Bucholz EM, Sue GR, Yeo H, Roman SA, Bell RH, Sosa JA. Our Trainees’ ConfidenceResults From a National Survey of 4136 US General Surgery Residents. Arch Surg. 2011;146(8):907-914. doi:10.1001/archsurg.2011.178
Author Affiliations: Department of Surgery, Yale University School of Medicine, New Haven, Connecticut (Mss Bucholz and Sue and Drs Yeo, Roman, and Sosa); and American Board of Surgery, Inc, Philadelphia, Pennsylvania (Dr Bell).
Objectives To characterize factors shaping surgery resident confidence and determine whether confidence is associated with future specialty training.
Design Cross-sectional study.
Setting Survey administered at the 2008 American Board of Surgery In-Service Training Examination.
Participants All categorical general surgery residents.
Interventions National Study of Expectations and Attitudes of Residents in Surgery survey. Participants reported demographics and level of agreement for 46 items regarding confidence, training, and professional plans.
Main Outcomes Measures Survey items “My operating skill level is appropriate” and “I may not feel confident enough to perform procedures independently before training completion.” We compared demographics and responses among residents who did/not feel confident.
Results Response rate was 77.4%. Residents who were female, single, or without children and at a lower postgraduate year had less confidence in their operating skill, as did residents at larger, university-based programs, in the northeastern United States. Residents worried about competence were more likely to believe specialty training was needed (71.2% vs 60.2%; P < .001). After adjustment, residents dissatisfied with training were less likely to believe their skills were level appropriate (odds ratio, 0.13; P < .001) as were residents not comfortable asking for help (odds ratio, 0.48; P < .001). After adjustment, women were twice more likely than men to worry about competence after training; single residents were 1.36 times more likely than married residents to believe their skills were not level appropriate. Program location, type, and size remained associated with confidence, as did satisfaction and comfort asking for help. Residents worried about skills were more likely to plan for fellowship.
Conclusions Sex, marital status, children, and postgraduate year are predictors of confidence, as are program location, type, and size. Residency programs may target modifiable factors contributing to low surgical confidence.
Self-confidence is an attitude that allows individuals to have a positive and realistic perception of themselves and their abilities. This has implications for job satisfaction and performance.1,2 As the transitional period bridging medical school and independent practice, residency training is critical for the development of physician confidence. A developed sense of confidence plays a potential role in a resident's career satisfaction and the decisions to change specialties or professions during residency and to pursue additional specialty training.
General surgery residency consists of 5 years of clinical training. Current literature suggests that confidence increases significantly during the internship year,3 in principle because interns have greater clinical responsibility for patient care compared with senior medical students.4,5 Confidence is important in surgical residency, because it encompasses several personal attributes that, if channeled appropriately, may lead to professional satisfaction and success. These include enthusiasm, assertiveness, independence, trust, the ability to handle criticism, and emotional maturity. The trajectory of developing surgical confidence during the span of surgical residency training is not well understood, because there is a paucity of research reporting the confidence levels of general surgery residents across all postgraduate training years.
The goal of the current study was to characterize the resident and residency program factors that influence confidence among surgical residents. We hypothesized that confidence level affects the decisions residents make regarding their training, such as attrition and further specialization.
To examine factors associated with resident confidence, we used data from the National Study of Expectations and Attitudes of Residents in Surgery survey, a cross-sectional survey of all US categorical general surgery residents that was administered in January 2008. Survey administration procedures and baseline data have been previously published.6 In brief, surveys were administered at the time of the American Board of Surgery In-Service Training Examination to maximize resident participation. All residency programs except 1 (n = 248) participated in the survey.
The survey consisted of 51 questions that inquired about residents' motivations for pursuing surgery, experience in the program, and intentions after training. Questions were structured as statements with a 5-point Likert response scale, ranging from strongly agree to strongly disagree. Questions about resident demographics (age, race/ethnicity, residency level, marital status, and number of children) also were included in the survey. All responses were confidential. Information on residency program characteristics (number of chief residents, type of residency, geographic region, and affiliation with postresidency training fellowships) was collected using the Accreditation Council for Graduate Medical Education online database. Data on resident attrition were obtained from the American Board of Surgery for the academic year of 2007-2008.
In this study, we examined demographic and programmatic factors related to resident surgical confidence and the association between surgery training experience and confidence levels. To characterize resident confidence, we used the following 2 survey items: “I feel my operating skill level is appropriate” and “I worry I will not feel confident enough to perform procedures by myself before I finish training.” These 2 items measured trust in one's skills and surgical independence. Surgery training experience was assessed using questions related to resident satisfaction with program training. Comfort asking their attending surgeons for help was used to measure ability to handle criticism and emotional maturity. For the purposes of these analyses, we excluded residents in their research years because they were remote from clinical activities.
Descriptive statistics were used to characterize the overall study sample with respect to individual and program-level factors. For ease of interpretation, survey responses to questions on confidence and surgery training experiences were divided into agree (strongly agree and agree) and disagree (neutral, disagree, and strongly disagree) categories. To assess differences in individual and program-level characteristics by surgical confidence category, we compared residents who agreed with the survey questions of interest with those who disagreed using the χ2 or the Fisher exact test for categorical variables and the unpaired, 2-tailed t test for continuous variables. Using χ2 tests, we compared residents with low vs high surgical confidence with respect to resident satisfaction and comfort asking their attending surgeons for help. Multivariate analyses used hierarchical logistic regression models and commercially available software (GLIMMIX procedure in SAS, version 9.2; SAS Institute, Cary, North Carolina). All demographic and program-related variables were included in the models to evaluate which variables were the strongest predictors of surgical confidence. This study was approved by the Yale University institutional review board and human investigation committee.
Of the 5345 US categorical general surgery residents eligible for our study, 4136 completed the survey, yielding a 77.4% response rate (Table 1). The mean age of respondents was 30.5 years; 31.2% were female; 51.2% were married; and 25.0% had children. Most of the respondents were white (62.3%), with Asians representing the next largest racial group (18.0%).
Most respondents were in programs in the Northeast (Table 2). University-based programs had the greatest percentage of residents (67.4%). Most residents were in training programs affiliated with subspecialty fellowship programs (60.5%). The mean number of chief residents per residency program was 4.
Nearly three-quarters (71.5%) of all categorical general surgery residents in postgraduate clinical training years believed that their operating skill was level appropriate but 26.1% worried that they would not feel confident enough to perform procedures by themselves before they finish training.
Age, sex, race, postgraduate year (PGY), marital status, and family size were all associated with the view that operating skill was level appropriate (Table 3 and Table 4). Older residents and residents further along in their training were more likely to believe that their operating skill was level appropriate, as were male residents compared with their female counterparts (75.4% vs 70.5%; P = .001). White and Hispanic residents were the most likely to believe that their surgical skills were level appropriate, whereas black residents were the least likely to believe it (76.4% and 77.2%, respectively, vs 58.1%; P < .001 for both comparisons). Married residents were more likely than single residents to believe that their surgical skills were level appropriate (78.6% vs 65.8%; P < .001), as were residents with children compared with residents without children (81.1% vs 71.5%; P = .001).
Several of the same demographic characteristics that were associated with resident views on operating skill also were associated with concern about the ability to perform procedures independently before finishing training. Men were less likely than women to have this concern (21.9% vs 37.0%; P < .001); married residents, less likely than single ones (23.4% vs 30.9%; P < .001); and residents with children, less likely than those without children (22.7% vs 28.3%; P = .006). Male residents were more likely to be married and have children than female residents (58.1% vs 37.4% and 31.1% vs 11.7%, respectively; P < .001 for both). A greater percentage of PGY1 and PGY5 residents were confident compared with PGY2, PGY3, and PGY4 residents.
Several program factors were associated with residents believing their operating skills were level appropriate (Table 4). Residents in programs located in the West, Midwest, and South were more likely to believe that their operating skill was level appropriate compared with residents at programs in the Northeast (75.6%, 75.7%, and 74.5%, respectively, vs 70.6%; P < .05 for all regions compared with the Northeast). Residents at community programs were more likely than their counterparts at university-based or military programs to feel confident in their skills (77.6% vs 72.2% and 68.9%, respectively; P < .001 for both). The larger the residency program as measured by the number of chief residents, the less likely its residents were to feel adequately skilled for their level of training.
Trainees at university-based and military programs and those with affiliated specialty fellowships felt less confident about being able to operate independently by the end of residency than their respective counterparts at community programs (29.6% and 29.3%, respectively, vs 20.4%; P < .01). Those without fellowships felt less confident than those with fellowships (22.9% vs 29.1%; P < .001). Fewer residents training at programs with 3 to 4, 5 to 6, or 7 or more chief residents felt confident compared with those training at programs with 1 to 2 chief residents (23.3%, 29.3%, and 31.3%, respectively, vs 18.3%; P < .05 for all comparisons).
Multivariate analyses were performed to examine the independent associations between responders' demographic and programmatic factors and resident confidence levels (Table 5 and Table 6). After adjustment, being single, black, or Asian remained independently associated with being less confident that operating skills were level appropriate than being white. In contrast, being further along in training was associated with higher confidence. Residents at training programs outside of the Northeast and those at community hospitals also were significantly more likely to be confident in their operating skills.
Many of the same resident and programmatic characteristics were associated with resident worry over performing procedures independently by the end of training. After adjustment, female and PGY2 through PGY4 residents were significantly more likely to worry about their surgical abilities at the end of training. Residents at community hospitals and of Hispanic origin were less likely to worry about their surgical abilities.
Associations also were examined between responses to other survey items and ratings of confidence levels (Tables 5 and 6). Residents who believed that their attending surgeons would think worse of them if they asked for help when performing procedures (odds ratio, 0.49; 95% confidence interval, 0.39-0.60; P < .001) or managing a patient (odds ratio, 0.48; 95% confidence interval, 0.39-0.60; P < .001) were less likely to believe that their operating skill was level appropriate than those who felt comfortable asking attending surgeons for help. These residents also were more likely to worry about feeling confident enough to perform procedures independently by the end of training. Residents who were not satisfied with their operative experience were less likely to feel that their operating skills were level appropriate and more likely to worry about feeling confident by the end of residency (Table 7).
An association was identified between surgical confidence during residency and career goals by studying resignation rates and intent to pursue fellowship training. Residents who believed that their operating skill was level appropriate tended to be less likely to drop out during residency (1.8% vs 2.8%; P = .06) and significantly less likely to pursue postresidency fellowship training (60.1% vs 71.2%; P < .001). Residents who had concerns about operating independently by the end of residency did not have higher dropout rates (2.7% vs 1.9%; P = .13), but were more likely to plan for postresidency specialty training (71.2% vs 60.2%; P < .001).
Based on the National Study of Expectations and Attitudes of Residents in Surgery survey of all general surgery residents in the United States, more than two-thirds of general surgery residents feel confident with their operating skills. However, several resident- and program-level factors were associated with greater trainee confidence, including male sex, being a more senior resident, being married, and having children. A higher percentage of residents at training programs at community hospitals, those with fewer chief residents, and programs that were not associated with fellowships were confident than residents at programs without these attributes. After adjustment in our multivariate logistic regression model, sex, PGY level, marital status, and residency program type remained independently associated with our measures of confidence.
Practical learning experiences increase confidence levels. Studies have demonstrated that surgery-based electives for senior medical students entering surgery residency are effective in developing surgical confidence and competence.4,7- 9 Clinical simulation experiences also have been shown to improve resident confidence with procedural skills.10,11 As residents progress from their internship to chief residency, they accumulate surgical experience. In our survey, resident confidence with their perceived operating skill being level appropriate increased during the course of training. Because senior residents usually operate on the most difficult or technically challenging cases, they have better access to opportunities that will enhance the quality of their operative experience. However, we observed that junior residents in the middle clinical training years were most worried about feeling confident enough to perform procedures independently by the end of training. This appears to be a reflection of the common perception that the middle years of general surgery residency training are particularly stressful. This is the time when residents are expected to assume increasing clinical and operative responsibilities while making decisions about career paths and applying for postresidency specialty training. The transition from first assistant to primary surgeon tends to occur most commonly during the PGY4 and PGY5 years. This transition likely enhances resident confidence in their skills.
In a 2007 study by Binenbaum et al,3 surgical and medical residents rated the extent to which various elements of internship training contributed to the development of physician confidence. Residents generally valued independence, responsibility, exposure to a broad spectrum of clinical cases, guidance from attendings, and backup support as factors that contributed to enhanced confidence. Surgical residents, in particular, placed great value on their operative experience, especially during clinically demanding surgical rotations. These elements of training presumably differ between the types of training programs and may explain the discrepancy observed between the confidence levels of residents in community-based vs academic training programs. In addition, this study supports our finding that residents who felt comfortable asking their attending surgeons for help reported higher levels of confidence. Residents in the Binnebaum study rated “working through decisions with residents/attendings” and “more senior-resident interactions” in the top 10 elements contributing to the development of physician confidence, highlighting the importance of mentorship.
More than 20% of PGY5 residents expressed concern about their surgical skills and independence by the end of surgical training. Because most of the PGY5 residents in our survey were part of the first cohort to start general surgery training in the era of the 80-hour workweek, lack of surgical confidence could, in part, be a result of the reduced operative experience and fewer learning opportunities available compared with their pre–80-hour workweek counterparts.12 Some studies have reported a reduction in the number of cases performed by chief and junior residents as surgeons since the inception of the restricted work hours.13 It is estimated that 77% of current general surgery residents plan to pursue postresidency specialty training. The rapid increase in the proportion of residents pursuing fellowships over recent years14 reflects, in part, their perceived need for additional training to bolster their surgical confidence and skills. Confidence is essential for optimal performance in the operating room, because critical decisions during an operation often must be made rapidly and under pressure, often without the opportunity to solicit additional data or colleague opinion.
Confidence can be learned. A lack of confidence may stem from feeling overly criticized, not being accorded enough independence, or feeling undermined. There are many personal factors that influence individual confidence levels, including sex, social factors, upbringing, and culture. Our study found that female residents were almost 2 times more likely than male residents to worry about their ability to perform procedures by the end of training. Prior studies have suggested that women may be more likely than men to drop out of general surgery training programs15; however, more recent studies have dispelled the association between sex and attrition.16 Studies have shown that women face different challenges during general surgery than men, such as having to make more sacrifices in terms of marriage and childbearing.17
Many of these challenges may manifest themselves in lower confidence levels during residency and may affect career decisions around specialty training. Married residents and residents with children were more likely to be confident. Male residents were more likely to be married and have children, which may add to feeling more competent in juggling multiple responsibilities and handling potentially stressful situations. In addition, having stronger home/family support may bolster confidence at work.
This study has several possible limitations, inherent to a self-report survey. Respondents were assured of strict study confidentiality, which we hoped encouraged truthful answers. Reported feelings of confidence do not necessarily correlate with objective measures of competence and skills.18 This correlation has been evaluated in PGY1 residents, and showed that PGY1 residents self-reported a greater improvement in skills than what was actually observed by an objective rubric over the course of their first year of surgery residency. Lacking confidence may sometimes be a catalyst for self-improvement and self-reflection, which are necessary traits in continuing professional development. Overconfidence can be a detriment to one's development. These factors could not be measured by our survey.
In summary, enhanced confidence in surgical trainees appears to be independently associated with being a senior resident, training at a community program, and training outside of the Northeast. Our study suggests that standardization of operative experience and clinical responsibilities, based on the individual and programmatic factors identified, have the potential to help reduce the variation observed in confidence levels across surgical trainees in the United States. Programs can encourage close, collegial, and respectful relationships between trainees and surgical faculty, facilitating resident-attending interactions and improving communication around cases. Feeling comfortable discussing failure with attending surgeons increases resident ability to learn from setbacks and increases resilience. This can bolster a resident's sense of involvement and importance and lead to more independence of thought.
Correspondence: Sanziana A. Roman, MD, 333 Cedar St, TMP 208, Department of Surgery, Yale University School of Medicine, New Haven, CT 06520 (firstname.lastname@example.org).
Accepted for Publication: April 8, 2011.
Author Contributions: Ms Bucholz 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. Study concept and design: Bucholz, Sue, Yeo, Roman, Bell, and Sosa. Acquisition of data: Bucholz, Yeo, and Bell. Analysis and interpretation of data: Bucholz, Sue, Yeo, Roman, and Sosa. Drafting of the manuscript: Bucholz, Sue, Roman, and Sosa. Critical revision of the manuscript for important intellectual content: Bucholz, Sue, Yeo, Roman, Bell, and Sosa. Statistical analysis: Bucholz. Obtained funding: Sue, Yeo, and Sosa. Administrative, technical, and material support: Yeo, Roman, and Bell. Study supervision: Roman, Bell, and Sosa.
Financial Disclosure: None reported.
Funding/Support: This study was supported by a Yale University School of Medicine Medical Student Research Fellowship (Ms Sue).
Previous Presentation: This study was presented in part at the 6th Academic Surgical Congress; February 3, 2011; Huntington Beach, California.