A-D, Data are shown as means (solid lines) with 95% CIs (shaded areas) for female and male trainees.
eTable 1. Case Number by Gender, Clinical Rotation Number, and Procedure Type
eTable 2. Unadjusted Mean Number of Cases/Trainee/Rotation
eTable 3. Statistical Significance for Factors Associated With Operative Autonomy on Bivariate Analysis
eTable 4. Statistical Significance (P values) for Factors Associated With Operative Autonomy on Multivariate Analysis (Type III ANOVA), Including Interactions Between Factors
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Joh DB, van der Werf B, Watson BJ, et al. Assessment of Autonomy in Operative Procedures Among Female and Male New Zealand General Surgery Trainees. JAMA Surg. 2020;155(11):1019–1026. doi:10.1001/jamasurg.2020.3021
Does trainee sex play a role in assessing the level of procedural autonomy a trainee receives during the New Zealand general surgery training program?
In this 5-year cohort study that included 119 380 general surgery procedures performed by 120 trainees, female trainees performed fewer endoscopic, major, and minor procedures autonomously than their male counterparts.
Results of this study suggest that female trainees are more likely to receive less procedural autonomy throughout their general surgery training program in New Zealand.
The need for trainee sex equality within surgical training has resulted in an appraisal of the training experience in the New Zealand general surgery training program.
To investigate the association between trainee sex and surgical autonomy in the operating room in the New Zealand general surgery training program.
Design, Setting, and Participants
Retrospective cohort study conducted from December 10, 2012, to December 10, 2017, examining all endoscopic, major, and minor procedures performed by all New Zealand general surgery trainees in every training hospital in New Zealand.
Main Outcomes and Measures
The primary outcome was the level of meaningful autonomy by each New Zealand general surgery trainee (ie, trainee as primary operator without the surgeon mentor scrubbed for the case). Outcomes were compared using multivariable analysis.
This study included 120 New Zealand general surgery trainees (42 women [35%] and 78 men [65%]) who were analyzed over 279.5 trainee-years (88.5 trainee-years for women and 191.0 trainee-years for men). Included were 119 380 general surgery procedures (17 465 endoscopic, 56 964 major, and 44 951 minor) in 18 hospitals. By the end of the 5-year training program, female trainees had a lower cumulative mean autonomous caseload than male trainees for endoscopic (284.0 [95% CI, 207.0-361.0] vs 352.2 [95% CI, 282.9-421.6], P = .03), major (139.9 [95% CI, 76.7-203.2] vs 198.1 [95% CI, 142.3-254.0], P = .02), and minor (456.3 [95% CI, 394.8-517.9] vs 519.9 [95% CI, 465.6-574.2], P = .007) procedures.
Conclusions and Relevance
After accounting for differences among trainees, hospital type, number of female and male surgeon mentors at each hospital, and trainee seniority, female trainees performed fewer cases with meaningful autonomy compared with male trainees. These findings support the need for pragmatic solutions to address this bias and further investigations on mechanisms contributing to discrepancies.
Sex inequality in the workplace remains prevalent in most societies, resulting in female workers being less likely to receive career advancement compared with their male colleagues.1 In addition, men are often given more authority for the same level of occupational status, likely because of conscious or unconscious bias among decision makers rather than because of the behavior of individual women.2 Sex inequality and discrimination have also been noted among surgeons and surgical trainees, with surveys documenting high rates of sex discrimination experienced by female surgeons at every level of training.3-5
An important aspect of surgical training is for trainees to be given appropriate autonomy in operative cases.6 This emphasis is because trainees who are able to act autonomously learn better and may achieve superior performance through the development of a range of technical and nontechnical skills.7 Ideally, surgical training programs should be designed for trainees to gradually gain more responsibility and autonomy in preparation for independent practice.8
With growing concerns that surgical programs are inadequately training surgeons, it is important that trainees be given adequate operative autonomy across the duration of their training.9 There is some evidence that male trainees receive more surgical autonomy or privileges during their training than their female colleagues.10-12 However, these studies are limited by small sample size or reliance on subjective evaluations of autonomy, and a more objective and larger-scale study is needed in this area. Therefore, the aim of this study was to investigate the association between trainee sex and surgical autonomy in the operating room in the New Zealand general surgery training program. We hypothesized that female surgical trainees receive less operative autonomy during their training than male trainees.
This retrospective longitudinal cohort study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline and was approved by the University of Auckland Human Participants Ethics Committee. The committee waived the requirement for informed consent because deidentified data were used. Included were all endoscopic, major, and minor procedures performed by all New Zealand general surgery trainees in every training hospital in New Zealand from December 10, 2012, to December 10, 2017.
This study included all operative procedures performed by general surgery trainees across all 18 training hospitals in New Zealand. Procedural data were retrieved through the Surgical Online Logbook and Assessments tool,13 which the Royal Australasian College of Surgeons (RACS) mandates that New Zealand general surgery trainees use to record all procedures. Operative procedures were classified as endoscopic, major, or minor according to standardized RACS criteria. Operative autonomy was classified by the trainee for each procedure in accordance with RACS guidelines using the following 5 categories: assisting senior resident, assisting surgeon mentor, primary operator with surgeon mentor scrubbed, primary operator with surgeon mentor in the operating room, and primary operator with surgeon mentor available. In this context, the term mentor is equivalent to attending surgeon and does not imply a formal mentorship relationship. There are no standard criteria used to assess the level of autonomy entrusted to trainees. The Surgical Online Logbook and Assessments tool data were recorded prospectively by the individual trainees after each procedure and were approved by their training supervisor. Data were deidentified regarding trainee and patient details, and trainees were assigned a unique study ID before transfer of the data to the study team.
Sex was recorded as that reported by the trainee at the time of registration with RACS. The duration of the General Surgery Surgical Education and Training (SET) program was 5 years at the start of the study period (SET 1-5) and was changed to 4 years in 2016 (SET 2-5). Trainees complete 6-month rotations with each surgical team; therefore, training comprised 10 clinical rotations in a 5-year program or 8 clinical rotations in a 4-year program. Because this investigation is a longitudinal study, individual trainees in the study contributed between 1 and 5 years of data, and some trainees took time out of the training program for 6 months or longer during the study. To account for this variation, data were considered in terms of trainee-year (ie, a trainee completing two 6-month full-time clinical rotations). When a trainee repeated a SET year, data from the first and last rotations of that SET year (regardless of how many times the year was repeated) were included in the analysis.
Training hospitals were categorized as major or peripheral centers based on the New Zealand Association of General Surgeons classification, whereby trainees in their final 2 years of training are placed in a major hospital in preparation for their fellowship examination. Major hospitals are all tertiary referral centers with a wide range of specialties present within the hospital. Trainees are only placed in training hospitals and do not perform procedures in non–training hospitals. The number and sex of trainers (ie, surgeon mentors) at each training hospital during each year of the study period were obtained from each hospital.
The primary outcome was the level of meaningful autonomy by each New Zealand general surgery trainee. Quiz Ref IDMeaningful autonomy was defined as the trainee performing the case as primary operator with the surgeon mentor in the operating room or available but not scrubbed for the case. This definition was used because it demonstrates a high level of trust in the trainee by the surgeon mentor. Outcomes were compared using multivariable analysis.
Continuous data are reported with 95% CIs. Procedure case numbers were used to calculate the unadjusted mean differences and relative differences in case numbers between female and male trainees, where a positive mean difference and a relative difference greater than 1 represented a higher number of cases performed by female trainees. Univariate and multivariable analyses were performed on the number of procedures per trainee, procedure type (endoscopy, major, or minor), autonomy level (autonomous or nonautonomous), and SET year. Linear mixed regression models were used in all analyses with trainee, hospital type, and calendar year included as random factors. Univariate analysis was performed by procedure type and autonomy level. To evaluate the same model for all combinations of procedure type and autonomy level, the best multivariable model was obtained using all the data while evaluating only terms with interaction between procedure type and autonomy level. All linear and quadratic models were evaluated to obtain the best-fitting multivariable model (ie, that with the lowest value for the Akaike information criterion14). All analyses were conducted using the function lmer from the lme4 package15 within R, version 3.5.1 (R Foundation for Statistical Computing).16 Cumulative values and values for totals, including 95% CIs, were calculated using the final regression model parameters with their covariances. Testing was 2-sided, with the threshold for statistical significance set at P = .05.
This study included 120 New Zealand general surgery trainees (42 women [35%] and 78 men [65%]) who were analyzed over 279.5 trainee-years (88.5 trainee-years for women and 191.0 trainee-years for men). Trainees contributed data for a median of 4 (interquartile range, 1-9) 6-month rotations. There were no missing values because all values within a rotation were available for each trainee. Four trainees interrupted their training for personal reasons, taking 6 to 12 months off before returning to the program. No differences were observed between female and male trainees with regard to SET year, rotation number, or hospital type (Table 1). The absolute number of cases performed by trainee sex and rotation number are listed in eTable 1 in the Supplement.
A total of 119 380 general surgery procedures (17 465 endoscopic, 56 964 major, and 44 951 minor) performed in 18 hospitals were included in the final analysis, of which 36 612 (30.7%) were performed by female trainees and 82 768 (69.3%) by male trainees. The unadjusted mean absolute differences in autonomous cases per trainee per year are listed in eTable 2 in the Supplement.
Bivariate analysis to investigate the statistical significance of trainee sex, SET year, rotation number, hospital type, number of female surgeon mentors, number of male surgeon mentors, and total number of surgeon mentors is summarized in eTable 3 in the Supplement. There were no missing data. Multivariable analysis was conducted controlling for these factors, including interaction terms between factors. The statistical significance of each factor is listed in eTable 4 in the Supplement. The mixed multivariable regression model is summarized in Table 2. Growth in cumulative autonomous caseload for female and male trainees is shown in the Figure. In this model, there were no statistically significant differences between female and male trainees in cumulative mean total caseload performed with any level of autonomy at the end of the 5-year training program (2102.2 vs 2238.0, P = .08) (Table 3). Quiz Ref IDHowever, regarding autonomous cases of any type, female trainees had a lower cumulative mean caseload than male trainees. This difference held true for endoscopic (284.0 [95% CI, 207.0-361.0] vs 352.2 [95% CI, 282.9-421.6], P = .03), major (139.9 [95% CI, 76.7-203.2] vs 198.1 [95% CI, 142.3-254.0], P = .02), and minor (456.3 [95% CI, 394.8-517.9] vs 519.9 [95% CI, 465.6-574.2], P = .007) procedures.Quiz Ref ID There was no difference in nonautonomous cumulative caseload between female and male trainees. After accounting for differences among trainees, hospital type, number of female and male surgeon mentors at each hospital, and trainee seniority, female trainees performed fewer cases with meaningful autonomy compared with male trainees. Compared with their male counterparts, female trainees had odds ratios of performing procedures with meaningful autonomy across the 5-year training program of 0.79 (95% CI, 0.70-0.89; P < .001) for performing any procedure, 0.77 (95% CI, 0.56-1.05; P = .12) for performing an endoscopy, 0.69 (95% CI, 0.54-0.87; P = .002) for performing a major procedure, and 0.80 (95% CI 0.65-0.99; P = .04) for performing a minor procedure.
This retrospective longitudinal cohort study of 119 380 general surgery procedures performed by 120 trainees over 279.5 trainee-years sought to investigate the association between trainee sex and operative autonomy. Female and male trainees performed equivalent numbers of procedures across the duration of their training (approximately 2000 cases). However, female trainees performed fewer autonomous endoscopic, major, and minor cases than their male colleagues, and this difference remained statistically significant after controlling for the individual variability among trainees, hospital type, number of female and male surgeon mentors at each hospital, and trainee seniority. These data demonstrate that female general surgery trainees in New Zealand performed a higher proportion of procedures without meaningful autonomy compared with their male colleagues.
In another study17 evaluating the association of a resident’s sex with faculty entrustment in the operating room, trainee sex was not found to be a statistically significant factor. However, the study was limited by a small caseload and the measurement tool (OpTrust), which required observation and assessment of each operative case by a third party. An earlier study18 of the feasibility of OpTrust found an increase in resident autonomy with observed cases; therefore, the results of the study by Thompson-Burdine et al17 may be subject to the Hawthorne effect. Furthermore, a 2019 study12 also observed that female residents received less autonomy than male residents, despite controlling for resident’s level of training, case complexity, faculty sex, and training program. In addition, a 2017 study10 revealed a bias against female residents based on resident assessments by both thoracic surgery trainees and faculty.
Although few studies10,12,17,18 have investigated the interaction between trainee sex and operative autonomy, numerous studies have investigated the sex equity of surgical training programs. A study19 in otolaryngology found that male residents spent on average more time in the operating room than female residents; however, this finding was in the context of men working longer hours than women. In a survey study20 of surgical residents, 91% of men believed that their operative experience was comparable to that of women, whereas only 75% of women (P < .001) thought that their experience was comparable to that of men. In addition, only 68% of women felt that they were evaluated by the same criteria as men during training, whereas 81% of men felt that women were evaluated by similar criteria as men (P < .001). Furthermore, more female residents and attending surgeons than male residents agreed with the statement that “in my residency program female residents are treated in an inferior manner compared with male residents.”
Despite efforts to counter sex discrimination in the workplace and during surgical training, the issue remains challenging because less obvious or covert forms of discrimination persist.21,22 Such discrimination is more difficult to identify because it is ambiguous and often cloaked by nondiscriminatory or rational reasons. Clinical judgment by a supervising surgeon to provide or withhold autonomy to a female trainee may be attributed to implicit bias. Studies23,24 on resident autonomy indicate that a resident’s observed clinical skill and level of training are the main factors that constitute a supervising surgeon’s decision to provide autonomy. If the provision of resident autonomy was based purely on this observation, then the autonomy given to female trainees should be at least equal to that of male trainees. To rectify discrepancies in the residency training experience between female and male residents, surgical programs must develop an objective method of evaluating each trainee’s entrustability.25 There is evidence that training supervisor behavior may be the primary contributor to whether a trainee is given autonomy, and this comportment may be amenable to change with appropriate education.26
Our study did not seek to identify the underlying causes of the difference in autonomy seen between female and male trainees. Mechanisms underlying the provision of resident autonomy have been addressed elsewhere,23 with major factors being a resident’s observed clinical skill, the attending surgeon’s level of confidence, and an increased focus on patient outcomes. Although trainee sex was not specifically addressed in that study, perceived resident skill levels may have been altered by trainee sex. Sex discrimination exists not only in residency training but persists from medical school to after training.27 Female trainees may also perceive their own skills differently than male trainees, such that female trainees may underestimate their abilities, despite having objective surgical skills that are comparable to those of their male counterparts.28,29 Whether the inequality in surgical autonomy stems from differences in trainees, supervising surgeons, or a combination of both remains to be investigated.
Our study has some limitations. This study relied on self-assessment of autonomy within the constraints of RACS predefined supervision categories and a requirement for sign off of cases by the training supervisor. Reporting error in case supervision may have been systematically different between female and male trainees (ie, female trainees could have underreported their autonomy, or male trainees could have overreported their autonomy). Furthermore, assessment of autonomy can differ between trainees and supervisors,30 and there was no external validation of autonomy at the time of case classification in our study. Case complexity is known to alter autonomy, and this factor was not included in our model beyond classifying procedures as endoscopic, major, or minor. However, the classification we used is defined by RACS in the Surgical Online Logbook and Assessments tool and cannot be changed, thereby removing any interrater variation. Quiz Ref IDIn addition, it is unclear whether our results stem from trainee factors or supervising surgeon factors. We were also unable to assess the implications of race/ethnicity and age of trainees and the consequences of race/ethnicity, age, and sex of the supervising surgeons on autonomy because this information was not included in our deidentified data. The data are from a single surgical specialty training program in New Zealand; therefore, the generalizability of the findings to other settings is unclear. However, the message of the results is applicable to all training programs: assess whether trainee sex bias exists and, if present, to address differences in surgical autonomy.
This study analyzed 119 380 general surgery procedures performed by 120 New Zealand trainees over 279.5 trainee-years to investigate the association between trainee sex and procedural autonomy. Quiz Ref IDAfter accounting for differences among trainees, hospital type, number of female and male surgeon mentors at each hospital, and trainee seniority, female trainees performed fewer cases with meaningful autonomy compared with male trainees. Our finding that trainee sex bias extends to operative autonomy in surgical training supports the need for action to address this discrimination. Further investigations on mechanisms contributing to these discrepancies are needed to assess whether interventions can alter training supervisor behavior.
Accepted for Publication: May 4, 2020.
Corresponding Author: Benjamin P. T. Loveday, MBChB, PhD, Department of Surgery, The University of Auckland, Auckland, New Zealand 1023 (email@example.com).
Published Online: August 26, 2020. doi:10.1001/jamasurg.2020.3021
Author Contributions: Dr Loveday had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Watson, Loveday.
Acquisition, analysis, or interpretation of data: Joh, van der Werf, French, Bann, Dennett, Loveday.
Drafting of the manuscript: Joh, van der Werf, Dennett, Loveday.
Critical revision of the manuscript for important intellectual content: van der Werf, Watson, French, Bann, Dennett, Loveday.
Statistical analysis: van der Werf, Loveday.
Administrative, technical, or material support: Dennett, Loveday.
Supervision: Bann, Loveday.
Conflict of Interest Disclosures: None reported.