Materials and Methods
Patients who underwent hysterectomy for benign disease from January 2010 through December 2012 and who were recorded in the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) database were included in this study. Primary Current Procedural Terminology (CPT) codes were used to identify patients who underwent hysterectomy and to classify patients by route of surgery ( Supplemental Table 1 ). CPT codes that include hysterectomy along with other procedures, such as lymphadenectomy, which would indicate the hysterectomy was being performed as treatment for a malignancy, were not included. Patients also were excluded if an International Classification of Diseases , 9th Revision code for malignancy was recorded.
The American College of Surgeons NSQIP database is a national surgical quality improvement program. Participation is voluntary, and participating institutions are able to track their own risk-adjusted outcomes after surgery. Trained clinical reviewers prospectively collect variables such as patient demographics, operative variables, and postoperative outcomes for 30 days after surgery. Periodic auditing ensures high-quality data, including for data points that occur after hospital discharge. Details of methods of data collection and reliability have been reported previously.
Demographic variables abstracted included age, race, and body mass index. Patient-related preoperative variables abstracted included hypertension requiring medication, diabetes mellitus requiring insulin or oral therapy, smoking in the last year, American Society of Anesthesiologists (ASA) score, and major medical comorbidity divided into the categories of neurologic, cardiac, pulmonary, and infectious. Major cardiac comorbidity was defined as congestive heart failure in the month before surgery, myocardial infarction 6 months before surgery, history of peripheral vascular disease, history of percutaneous cardiac intervention, or previous cardiac surgery. Major pulmonary comorbidity was defined as history of severe chronic obstructive pulmonary disease or pneumonia in the last 30 days. Major neurologic comorbidity was defined as history of stroke with or without neurologic deficit or history of transient ischemic attack. Major infectious morbidity was defined as preoperative sepsis, or a preoperative open wound. Definitions of these patient history variables, such as the criteria for history of myocardial infarction or stroke, were per the NSQIP data participant use file.
Operative variables abstracted included operative time, surgical approach, trainee involvement, and surgical complexity. Trainee involvement was defined as trainee intraoperative involvement as coded in the NSQIP database. Trainees were further classified as either residents or fellows on the basis of postgraduate year. Residents were defined as postgraduate year 1−4 and fellows were defined as postgraduate year 5 and greater. Surgical complexity was defined by the work relative value unit (wRVU), which is an estimate of the amount of physician work per CPT code defined by Medicare. The wRVU for each procedure is the sum of the assigned value to each CPT code for the procedure; thus, greater wRVU is associated with increased surgical complexity.
Perioperative surgical complications were defined as complications that occurred from the start of surgery up to 30 days postoperatively. Major complications were defined as grade 3 or greater on the validated Clavien-Dindo grading scale. Minor complications were defined as grade 2 or less. Major complications included myocardial infarction, pneumonia, venous thromboembolism, deep or organ space surgical-site infection, stroke, fascial dehiscence, unplanned return to the operating room, renal failure, cardiopulmonary arrest, sepsis, intubation greater than 48 hours, and death. Minor complications included urinary tract infection, blood transfusion, and superficial wound infection. Intraoperative complications such as accidental puncture or laceration are not recorded in the NSQIP database and thus were not included in the definition of perioperative complication. Specific definitions of the diagnostic criteria for each of these complications can be found in the NSQIP data participant use file.
This was a secondary analysis cohort study of prospectively collected surgical-quality data. The primary outcome was major perioperative complication and secondary outcome was minor perioperative complication. Patients with intraoperative trainee participation were compared with those without with respect to outcomes. For bivariable analysis, 2-tailed t-tests were used for continuous variables and Pearson χ 2 tests for categorical variables. Stratified analysis was performed by surgical approach, given the association between surgical approach and both trainee involvement and complications. Associations between trainee participation and complications were analyzed by the use of binary logistic regression to examine for potential confounding. Confounders were selected on the basis of known associations with complications in the literature. A P value of less than .05 was considered significant for all analyses. SPSS version 20.0 (IBM Corp, Armonk, NY) was used for all analyses. The Institutional Review Board of University of North Carolina at Chapel Hill declared this study exempt from formal review because it does not constitute human subjects research.
Results
We identified 22,499 patients who underwent hysterectomy for benign disease. Demographic and operative characteristics of the overall study population are provided in Table 1 . Patients with intraoperative trainee involvement represent a population at greater risk of complication with older age, greater surgical complexity scores based on work relative value units, and greater rates of comorbidities (hypertension, pulmonary and neurologic comorbidities, and ASA score ≥3).
| Patient characteristics | Overall n = 22,499 | No trainee participation n = 13,028 | Trainee participation n = 9471 | P value |
|---|---|---|---|---|
| Age, y | 47.5 ± 10.8 | 46.5 ± 10.8 | 48.59 ± 10.7 | <.0001 |
| BMI, kg/m 2 | 30.1 ± 7.5 | 30.1 ± 7.3 | 30.1 ± 7.7 | NS |
| Race | <.0001 | |||
| White | 13,579 (60%) | 8583 (66%) | 4996 (53%) | |
| Hispanic | 2505 (11%) | 2003 (15%) | 502 (5.3%) | |
| Black | 2636 (12%) | 924 (7.1%) | 1712 (18%) | |
| Asian | 371 (1.7%) | 160 (1.2%) | 211 (2.2%) | |
| Other | 3408 (15%) | 1358 (10%) | 2050 (22%) | |
| Hypertension | 5740 (26%) | 3053 (23%) | 2687 (28%) | <.0001 |
| Current smoker | 4334 (19%) | 2619 (20%) | 1715 (18%) | <.0001 |
| Diabetes | NS | |||
| Noninsulin | 1144 (5.1%) | 629 (4.8%) | 515 (5.4%) | |
| Insulin | 357 (1.6%) | 205 (1.6%) | 152 (1.6%) | |
| Cardiac comorbidity | 295 (1.3%) | 163 (1.3%) | 132 (1.4%) | NS |
| Pulmonary comorbidity | 1024 (4.6%) | 659 (5.1%) | 365 (3.9%) | <.0001 |
| Neurologic comorbidity | 347 (1.5%) | 172 (1.3%) | 175 (1.9%) | <.002 |
| Infectious comorbidity | 156 (0.7%) | 81 (0.6%) | 75 (0.8%) | NS |
| ASA score ≥3 | 3825 (17%) | 1892 (15%) | 1933 (20%) | <.0001 |
| Surgical complexity (wRVU) | 24.3 ± 12.1 | 23.7 ± 11.9 | 24.9 ± 12.1 | <.0001 |
| Operative time, minutes | 122.5 ± 63.7 | 108.3 ± 58.1 | 141.8 ± 65.9 | <.0001 |
Overall, 42.1% (n = 9471) of patients had intraoperative trainee involvement in their hysterectomy. Of the 9471 hysterectomies with trainee participation, 8375 (88.4%) had data available for postgraduate year of the highest-level trainee involved. This was a fellow in 14.5% (n = 1375) and a resident in 73.9% (n = 7000). Surgical approach was 22.7% vaginal (n = 5112), 47.1% abdominal (n = 6803), and 30.2% laparoscopic (n = 10,584). Rates of trainee involvement differed depending on surgical approach with trainee involvement in 45.1% (n = 4272) of laparoscopic hysterectomies, 32.1% (n = 3038) of open hysterectomies, and 22.8% (n = 2161) of vaginal hysterectomies ( P < .001).
The overall rate of major perioperative complication was 3.2% (n = 726), and the rate of minor perioperative complication was 7.2% (n = 1616). Rates of complications also differed depending on surgical approach with more complications occurring in the open hysterectomy group. Rates of major perioperative complication were 4.0% for open hysterectomy (n = 270), 3.0% for laparoscopic hysterectomy (n = 316), and 2.7% for vaginal hysterectomy (n = 140). Rates of minor perioperative complication were 11.4% for open hysterectomy (n = 775), 6.2% for vaginal hysterectomy (n = 317), and 5.0% for laparoscopic hysterectomy (n = 524).
Given the observed differences in intraoperative trainee involvement and major perioperative complications depending on surgical approach, a stratified analysis was performed by surgical approach. In bivariable analysis, trainee involvement was associated with major perioperative complication among patients undergoing vaginal hysterectomy, but not abdominal or laparoscopic hysterectomy ( Table 2 ). Trainee involvement also was associated with increased rates of minor complications among patients undergoing vaginal, laparoscopic, and open hysterectomies.
| No major complication | Major complication | P value | No minor complication | Minor complication | P value | |
|---|---|---|---|---|---|---|
| Vaginal | ||||||
| No trainee | 2,883 | 68 (2.3%) | .03 a | 2791 | 160 (5.4%) | .007 a |
| Trainee | 2,089 | 72 (3.3%) | 2004 | 157 (7.3%) | ||
| Laparoscopic | ||||||
| No trainee | 6,126 | 186 (2.9%) | .78 | 6038 | 274 (4.3%) | <.001 a |
| Trainee | 4,142 | 130 (3.0%) | 4022 | 250 (5.9%) | ||
| Open | ||||||
| No trainee | 3,630 | 135 (3.6%) | .07 | 3417 | 348 (9.2%) | <.001 a |
| Trainee | 2,903 | 135 (4.4%) | 2611 | 427 (14.1%) |
We then compared the association between trainee level and perioperative complication. We found that there was no difference for fellows versus residents in rates of major complications (4.0% vs 3.5%, P = .33) or minor complications (9.2% vs 9.0%, P = .89). The specific perioperative complications experienced by patients with and without trainee involvement are listed in Table 3 .
| No trainee (n = 13,028) | Trainee (n = 9471) | P value | |
|---|---|---|---|
| Major complication | |||
| MI | 5 (0.04%) | 3 (0.03%) | NS |
| PNA | 26 (0.2%) | 13 (0.1%) | NS |
| VTE | 42 (0.3%) | 41 (0.4%) | NS |
| Deep/organ space SSI | 125 (0.2%) | 109 (0.2%) | NS |
| Fascial dehiscence | 36 (0.3%) | 27 (0.3%) | NS |
| Return to OR | 193 (1.5%) | 149 (1.6%) | NS |
| Stroke | 2 (0.02%) | 0 (0.0%) | NS |
| Renal failure | 8 (0.06%) | 5 (0.05%) | NS |
| Cardiopulmonary arrest | 1 (0.008%) | 1 (0.01%) | NS |
| Sepsis | 52 (0.4%) | 61 (0.6%) | .02 a |
| Prolonged/reintubation | 9 (0.1%) | 19 (0.1%) | NS |
| Nerve injury | 7 (0.05%) | 6 (0.06%) | NS |
| Minor complication | |||
| Superficial wound infection | 164 (1.3%) | 160 (1.7%) | .007 a |
| Transfusion | 329 (2.5%) | 404 (4.3%) | <.001 a |
| Urinary tract infection | 313 (2.4%) | 320 (3.4%) | <.001 a |
To address potential confounders, logistic regression was performed in which we adjusted for age, body mass index, hypertension, smoking, cardiac comorbidity, pulmonary comorbidity, neurologic comorbidity, infectious comorbidity, ASA score, and surgical complexity ( Table 4 ). After we adjusted for these confounders, trainee involvement remained associated with major perioperative complications among patients undergoing vaginal hysterectomy. Trainee involvement also remained associated with minor perioperative complications among all 3 surgical approaches.
| Unadjusted OR | 95% CI | Adjusted OR | 95% CI | |
|---|---|---|---|---|
| Major complication | ||||
| Vaginal | 1.46 a | 1.04−2.04 a | 1.4 a | 1.03−2.04 a |
| Laparoscopic | 1.03 | 0.82−1.30 | 1.09 | 0.86−1.37 |
| Open | 1.25 | 0.98−1.60 | 1.19 | 0.92−1.53 |
| Minor complication | ||||
| Vaginal | 1.37 a | 1.09−1.72 a | 1.30 a | 1.03−1.63 a |
| Laparoscopic | 1.37 a | 1.15−1.63 a | 1.36 a | 1.14−1.63 a |
| Open | 1.61 a | 1.38−1.87 a | 1.56 a | 1.33−1.82 a |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
