy by resident and fellow participation before matching
| Esophagectomy (n=1169) | Gastrectomy (n=2470) | |||||
|---|---|---|---|---|---|---|
| Resident participation (N=706) | Fellow participation (N=463) | p-value | Resident participation (N=2143) | Fellow participation (N=327) | p-value | |
| Baseline characteristics | ||||||
| Age in years, median (IQR) | 63 (57-71) | 63 (56-70) | 0.40 | 64 (53-73) | 62 (49-71) | 0.002 |
| Female sex, n (%) | 137 (19.4) | 98 (21.2) | 0.46 | 1099 (51.3) | 176 (53.8) | 0.39 |
| White race, n (%) | 643 (91.1) | 439 (94.8) | 0.02 | 1527 (71.3) | 261 (79.8) | 0.001 |
| BMI >30, n (%) | 271 (38.4) | 175 (37.8) | 0.84 | 688 (32.1) | 112 (34.3) | 0.44 |
| ASA III-V, n (%) | 63 (8.9) | 42 (9.1) | 0.93 | 144 (6.7) | 11 (3.4) | 0.02 |
| Diabetes mellitus, n (%) | 84 (11.9) | 56 (12.1) | 0.92 | 227 (10.6) | 46 (14.1) | 0.06 |
| Vascular comorbidities, n (%) | 379 (53.7) | 251 (54.2) | 0.86 | 1145 (53.4) | 158 (48.3) | 0.08 |
| Cardiac comorbidities, n (%) | 98 (13.9) | 70 (15.1) | 0.56 | 262 (12.2) | 36 (11.0) | 0.53 |
| Pulmonary comorbidities, n (%) | 115 (16.3) | 64 (13.8) | 0.25 | 202 (14.1) | 50 (15.3) | 0.58 |
| Current smoker, n (%) | 193 (27.3) | 103 (22.3) | 0.05 | 435 (20.3) | 67 (20.5) | 0.94 |
| Malnourishment, n (%) | 126 (17.9) | 85 (18.4) | 0.82 | 311 (14.5) | 41 (12.5) | 0.34 |
| Neoadjuvant chemotherapy, n (%) | 89 (12.6) | 117 (25.3) | <0.001 | 141 (6.6) | 32 (9.8) | 0.03 |
| Surgical outcomes | ||||||
| Mortality within 30 days, n (%) | 25 (3.5) | 11 (2.4) | 0.26 | 63 (2.9) | 10 (3.1) | 0.91 |
| Minor complications, n (%) | 124 (17.6) | 70 (15.1) | 0.27 | 235 (11.0) | 47 (14.4) | 0.07 |
| Major complications, n (%) | 261 (37.0) | 150 (32.4) | 0.11 | 481 (22.5) | 73 (22.3) | 0.96 |
| Surgical site infections, n (%) | 136 (19.3) | 70 (15.1) | 0.07 | 321 (15.0) | 61 (18.7) | 0.09 |
| Operating time, median (IQR) | 328.5 (249-422) | 338 (263-436) | 0.14 | 208 (153-276) | 232 (182-301) | <0.001 |
| Reoperation, n (%) | 96 (13.6) | 50 (10.8) | 0.16 | 170 (7.9) | 25 (7.7) | 0.86 |
Table 2b:
Baseline characteristics and surgical outcomes for hepatectomy and pancreatectomy by resident and fellow participation before matching
| Hepatectomy (n=4700) | Pancreatectomy (n=7558) | |||||
|---|---|---|---|---|---|---|
| Resident participation (N=3535) | Fellow participation (N=1165) | p-value | Resident participation (N=6503) | Fellow participation (N=1055) | p-value | |
| Baseline characteristics | ||||||
| Age in years, median (IQR) | 60 (51-68) | 59 (49-68) | 0.20 | 64 (54-72) | 63 (54-72) | 0.16 |
| Female sex, n (%) | 1846 (52.2) | 587 (50.4) | 0.28 | 3334 (51.3) | 530 (50.2) | 0.53 |
| White race, n (%) | 651 (18.4) | 188 (16.1) | 0.08 | 5560 (85.5) | 911 (86.4) | 0.46 |
| BMI >30, n (%) | 1378 (39.0) | 456 (39.1) | 0.92 | 2203 (33.9) | 364 (34.5) | 0.69 |
| ASA III-V, n (%) | 162 (4.6) | 55 (4.7) | 0.85 | 307 (4.7) | 36 (3.4) | 0.06 |
| Diabetes mellitus, n (%) | 382 (10.8) | 130 (11.2) | 0.74 | 1099 (16.9) | 171 (16.2) | 0.58 |
| Vascular comorbidities, n (%) | 1719 (48.6) | 559 (48.0) | 0.70 | 3540 (54.4) | 527 (50.0) | 0.007 |
| Cardiac comorbidities, n (%) | 263 (7.4) | 114 (9.8) | 0.01 | 680 (10.5) | 121 (11.5) | 0.32 |
| Pulmonary comorbidities, n (%) | 349 (9.9) | 101 (8.7) | 0.23 | 761 (11.7) | 104 (9.9) | 0.08 |
| Current smoker, n (%) | 584 (16.5) | 201 917.3) | 0.56 | 1399 (21.5) | 219 (20.8) | 0.58 |
| Malnourishment, n (%) | 145 (4.1) | 51 (4.4) | 0.68 | 998 (15.4) | 155 (14.7) | 0.58 |
| Neoadjuvant chemotherapy, n (%) | 383 (10.8) | 154 (13.2) | 0.03 | 220 (3.4) | 50 (4.7) | 0.03 |
| Surgical outcomes | ||||||
| Mortality within 30 days, n (%) | 68 (1.9) | 35 (3.0) | 0.03 | 115 (1.8) | 19 (1.8) | 0.94 |
| Minor complications, n (%) | 324 (9.2) | 132 (11.3) | 0.03 | 941 (14.5) | 195 (18.5) | <0.001 |
| Major complications, n (%) | 625 (17.7) | 186 (16.0) | 0.18 | 1530 (23.5) | 266 (25.2) | 0.23 |
| Surgical site infections, n (%) | 403 (11.4) | 155 (13.3) | 0.08 | 1311 (20.2) | 273 (25.9) | <0.001 |
| Operating time, median (IQR) | 220 (157-303) | 219 (157-293) | 0.46 | 318 (230-413) | 330 (244-422) | 0.02 |
| Reoperation, n (%) | 131 (3.7) | 42 (3.6) | 0.87 | 361 (5.6) | 80 (7.6) | 0.009 |
Propensity Matched Analysis
After propensity matching, overall rates of major complications were equivalent across all four anatomic locations between cases performed with the participation of a senior resident versus a surgical fellow: esophagectomy (37.0%% vs 31.6%, p=0.10), gastrectomy (22.6% vs 22.3%, p=0.93), hepatectomy (15.8% v 16.0%, p=0.91), and pancreatectomy (23.9% vs 25.2%, p=0.48) (Table 3a and Table 3b). For all anatomic locations except hepatectomy, mortality rates were comparable among cases performed with assistance of a fellow or a senior resident. Mortality rates (1.7% vs 3.0%, p=0.04) and minor complication rates (8.4% vs 11.3%, p=0.02) after hepatectomy were higher for cases with fellow participation. Reoperation rates were higher for esophagectomies involving resident participation (14.9% vs 10.5%, p=0.05) and pancreatectomies involving fellow participation (5.4% vs 7.6%, p=0.04). Reoperation rates were comparable for complex gastric and hepatic cases involving resident or fellow participation. Median operative time for gastrectomy was shorter involving resident participation (212 vs 232 min, p=0.004).
Table 3a:
Baseline characteristics and surgical outcomes for esophagectomy and gastrectomy by resident and fellow participation after matching
| Esophagectomy (n=860) | Gastrectomy (n=654) | |||||
|---|---|---|---|---|---|---|
| Resident participation (N=430) | Fellow participation (N=430) | p-value | Resident participation (N=327) | Fellow participation (N=327) | p-value | |
| Baseline characteristics | ||||||
| Age in years, median (IQR) | 63 (57-71) | 63 (56-71) | 0.73 | 60 (50-72) | 62 (49-71) | 0.85 |
| Female sex, n (%) | 87 (20.2) | 94 (21.9) | 0.56 | 179 (54.7) | 176 (53.8) | 0.81 |
| White race, n (%) | 396 (92.1) | 406 (94.4) | 0.17 | 241 (73.7) | 261 (79.8) | 0.06 |
| BMI >30, n (%) | 163 (37.9) | 164 (38.1) | 0.94 | 102 (31.2) | 112 (34.3) | 0.40 |
| ASA III-V, n (%) | 32 (7.4) | 37 (8.6) | 0.53 | 8 (2.5) | 11 (3.4) | 0.48 |
| Diabetes mellitus, n (%) | 57 (13.3) | 53 (12.3) | 0.68 | 30 (9.2) | 46 (14.1) | 0.05 |
| Vascular comorbidities, n (%) | 231 (53.7) | 236 (54.8) | 0.73 | 148 (45.3) | 158 (48.3) | 0.43 |
| Cardiac comorbidities, n (%) | 63 (14.7) | 66 (15.4) | 0.77 | 34 (10.4) | 36 (11.0) | 0.80 |
| Pulmonary comorbidities, n (%) | 71 (16.5) | 63 (14.7) | 0.45 | 37 (11.3) | 50 (15.3) | 0.13 |
| Current smoker, n (%) | 103 (24.0) | 102 (23.7) | 0.94 | 75 (22.9) | 67 (20.5) | 0.45 |
| Malnourishment, n (%) | 76 (17.7) | 79 (18.4) | 0.79 | 55 (16.8) | 41 (12.5) | 0.12 |
| Neoadjuvant chemotherapy, n (%) | 85 (19.8) | 84 (19.5) | 0.93 | 30 (9.2) | 32 (9.8) | 0.79 |
| Surgical outcomes | ||||||
| Mortality within 30 days, n (%) | 15 (3.5) | 10 (2.3) | 0.31 | 11 (3.4) | 10 (3.1) | 0.82 |
| Minor complications, n (%) | 75 (17.4) | 64 (14.9) | 0.31 | 44 (13.5) | 47 (14.4) | 0.73 |
| Major complications, n (%) | 159 (37.0) | 136 (31.6) | 0.10 | 74 (22.6) | 73 (22.3) | 0.93 |
| Surgical site infections, n (%) | 77 (17.9) | 66 (15.4) | 0.32 | 49 (15.0) | 61 (18.7) | 0.21 |
| Operating time, median (IQR) | 329.5 (247-422) | 336 (263-434) | 0.41 | 212 (153-277) | 232 (182-301) | 0.004 |
| Reoperation, n (%) | 64 (14.9) | 45 (10.5) | 0.05 | 27 (8.3) | 25 (7.7) | 0.77 |
Table 3b:
Baseline characteristics and surgical outcomes for hepatectomy and pancreatectomy by resident and fellow participation after matching
| Hepatectomy (n=2330) | Pancreatectomy (n=2110) | |||||
|---|---|---|---|---|---|---|
| Resident participation (N=1165) | Fellow participation (N=1165) | p-value | Resident participation (N=1055) | Fellow participation (N=1055) | p-value | |
| Baseline characteristics | ||||||
| Age in years, median (IQR) | 59 (50-68) | 59 (49-68) | 0.82 | 64 (54-72) | 63 (54-72) | 0.63 |
| Female sex, n (%) | 596 (51.2) | 587 (50.4) | 0.71 | 518 (49.1) | 530 (50.2) | 0.60 |
| White race, n (%) | 959 (82.3) | 977 (83.9) | 0.32 | 149 (14.1) | 144 (13.7) | 0.75 |
| BMI >30, n (%) | 476 (40.9) | 456 (39.1) | 0.40 | 390 (37.0) | 364 (34.5) | 0.24 |
| ASA III-V, n (%) | 53 (4.6) | 55 (4.7) | 0.84 | 36 (3.4) | 36 (3.4) | 1.00 |
| Diabetes mellitus, n (%) | 126 (10.8) | 130 (11.2) | 0.79 | 174 (16.5) | 171 (16.2) | 0.86 |
| Vascular comorbidities, n (%) | 536 (46.0) | 559 (48.0) | 0.34 | 565 (53.6) | 527 (50.0) | 0.10 |
| Cardiac comorbidities, n (%) | 92 (7.9) | 114 (9.8) | 0.11 | 120 (11.4) | 121 (11.5) | 0.95 |
| Pulmonary comorbidities, n (%) | 113 (9.7) | 101 (8.7) | 0.39 | 117 (11.1) | 104 (9.9) | 0.36 |
| Current smoker, n (%) | 217 (18.6) | 201 (17.3) | 0.39 | 228 (21.6) | 219 (20.8) | 0.63 |
| Malnourishment, n (%) | 34 (2.9) | 51 (4.4) | 0.06 | 165 (15.6) | 155 (14.7) | 0.54 |
| Neoadjuvant chemotherapy, n (%) | 133 (11.4) | 154 (13.2) | 0.19 | 28 (2.7) | 50 (4.7) | 0.01 |
| Surgical outcomes | ||||||
| Mortality within 30 days, n (%) | 20 (1.7) | 35 (3.0) | 0.04 | 24 (2.3) | 19 (1.8) | 0.44 |
| Minor complications, n (%) | 98 (8.4) | 132 (11.3) | 0.02 | 156 (14.8) | 195 (18.5) | 0.02 |
| Major complications, n (%) | 184 (15.8) | 186 (16.0) | 0.91 | 252 (23.9) | 266 (25.2) | 0.48 |
| Surgical site infections, n (%) | 127 (10.9) | 155 (13.3) | 0.08 | 224 (21.2) | 273 (25.9) | 0.01 |
| Operating time, median (IQR) | 217 (157-299) | 219 (157-293) | 0.85 | 320 (235-417) | 330 (244-422) | 0.43 |
| Reoperation, n (%) | 46 (4.0) | 42 (3.6) | 0.66 | 57 (5.4) | 80 (7.6) | 0.04 |
Discussion
In this study, we utilized the ACS-NSQIP database from 2007 to 2012 to compare perioperative outcomes of complex surgical oncologic operations involving senior residents and surgical fellows. Overall, we observed that senior resident participation in complex cancer operations, specifically esophagectomy, gastrectomy, hepatectomy and pancreatectomy, does not appear to negatively impact operative time or perioperative outcomes, compared to the participation of a fellow. Regardless of whether a senior resident or surgical fellow participated in the case, the overall rate of major complications was equivalent across all four anatomic locations. For several operation types, secondary outcomes, including rate of minor complications, reoperation rates, operative time and mortality rate, favored resident participation, compared to fellow participation. While these secondary findings may seem counter-intuitive, they may be explained by differences in the extent of trainee autonomy and/or participation. Nonetheless, the current study suggests that relative to fellows, resident involvement in complex surgical oncology is safe. Our results demonstrate that senior residents and surgical fellows can equally safely participate in complex cancer operations, including esophagectomy, gastrectomy, hepatectomy and pancreatectomy. These findings are largely consistent with prior studies.9,13–15 In particular, several large studies examining outcomes from the ACS-NSQIP database across other surgical subspecialties have identified equivalent patient outcomes in cases involving either senior residents or fellows. Our group previously described how surgical outcomes following common minimally invasive cases, including laparoscopic Roux-en-Y gastric bypass, laparoscopic cholecystectomy and laparoscopic hiatal hernia repair, were similar between fellow and chief resident assistance.15 Resident participation in partial nephrectomies, either open or laparoscopic, was associated with lower major complication rates, compared to fellows.7 With regard to thyroid and parathyroid surgery, operations involving senior residents were associated with no statistically significant difference in the rate of neurologic or bleeding complications when compared to operations involving fellow participation.13 These findings and our data provide robust evidence that resident involvement does not adversely impact outcomes relative to fellow participation.
Over the past 20 years, notable changes in surgical training have occurred;19,20 in particular, there has been a substantial decrease in operative trauma and a major shift from open abdominal and vascular operations to minimally invasive and endovascular surgeries, respectively.19–23 As a result, the value of complex surgical oncology in the operative training of general surgery residents has increased. While minimally invasive approaches are increasingly common in complex surgical oncology, the role of complex surgical oncology in teaching safe and effective open thoracic and abdominal surgical technique is essential for trainees. The surgical technique and training widely extends to other surgical subspecialties, which generally offer less opportunities in these domains, but equally require proficiency when alternatives to open surgery are not indicated or appropriate. Accordingly, it is critical that surgical residents continue to meaningfully participate in complex surgical oncology operations, and the findings of the current study support their participation relative to fellows.
Beyond this overarching finding, some additional results warrant consideration. First of all, we observed that median operative time was shorter by 20 minutes for gastric resections involving residents, compared to fellows. While this difference is relatively modest, increased operative duration is associated with negative short-term outcomes and increased hospital costs.24–27 This finding likely relates to differences in the nature of resident and fellow participation in cases; however, data on trainee autonomy and intraoperative activities is limited. One study by Fieber et al. found that perceived trainee autonomy was associated with improved patient outcomes following colorectal surgery.28 Further research should aim to more directly evaluate the nature of trainee autonomy and intraoperative tasks in order to inform surgical training and understand the differential impact of surgical trainees of varying experience. Second, we found that certain secondary outcomes, including reoperation rates following esophagectomy, as well as mortality and minor complication rates after hepatectomy, were lower for cases involving residents compared to surgical fellows. This was not expected given the nature of surgical fellowship, both in terms of greater domain operative experience and overall training experience. While our analysis was matched using propensity scores, this likely reflects differences in the acuity and complexity of cases in which residents and fellows participate, and their autonomy and degree of operative involvement. Recent data from Simpson et al. evaluating the effect of assistant’s training level on outcomes following hepatectomy found that relative to surgical residents, fellows performed higher complexity cases with longer operative time; however, outcomes were similar regardless of assistant training level.29
Despite the current findings that senior residents do not adversely affect patient outcomes relative to surgical fellows, it is important to consider prior data demonstrating that trainee participation in complex oncologic surgery is associated with significantly higher rates of 30-day postoperative outcomes.16 While our data is not consistent with several other studies examining the impact of resident participation versus attending-only involvement on outcomes in other surgical specialties, our study does not compare outcomes of operations involving trainees versus attending-only. Complex cancer operations are distinctly sophisticated, and additional research is needed to assess the impact of trainee participation in these operations in order to guide the safe involvement of trainees while also ensuring optimal patient outcomes. Ultimately, we recommend that resident involvement in complex cancer operations should be graduated, closely supervised and modified when appropriate.
Importantly, this study does not evaluate the important question of how co-occurring residency and fellowship programs influence their respective training and clinical experience. Several groups have studied this relationship, highlighting the uncertainty in effect. A 2012 systematic review concluded that the overall impact of advanced surgical fellowships on resident education and training remains unclear, reporting that 13/23 (56.5%) studies found mixed results, 9/23 (39.1%) found a negative impact and only 1/23 (4.3%) found a generally positive effect.30 A second study found that the implementation of a pediatric surgery fellowship resulted in a decrease in general surgery resident index and overall case volume in pediatric surgery, despite a concurrent increase in the overall number of attending surgeons and relative value units.31 Conversely, it has been shown that implementation of an Acute Care Surgery fellowship did not compromise operative experience, and even had a positive impact on operative and non-operative clinical experiences.32 Despite these studies, data regarding the impact of co-occurring residency and fellowship programs is limited and further research is needed to inform surgical education.
We acknowledge that there are several limitations to our study. Most notably, our analysis is limited by the nature of the ACS-NSQIP database and the fact that the database only captures a trainee’s presence in an operation, and does not describe or quantify the extent of their involvement. Despite this limitation, the current findings demonstrate that senior resident participation in complex surgical oncology – under the guidance and direction of an attending surgeon – is safe, and not associated with worse clinical outcomes, which is an important contribution to our understanding graduate surgical education. Future research investigating the impact of experience level of the attending surgeon on the surgical training and operative experience of trainees may be insightful. In addition, this study utilized a non-randomized, large database, which likely suffers selection bias. However, we attempted to minimize this limitation by performing propensity score matching based on the likelihood that an operation would be performed with the participation of a surgical fellow. With that said, we are unable to control for other institutional and/or local factors, and are not directly comparing residents and fellows in a single setting. Furthermore, the data may not fully reflect the present era, as NSQIP stopped collecting data on trainee involvement at the end of 2012. While there have been marked changes in healthcare over the past decade, any major changes to intra-operative practices have likely affected surgical residents and fellows equivalently. For this reason, we suspect our findings reasonably describe relevant outcomes, as they relate to current surgical practice and training. However, the current study does not investigate more recent changes to surgical education and training, In particular, surgical simulation, in the form of traditional approaches or newer technologies such as the use of augmented reality has become an increasing educational tool for surgical training.33–36 Simulation affords new opportunities for hands-on operative training and development, outside the realm of actual patient care. There is evidence that stimulation improves surgical outcomes and the safety of involving trainees in operative care.33,35 Surgical simulation may alters the narrative regarding the optimal role of trainees in direct operative care, and further investigation regarding its impact is warranted. Lastly, the ACS-NSQIP database only includes 30-day outcomes. For cancer operations, these outcomes are inherently lacking certain important endpoints. Despite these limitations, the NSQIP database is a validated and risk-adjusted database that confers less risk than accruing data across single-institutions.
In conclusion, we found that senior resident participation in complex cancer operations does not negatively impact operative time or perioperative outcomes, compared to the participation of a fellow. Regardless of whether a senior resident or surgical fellow participated in the case, the overall rate of major complications was equivalent across all four anatomic locations. Our results suggest that relative to fellows, resident involvement in complex surgical oncology remains safe and appropriate. Future research, including qualitive approaches, is needed at co-occurring residency and fellowship programs to investigate the intra-operative experiences and outcomes for surgical residents and fellows.
Highlights.
We compared surgical outcomes among propensity-matched cases involving residents or fellows
Overall rates of complications were comparable for cases involving senior residents or fellows
Senior resident participation in complex cancer surgery does not have a negative impact on care
Further research is needed to assess the impact of co-occurring residency and CSO fellowship programs on the training of residents and fellows – and their respective involvement in operations
Funding:
Susanna W.L. de Geus is supported by the Perlman scholarship for pancreatic cancer research. Alison Woods is supported by a T32 grant through Johns Hopkins University School of Medicine, from the National Institutes of Health (National Cancer Institute award #T32CA126607). Marianna Papageorge is supported by a T32 grant through Boston University School of Medicine (award #T32HP10028).
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.American Board of Surgery. Speciality of Complex General Surgical Oncology.
- 2.Lee AY, Delman KA, Berman RS. Development of a surgical oncology training curriculum for accreditation. J Surg Oncol. 2020;122(1):15–20. doi: 10.1002/JSO.25933 [DOI] [PubMed] [Google Scholar]
- 3.Society of Surgical Oncology. History of the Society of Surgical Oncology. [Google Scholar]
- 4.Balch CM, Coit DG, Berman RS. 2015 James Ewing Lecture: 75-Year History of the Society of Surgical Oncology-Part III: The Transformative Years (1991-2015). Ann Surg Oncol. 2016;23(5):1409–1417. doi: 10.1245/S10434-015-5011-4 [DOI] [PubMed] [Google Scholar]
- 5.Society of Surgical Oncology. Society of Surgical Oncology: Fellowship Program List.
- 6.Sachs TE, Ejaz A, Weiss M, et al. Assessing the experience in complex hepatopancreatobiliary surgery among graduating chief residents: Is the operative experience enough? Surgery. 2014;156(2):385. doi: 10.1016/J.SURG.2014.03.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kern SQ, Lustik MB, Mcmann LP, Thibault GP, Sterbis JR. Comparison of outcomes after minimally invasive versus open partial nephrectomy with respect to trainee involvement utilizing the American College of Surgeons National Surgical Quality Improvement Program. J Endourol. 2014;28(1):40–47. doi: 10.1089/END.2013.0051 [DOI] [PubMed] [Google Scholar]
- 8.Srour MK, Manguso N, Mirocha J, Chung A, Giuliano AE, Amersi F. Impact of Resident and Fellow Participation on Surgical Outcomes in Breast Conserving Surgery for Invasive Breast Cancer. J Surg Educ. 2020;77(1):144–149. doi: 10.1016/J.JSURG.2019.07.014 [DOI] [PubMed] [Google Scholar]
- 9.Venkat R, Valdivia PL, Guerrero MA. Resident participation and postoperative outcomes in adrenal surgery. J Surg Res. 2014;190(2):559–564. doi: 10.1016/J.JSS.2014.05.025 [DOI] [PubMed] [Google Scholar]
- 10.Itani KMF, DePalma RG, Schifftner T, et al. Surgical resident supervision in the operating room and outcomes of care in Veterans Affairs hospitals. Am J Surg. 2005;190(5):725–731. doi: 10.1016/J.AMJSURG.2005.06.042 [DOI] [PubMed] [Google Scholar]
- 11.Fischer CP, Hong JC. Early perioperative outcomes and pancreaticoduodenectomy in a general surgery residency training program. J Gastrointest Surg. 2006;10(4):478–482. doi: 10.1016/J.GASSUR.2006.01.010 [DOI] [PubMed] [Google Scholar]
- 12.Kshirsagar RS, Chandy Z, Mahboubi H, Verma SP. Does resident involvement in thyroid surgery lead to increased postoperative complications? Laryngoscope. 2017;127(5):1242–1246. doi: 10.1002/LARY.26176 [DOI] [PubMed] [Google Scholar]
- 13.Feeney T, Price LL, Chen L, Graham R, Chatterjee A. Resident and fellow participation in Thyroid and Parathyroid surgery: an ACS-NSQIP clinical outcomes analysis. J Surg Res. 2017;220:346–352. doi: 10.1016/J.JSS.2017.07.030 [DOI] [PubMed] [Google Scholar]
- 14.Chatterjee A, Pyfer B, Chen L, Czerniecki B, Tchou J, Fisher C.Resident and Fellow Participation in Breast Surgery: An American College of Surgeons NSQIP Clinical Outcomes Analysis. J Am Coll Surg. 2015;221(5):988–994. doi: 10.1016/J.JAMCOLLSURG.2015.08.008 [DOI] [PubMed] [Google Scholar]
- 15.Aly S, de Geus SWL, Carter CO, et al. Impact of fellow compared to resident assistance on outcomes of minimally invasive surgery. Surg Endosc. 2022;36(2):1554–1562. doi: 10.1007/S00464-021-08444-8 [DOI] [PubMed] [Google Scholar]
- 16.Castleberry AW, Clary BM, Migaly J, et al. Resident education in the era of patient safety: a nationwide analysis of outcomes and complications in resident-assisted oncologic surgery. Ann Surg Oncol. 2013;20(12):3715–3724. doi: 10.1245/S10434-013-3079-2 [DOI] [PubMed] [Google Scholar]
- 17.Sippey M, Spaniolas K, Manwaring ML, Pofahl WE, Kasten KR. Surgical resident involvement differentially affects patient outcomes in laparoscopic and open colectomy for malignancy. Am J Surg. 2016;211(6):1026–1034. doi: 10.1016/J.AMJSURG.2015.07.019 [DOI] [PubMed] [Google Scholar]
- 18.American College of Surgeons. ACS NSQIP Participant Use Data File.
- 19.Drake FT, Horvath KD, Goldin AB, Gow KW. The general surgery chief resident operative experience: 23 years of national ACGME case logs. JAMA Surg. 2013;148(9):841–847. doi: 10.1001/JAMASURG.2013.2919 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Cortez AR, Katsaros GD, Dhar VK, et al. Narrowing of the surgical resident operative experience: A 27-year analysis of national ACGME case logs. Surgery. 2018;164(3):577–582. doi: 10.1016/J.SURG.2018.04.037 [DOI] [PubMed] [Google Scholar]
- 21.Richards MK, McAteer JP, Drake FT, Goldin AB, Khandelwal S, Gow KW. A national review of the frequency of minimally invasive surgery among general surgery residents: assessment of ACGME case logs during 2 decades of general surgery resident training. JAMA Surg. 2015;150(2):169–172. doi: 10.1001/JAMASURG.2014.1791 [DOI] [PubMed] [Google Scholar]
- 22.Schanzer A, Steppacher R, Eslami M, Arous E, Messina L, Belkin M. Vascular surgery training trends from 2001–2007: A substantial increase in total procedure volume is driven by escalating endovascular procedure volume and stable open procedure volume. J Vasc Surg. 2009;49(5):1339–1344. doi: 10.1016/J.JVS.2008.12.019 [DOI] [PubMed] [Google Scholar]
- 23.Fan EY, Crawford AS, Judelson DR, et al. Trends in General Surgery Operative Experience Obtained by Integrated Vascular Surgery Residents. J Surg Educ. 2021;78(6):2127–2137. doi: 10.1016/J.JSURG.2021.05.010 [DOI] [PubMed] [Google Scholar]
- 24.Cheng H, Chen BPH, Soleas IM, Ferko NC, Cameron CG, Hinoul P. Prolonged Operative Duration Increases Risk of Surgical Site Infections: A Systematic Review. Surg Infect (Larchmt). 2017;18(6):722–735. doi: 10.1089/SUR.2017.089 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Held MB, Boddapati V, Sarpong NO, Cooper HJ, Shah RP, Geller JA. Operative Duration and Short-Term Postoperative Complications after Unicompartmental Knee Arthroplasty. J Arthroplasty. 2021;36(3):905–909. doi: 10.1016/J.ARTH.2020.09.007 [DOI] [PubMed] [Google Scholar]
- 26.Philip S, Jackson N, Mittal V. Outcomes after laparoscopic or robotic colectomy and open colectomy when compared by operative duration for the procedure. Am J Surg. 2018;215(4):577–580. doi: 10.1016/J.AMJSURG.2017.04.020 [DOI] [PubMed] [Google Scholar]
- 27.Guidolin K, Spence RT, Azin A, et al. The effect of operative duration on the outcome of colon cancer procedures. Surg Endosc. 2022;36(7):5076–5083. doi: 10.1007/S00464-021-08871-7 [DOI] [PubMed] [Google Scholar]
- 28.Fieber JH, Bailey EA, Wirtalla C, et al. Does Perceived Resident Operative Autonomy Impact Patient Outcomes? J Surg Educ. 2019;76(6):e182–e188. doi: 10.1016/J.JSURG.2019.06.006 [DOI] [PubMed] [Google Scholar]
- 29.Simpson RE, Carpenter KL, Wang CY, et al. Is resident assistance equivalent to fellows during hepatectomy? Surg Endosc. 2021;35(1):260–269. doi: 10.1007/S00464-020-07388-9 [DOI] [PubMed] [Google Scholar]
- 30.Plerhoples TA, Greco RS, Krummel TM, Melcher ML. Symbiotic or parasitic? A review of the literature on the impact of fellowships on surgical residents. Ann Surg. 2012;256(6):904–908. doi: 10.1097/SLA.0B013E318262EDD5 [DOI] [PubMed] [Google Scholar]
- 31.Snyder RA, Phillips SE, Terhune KP. Impact of implementation of a pediatric surgery fellowship on general surgery resident operative volume. J Surg Educ. 2012;69(6):753–758. doi: 10.1016/J.JSURG.2012.03.007 [DOI] [PubMed] [Google Scholar]
- 32.Dinan KA, Davis JW, Wolfe MM, Sue LP, Cagle KM. An acute care surgery fellowship benefits a general surgical residency. J Trauma Acute Care Surg. 2014;77(2):209–212. doi: 10.1097/TA.0000000000000293 [DOI] [PubMed] [Google Scholar]
- 33.Meling TR, Meling TR. The impact of surgical simulation on patient outcomes: a systematic review and meta-analysis. Neurosurg Rev. 2021;44(2):843–854. doi: 10.1007/S10143-020-01314-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Zevin B, Aggarwal R, Grantcharov TP. Surgical simulation in 2013: why is it still not the standard in surgical training? J Am Coll Surg. 2014;218(2):294–301. doi: 10.1016/J.JAMCOLLSURG.2013.09.016 [DOI] [PubMed] [Google Scholar]
- 35.Trehan K, Kemp CD, Yang SC. Simulation in cardiothoracic surgical training: where do we stand? J Thorac Cardiovasc Surg. 2014;147(1):18–24.e2. doi: 10.1016/J.JTCVS.2013.09.007 [DOI] [PubMed] [Google Scholar]
- 36.Balla A, Sartori A, Botteri E, et al. Augmented reality (AR) in minimally invasive surgery (MIS) training: where are we now in Italy? The Italian Society of Endoscopic Surgery (SICE) ARMIS survey. Updates Surg. 2023;75(1):85–93. doi: 10.1007/S13304-022-01383-6 [DOI] [PubMed] [Google Scholar]