Laboratory-Based Instruction for Skin Closure and Bowel Anastomosis for Surgical Residents

Hypothesis  Multimedia delivery of cognitive content paired with faculty-supervised partial task simulation for both excision of a simulated skin lesion with subsequent wound closure and hand-sewn bowel anastomosis would be an effective method for developing appropriate procedural skills among junior residents.

Design  Prospective cohort study.

Setting  University-based surgical residency.

Participants  First- and second-year surgical residents (n = 45).

Interventions  Surgical residents were given comprehensive instructional materials, including structured curricula with goals and objectives, text, figures, and narrated expert digital video, before the training session. A 4-hour, standardized, laboratory-based instruction session was then performed in small groups, which emphasized faculty-supervised practice. Residents were asked to (1) excise a skin lesion and close the wound and (2) perform hand-sewn bowel anastomosis. These 2 tasks were assessed before and after supervised practice. Performances were video recorded. Residents were surveyed before and after training.

Main Outcome Measures  Time to completion and Objective Structured Assessment of Technical Skill global rating scale score based on video recordings were evaluated by blinded reviewers. Final product quality was measured by anastomotic leak pressure and by wound closure aesthetic quality.

Results  Residents perceived the laboratory training to be equal to training in the operating room for skin closure and superior to training in the operating room for bowel anastomosis. Residents perceived time spent on both tasks to be “perfect.” Mean objective scores improved significantly on 5 of 6 outcome measures.

Conclusions  Junior resident surgical performance improved substantially with 4 hours of laboratory-based, faculty-supervised practice. Both first- and second-year residents benefited from this training. These data show that curriculum-driven, faculty-supervised instruction in a laboratory setting is beneficial in the training of junior surgical residents.

Laboratory-based training for technical skills is becoming increasingly integrated into surgical residency programs. Simulation-based training for laparoscopic skills has been extensively studied. Little is known, however, about the effects of this mode of teaching as it relates to the acquisition of skills for open operations despite the extensive use of this method in surgical training programs. We hypothesized that multimedia delivery of cognitive content paired with faculty-supervised partial task simulation for both excision of a simulated skin lesion with subsequent wound closure and hand-sewn bowel anastomosis would be an effective method for developing appropriate procedural skills among junior residents. To that end, we performed a prospective study of skill acquisition for these tasks based on an analysis of technical skill, time to completion, and final product quality. In addition to these objective measures, we measured subjective, posttraining, survey-based outcomes to determine the perceived value of the training session. We also aimed to determine if outcomes were similar when this mode of training was applied to first- and second-year residents and the importance of individual baseline performance as it relates to educational gain for these particular tasks.

This prospective cohort study was performed over 1 year at the Institute for Surgical and Interventional Simulation (ISIS) at the University of Washington. After multimedia-based cognitive pretraining, laboratory-based technical skill training was performed with comparisons between pretraining and posttraining objective measures, as well as between educational gains for residents of differing training levels and differing degrees of pretraining skill. Subjective perceptions of training were surveyed after instruction.

First- and second-year surgical residents attended 1 day of formal laboratory-based training as a part of a dedicated technical skills rotation (EVATS [emergency coverage, vacation coverage, academic time, competency training, and technical skills training], as previously described).¹ Technical skills rotations occurred over 1 academic year, and residents from each rotation attended the laboratory together in small groups (2-4 residents per group; 14 groups total). Laboratory sessions covered (1) skin excision and closure and (2) hand-sewn small-bowel anastomosis. Before attending the laboratory session, trainees were given instructional materials, including a structured curriculum with goals and objectives for each laboratory session and text and figures related to the technical details of the tasks to be performed, akin to a surgical atlas description. Trainees were also given an expert demonstration in the form of narrated digital video for each of the 2 tasks. Trainees were explicitly told that they were expected to know how to perform the tasks at the beginning of the day and that the laboratory session should be used to refine skills, not learn how to perform the task.

Each laboratory session began with self-directed practice, including a trial of each task (Figure 1). These trials were intended to serve as learner needs assessments, allowing the trainees to gain understanding of which areas of the tasks they needed assistance with. Trainees were given a piece of porcine abdominal skin (10 × 20 cm) with a marked 2-cm “lesion” and were asked to excise the lesion and close the wound using an interrupted vertical mattress technique. Trainees were then given a piece of harvested porcine small bowel fixed to a jig and asked to perform 2-layer hand-sewn bowel anastomosis without the aid of an assistant. Trainees were given a maximum of 65 minutes to practice each task. Faculty proctors were not present during self-directed practice.

After 2 hours of self-directed practice, trainees had 2 formal supervised practice sessions with surgical faculty (1 session each for skin closure and bowel anastomosis). Each 2-hour session was proctored by an individual faculty member. Faculty proctors were instructed to circulate around the laboratory while observing trainee technical performance and answering questions. Faculty proctors were instructed to give real-time formative feedback (both positive and negative) and were encouraged to physically work (ie, lay hands over trainee's hands) with the residents as they practiced, similar to the coaching style of athletic coaches.

At the conclusion of the supervised practice sessions, trainees were again given porcine tissues to perform the tasks. This final repetition was designed to allow trainees to attempt the tasks 1 final time without assistance to further solidify skills they had developed. Faculty proctors were not present during the final attempts.

Study procedures associated with training included objective assessment during the 65-minute, self-directed practice sessions (Figure 1) and survey assessment before and after training. During the 65-minute sessions, pretraining assessment was performed on the first attempt of each task and posttraining assessment was performed on the last attempt of each task. Most trainees performed only 1 attempt at bowel anastomosis because of time constraints. Trainees were surveyed at the beginning of the day for prior experience and at the end of the day for perceptions relating to the training procedures.

Three independent objective assessments were performed for each task before and after training (during the 65-minute practice sessions) and included (1) time to completion, (2) blinded assessment of technical skill (based on digital video recordings), and (3) final product quality. Time to completion was recorded real time in the laboratory and measured the time between placement of the specimen on the table and the time the resident stated that he or she was finished.

Digital video recordings were made of the operative field during task performance for blinded technical skill assessment (before and after training, as described herein). These digital files were coded with random numbers by the primary investigator (A.R.J.) and subsequently distributed in batches of 10 to 15 videos to independent blinded expert reviewers (L.K.M. and A.E.L.), who evaluated performance using a modified Objective Structured Assessment of Technical Skill (OSATS) global rating scale.² All videos for each task were reviewed by the same blinded expert (1 expert each for skin closure and bowel anastomosis). Reviewers had no knowledge of the temporal sequence (before and after testing) or level of training of the resident. Global rating scales were modified to exclude the instrument knowledge subscale as audio was deleted from all video recordings for blinding purposes, rendering the use of this particular scale not possible.

Final product quality for bowel anastomosis was assessed by testing for leak pressure by fixing the specimen to a pressure source and clamping the lumen. Pressure was gradually increased across the anastomosis until leakage was visible and the pressure at which the anastomosis first leaked was recorded. Final product quality for skin excision and closure was determined by blinded review of digital photographs. Three blinded reviewers (A.R.J., L.K.M., and D.J.A.) scored each specimen independently using a 7-item 5-point rating scale specifically assessing suture spacing, mattress limb symmetry, gaps and misalignments in closure, dog ears, scar length, suture alignment, and tissue eversion. Scores from individual scales were summed to form a composite aesthetic rating.

Subjective assessment included pretraining background information and posttraining evaluation. Survey data collected before training included information regarding prior skin closure and bowel anastomosis experience, including prior laboratory-based training and number of cases performed. Posttraining survey items included subjective rating of the amount of time spent (too little or too much) and value of laboratory-based training compared with operating room experience (less valuable, equal to, or more valuable). All survey items used a 5-point rating scale.

Resident trainees were recruited the week before the scheduled sessions. Institutional review board approval was obtained from the University of Washington Human Subjects Division. Study procedures were explained, and written informed consent to acquire data was obtained from all residents. Participation in the study associated with the training laboratory was optional, and participation was kept confidential from the faculty. Objective outcome measures and survey responses were not shared with the faculty and were explicitly kept separate from the residents' files. To preserve confidentiality, each workstation had a video camera above it, with no cassette placed into the cameras of residents electing not to participate. All study recruitment and consent procedures occurred before the arrival of faculty.

All calculations were performed using a commercially available software program (SPSS for Windows, version 15; SPSS Inc, Chicago, Illinois). All 6 objective outcome measures were treated as continuous ratio data and analyzed using repeated-measures 2-way analysis of variance for differences between pretest and posttest assessments, with differences in skill acquisition between first- and second-year residents signified by an interaction between resident level and improvement for each of the 6 objective measures. Additional analysis for differential learning as a function of baseline ability was performed by grouping residents by pretest quartile and analyzing for the presence of an interaction between pretest quartile and improvement using 2-way, repeated-measures analysis of variance. Interrater reliability for skin aesthetic quality was assessed for internal consistency considering a Cronbach α of .7 as acceptable, and the mean score from all 3 reviewers was used for pretraining vs posttraining comparisons. Survey data were significantly nonnormal and were analyzed for differences between first- and second-year trainees using the Mann-Whitney test. To control for experiment-wide type I error (α = .05), a Bonferroni correction for 6 comparisons was used, with P < .008 (2-tailed) per comparison considered statistically significant. Effect sizes were calculated and expressed as Cohen d.

Forty-five residents (28 first-year and 17 second-year residents) attended the laboratory. All residents elected to participate in the study. Significant mean differences were seen between pretest and posttest assessments for 5 of 6 objective outcome measures (Table 1). For skin excision and closure, significant improvements were seen in both time to completion and OSATS score, but no difference was demonstrated for aesthetic rating. Pretest vs posttest comparison for each of the 7 aesthetic rating scales also failed to demonstrate a significant difference. Interrater reliability for skin aesthetic rating was acceptable (Cronbach α = .79). For bowel anastomosis, significant improvements were seen in time to completion, OSATS score, and anastomotic leak pressure. Significant interaction between magnitude of improvement and level of training was not found for any outcome measure (P > .008) (Figure 2 and Table 2). A significant interaction or strong trend toward a significant interaction was demonstrated between pretest quartile rank and improvement for 5 of 6 objective outcome measures, with low performers generally achieving greater educational gain than high performers (Table 3).

Pretraining survey data were substantially skewed and were, therefore, not included in any analysis. All residents had uniformly performed many skin closures and a few bowel anastomoses. Posttraining evaluation data demonstrated that residents perceived the amount of time spent on each task to be appropriate and that the value of training was equivalent to that of training in the operating room for skin closure and slightly more valuable than training in the operating room for bowel anastomosis. Significant differences in perceptions were not seen between first- and second-year trainees, but a trend toward a difference in value of the skin closure training was demonstrated (Table 4).

This study shows that for 2 procedures junior residents are expected to master, cognitive pretraining sessions relating to the knowledge required for technical performance followed by a 4-hour, faculty-supervised laboratory training session significantly benefits trainees in terms of psychomotor skill development. Furthermore, we were able to quantify, under the conditions of the study, the groupwide magnitude of improvement that can be expected at this level of training. Individual trainee improvement, however, differs substantially from one trainee to another. In general, those who perform at the lower end of the scale on the pretraining performance gain more from the exercise than those whose initial performance is at the top of the scale. Despite this increased gain, however, those who initially score higher tend to end with the highest scores at the end of the exercise. These data argue that simulation as described is more useful in the initial training and its benefits may decrease when someone has already acquired substantial technical expertise.

In the design of this training model, elements we believed to be important to instruction included the following: (1) a curriculum-driven educational experience, (2) availability of the appropriate cognitive learning tools to be completed before the initiation of the psychomotor portion of the training, (3) an initial period of adaptation to the laboratory setting with a self-performed needs assessment, and (4) availability of faculty to guide the residents through the procedural portion of their training.

Experience with laboratory-based technical skills instruction has gone largely undocumented. Existing literature^3-8 focuses largely on descriptions of techniques, simulation model fidelity, and validation of assessment techniques but not necessarily on the study of instructional techniques or educational gains achieved. Limited experience with laboratory-based vascular anastomosis instruction has demonstrated immediate improvements in final product quality for junior residents⁹ and has correlated certain measures of laboratory performance with live operating room performance.¹⁰ Laboratory-based training of skin-flap creation has been shown to improve scores on a written examination, but effect on technical skill was not measured.¹¹

This study provides a model for evaluation of open technical skills in the context of standardized training. These data provide a baseline example of the magnitude of improvement (groupwide SDs of 0.5-1.1) (Table 1) that can be expected with simulation-based training for skin excision and closure and hand-sewn bowel anastomosis. Survey data demonstrate that laboratory-based training is perceived by junior residents to be approximately equal to training in the operating room (which would equal a subjective transfer factor of 100%) for skin closure and slightly more valuable than training in the operating room for bowel anastomosis. In addition, the data reflect that 4 hours of training and practice per task is perceived to be appropriate.

The lack of improvement in skin aesthetic quality may be due to a lack of sensitivity of the rating scale instrument. The lack of improvement on any of the 7 rating scales, however, would suggest that this is likely not the case. Alternatively, the poor elasticity and thickness of porcine skin lead to difficulty in obtaining an ideal cosmetic outcome, possibly rendering this assessment invalid. In addition, most residents had substantial prior experience with skin closure, and cosmetic outcome may not have been improved by this intervention. Despite this substantial prior experience, few residents were able to achieve a perfect aesthetic score. Future study of this assessment with faculty expert performance to investigate the validity of aesthetic assessment with porcine skin is in progress.

Analysis for differences in task acquisition for the 2 levels of trainees failed to show any significant difference for any of the 6 objective outcome measures. This finding is possibly a type II error attributable to the small sample size. Trends toward differential learning for time to completion for both tasks suggest that there may be a greater gain from training first-year residents in skin excision and closure and second-year residents in bowel anastomosis. This theory was reflected also by the survey data demonstrating a trend toward differences in perception of the value of skin excision training between resident levels. Regardless, both first- and second-year residents benefited from the training, and it is appropriate to use this model of teaching for both groups of residents.

This study was limited to immediate outcomes and did not include long-term follow-up for retention of acquired skills. In addition, we limited our observations to the analysis of the response to only a single instructional technique. Future studies are needed to document any potential lasting effect of this intervention and to investigate other instructional techniques, including that of self-directed practice alone. Furthermore, assessment of cognitive content acquired was not performed. Correlation between cognitive knowledge and technical skill acquisition was, therefore, not possible. The use of OSATS with video-recorded performance of open surgical tasks has been previously described.^12-14 This technique permits convenient blinded technical skill evaluation but has yet to be validated. Video-based OSATS has been shown to be a poor discriminator among levels of junior trainees for suturing or tendon repair.¹⁴ Furthermore, the discriminative validity of the skin assessment rating scale has not been studied, and bowel anastomotic leak testing has recently been demonstrated to have poor discriminative validity across first- through fifth-year training levels.³ The validity of this assessment in the early portion of the learning curve, however, has not been studied. Our data suggest that it may be a useful discriminator in the context of this early learning period. Finally, as with all single-center studies, these results will need to be demonstrated in other settings before generalization.

In conclusion, this study demonstrates the immediate benefit of simulation-based training for skin excision and closure and hand-sewn bowel anastomosis. Both first- and second-year residents benefited from this training. Residents with a lower level of measured baseline skill derived greater benefit from the training. These data support laboratory-based instruction for junior surgical residents with faculty supervision in addition to standardized multimedia curricula. With the recent release of the Association for Program Directors in Surgery National Technical Skills Curriculum,¹⁵ further evaluation of open skills curricula is essential. As technical skills training in the laboratory setting evolves, continuous evaluation of educational methods will be necessary to demonstrate the efficacy of these techniques.

Correspondence: Aaron R. Jensen, MD, Department of Surgery, Room BB-487, University of Washington, 1959 NE Pacific St, Box 356410, Seattle, WA 98195 (arjensen@u.washington.edu).

Accepted for Publication: April 7, 2008.

Author Contributions: Dr Jensen 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: Jensen, Wright, Anastakis, Pellegrini, and Horvath. Acquisition of data: Jensen, Wright, McIntyre, Levy, Foy, and Horvath. Analysis and interpretation of data: Jensen, Wright, Anastakis, Pellegrini, and Horvath. Drafting of the manuscript: Jensen, Wright, and Horvath. Critical revision of the manuscript for important intellectual content: McIntyre, Levy, Foy, Anastakis, and Pellegrini. Statistical analysis: Jensen. Study supervision: Wright, McIntyre, Levy, Foy, Anastakis, Pellegrini, and Horvath.

Financial Disclosure: None reported.

Previous Presentation: This paper was presented at the Pacific Coast Surgical Association 79th Annual Meeting; February 16, 2008; San Diego, California; and is published after peer review and revision. The discussions that follow this article are based on the originally submitted manuscript and not the revised manuscript.

Additional Contributions: Karthik Balakrishnan, MD, and Farhood Farjah, MD, assisted with video acquisition.

Financial Disclosure: None reported.