Introduction

Surgery is stressful, and stress can affect human performance [1,2,3]. Stress is “a process by which environmental demands evoke an appraisal process in which perceived demand exceeds resources and results in undesirable outcomes” [4]. The concept of stress in surgery is not widely discussed, and surgical trainees learn stress management techniques from directly observing their superiors [1]. With the advent of competency by design, surgical residency programs across North America are developing curricula to teach surgical skills through simulation. High-fidelity simulation, which may mimic the real-life stressful environment of an operating room (OR), is costly and does not improve basic surgical skills above and beyond low-fidelity simulations [5,6,7]. Therefore, many programs have adopted low-fidelity simulation such as Fundamentals of Laparoscopic Surgery (FLS) training into their curricula. We know that stress can affect performance on a simulator even in low-fidelity simulations such as FLS [3]. Military research shows that building positive performance expectations is crucial to prepare operating personnel for high-demand conditions [4]. Positive performance expectations are present when the perception of available resources exceeds the perceived threat, and this is also known as self-efficacy in the psychology literature [8]. Improving self-efficacy in athletes is desirable, as it is linked to enhanced effort investment, greater persistence and improved performance [9]. Many stress-coping training programs for industries such as aviation and sport incorporate self-efficacy or self-confidence improvement into their performance training [4]. It is currently largely unknown if simulation programs we develop are effective at optimizing self-efficacy in surgical trainees. We developed and partially validated three procedure-specific competency-based modules to teach vaginal hysterectomy (VH), anterior (AR) and posterior repair (PR) to novice gynecology residents [10]. Our objectives in the current study were to evaluate whether each module improved self-confidence and satisfaction during performance of the index procedure in the real operating room (OR).

Materials and methods

This was an ancillary study of a larger randomized controlled trial of competency-based procedure-specific modules designed to teach vaginal surgery to junior residents in gynecology.

Residents at three Canadian universities (University of British Columbia, University of Calgary and University of Alberta) were included if they had previously performed fewer than five index procedures independently. Local Institutional Review Board approval was received at each participating site. Residents were randomly allocated to one of the two groups using a web-based randomization scheme. Intervention residents received educational modules; controls engaged in self-directed learning. Each module was procedure specific for VH, AR or PR and has been previously described [10]. Baseline knowledge was tested in both groups at recruitment. Following the didactic online portion, knowledge was again tested in the intervention residents. A 20-question multiple choice test designed for each module was used to test knowledge pre- and post-didactic teaching. If a passing grade was achieved (>60%), the intervention residents pursued guided practice on low-fidelity models with a trained gynecologic staff surgeon. All three mentoring staff surgeons (one at each institution) were fellowship-trained urogynecologic surgeons with high-volume clinical practices and academic interest. All three are routinely engaged in teaching of trainees at various levels (medical students, residents and fellows) on a daily basis as part of their practice in tertiary care university centers. When deemed ready by each staff surgeon, the intervention residents performed the index procedure (VH or AR or PR) in the real OR under supervision. Control residents engaged in usual self-directed learning and also performed VH, AR or PR in the OR under supervision. Immediately after the OR procedure, all residents filled out a validated self-confidence scale [11] and a satisfaction scale. Possible scores on the self-confidence scale varied between 0 and 30, with higher scores reflecting increased confidence and a score of 30 reflecting the self-confidence of an expert gynecologic surgeon [11]. Satisfaction was measured on a Likert scale, with scores from 0 to 4 and higher scores reflecting increased satisfaction with training. Scores were compared between groups. The knowledge score and self-confidence score were compared between groups using two-sample t-test. Satisfaction scores between groups were compared based on the chi-square of Fisher’s exact test as appropriate. Correlations were sought between self-confidence and various variables using Spearman correlation. Analyses were conducted using SAS software, version 9.4 (SAS Institute).

Results

Forty-six residents were recruited, 21 intervention and 25 control (Table 1). Each intervention resident completed one or more of the three modules followed by the corresponding index procedure. Most residents were in the second or third year of Obstetrics and Gynecology residency, and most had not performed the index procedure independently before enrollment in the trial.

Table 1 Baseline resident demographics
Full size table

Baseline knowledge scores were similar between the two groups (mean score: 12.3/20 versus 11.7/20, p = 0.419). Following the didactic portion of the modules, knowledge improved significantly in the intervention group and for each module tested (mean score increased by 3.3, 2.5 and 5.2 for VH, AR and PR, respectively; p < 0.001 for all three). Overall, the mean knowledge score improved from 58% to 76% (p < 0.001; 95% CI for improvement: 12–24%).

Scores on the post-test or improvement in knowledge in the intervention group did not show any significant strong correlations with self-confidence on any of the modules (Spearman correlation coefficient range − 0.24 to 0.48).

Overall, the VH module significantly improved self-confidence for performance of a vaginal hysterectomy (mean 20.2 intervention vs. 16.6 control, p = 0.021; Fig. 1). The AR and PR modules did not significantly improve self-confidence for performance of an AR (p = 0.938) or a PR (p = 0.116).

Fig. 1
figure 1

Overall procedure-specific self-confidence score (all residents)

Full size image

When the analysis was repeated only in residents who had never performed these procedures independently before, both VH and PR modules significantly improved self-confidence (mean 20.2 intervention vs. 16.4 control, p = 0.026 and mean 19.4 intervention vs. 16.6 control, p = 0.027). Again, the AR module did not significantly improve self-confidence for performance of an AR (p = 0.899) (Fig. 2).

Fig. 2
figure 2

Procedure-specific self-confidence score (residents who had never performed procedures independently before)

Full size image

Overall satisfaction was significantly better in intervention vs. control residents for the VH module (p < 0.001) but not for AR or PR modules. In residents who had never performed these procedures before, satisfaction improved through training on the VH (93% scored 3 or higher vs. 25%; p = 0.002) and PR (75% scored 3 or higher vs. 33%; p = 0.047) modules, but not on the AR (63% scored 3 or higher vs. 73%; p = 0.58) model.

Self-confidence or satisfaction was not significantly improved for any of the modules when higher level residents (PGY 3 and 4 combined) were examined separately (Fig. 3).

Fig. 3
figure 3

Procedure-specific self-confidence score by resident level

Full size image

There were significant positive correlations between self-confidence and satisfaction scores for each of the three modules and overall. The Spearman correlation coefficient was 0.64 (p < 0.001) for all residents and the average of all three procedures.

There were varying ranges of self-confidence scores in both control and intervention groups. For example, in control residents who had never performed these procedures, self-confidence for the average of all three procedures ranged from 9/30 to 23/30, whereas in intervention residents, after completion of the modules, self-confidence ranged from 15/30 to 24/30).

Discussion

This ancillary study of procedure-specific modules for vaginal surgery training in junior gynecology residents showed that self-confidence for actual surgical procedures was improved through training, preferentially in residents who had never performed these procedures independently, for vaginal hysterectomy and posterior repair but not for anterior repair. In addition, procedure-specific knowledge acquired through the didactic online portion of the module did not strongly correlate with improved self-confidence, suggesting that simulation training is needed, in addition to didactic training, for improved self-confidence.

The finding that self-confidence in the actual OR was only improved in the most novice trainees and for vaginal hysterectomy and posterior repair over anterior repair can be partially explained by the fact that these procedures have different perceived levels of difficulty and different complication types and rates. Among the three procedures, vaginal hysterectomy is likely the highest stakes procedure and thus more intimidating for trainees to undertake. It is not surprising that preparation for vaginal hysterectomy was associated with the largest improvements in self-confidence and satisfaction. The fact that these improvements disappeared in more senior residents is likely attributable to the basic nature of our practical models, designed to teach the most elementary yet essential surgical skills required. Presumably, the PGY 3 and 4 level residents felt they already possessed these skills despite the fact they had never had the opportunity to perform the index procedures independently.

Our findings are consistent with existing literature on low-fidelity simulation enhancing self-confidence and satisfaction in Obstetrics and Gynecology [12,13,14]. For example, a low-fidelity simulator used to teach primary care gynecologic procedures such as endometrial biopsy increased the self-confidence and satisfaction of trainees [12]. This holds true in other surgical and acute medical specialties as well [15, 16]. All these studies evaluated self-confidence immediately after simulation training rather than during actual OR performance of the procedure being taught. On the other hand, as far as we can tell from the literature search, our study is the first to assess self-confidence in a real OR environment after low-fidelity procedure-specific module training. We used a detailed, validated questionnaire for self-confidence assessment, which increases the value of our findings. Our trial was conducted at three different institutions across Canada and involved multiple trainees, trainers and evaluators, which improves the generalizability of our findings. Limitations of our study include the fact that our modules were designed for teaching and practice of basic surgical skills and did not include self-confidence-building strategies per se. In addition, we did not evaluate self-confidence at baseline; however, this can be explained by our trial design, calling for enrollment of the most novice residents who had preferably not performed the index procedures prior to training. The relatively small number of residents enrolled is also a limitation of our study; larger surgical education studies may refine our ability to draw conclusions on trainee self-confidence for various surgical procedures.

The wide range of self-confidence scores in the control residents, who did not undergo any standardized training before performing these procedures, suggests that self-efficacy beliefs in surgery do not rely on training alone but are multifactorial. Baseline technical competence and personality may also influence self-efficacy, just like in other hazardous or high-stress occupations [4]. There is some evidence to suggest that, before instruction, greater trainee confidence is significantly correlated to poorer performance [17]. On the other hand, many surgical surveys of trainee readiness for practice report low confidence among surgical graduates [18]. A recent systematic review of these surveys made a recommendation for more robust self-efficacy research and the development of “measurable benchmarks” [18]. Surgical self-efficacy is clearly a complex construct, and its relationship to performance is not yet fully elucidated. Future competency-based training programs may research ways to optimize self-confidence when it is too high or too low in individual trainees. Our research group is hoping to better delineate the relationship between self-confidence and actual OR performance during vaginal surgery procedures in the near future.