Podium IIIA - Curriculum Development
(S048) UNVEILING SELF-REGULATED LEARNING PRACTICES IN SURGICAL RESIDENTS: A SCOPING REVIEW
Ye Lim Sarah Lee, Mandeep Kaur, Brandon Cowan, MD, Aileen Gozali, Jacquelyn Knox, MD, Patricia O'Sullivan, EdD, MS, Hueylan Chern, MD, Adnan Alseidi, MD, EdM, Shareef Syed, MBChB, MRCS; University of California San Francisco
Introduction:
Self-regulated learning (SRL) is a challenging but proactive approach that can help surgical residents maximize learning among their many competing demands. Frameworks for SRL in this population are discussed but lack a comprehensive list of strategies. This scoping review aims to map SRL strategies used by surgical residents into a single collated resource.
Methods:
Following the Arksey and O'Malley methodological guidelines, we searched five electronic databases (PubMed, Embase, Web of Science, Cochrane, CINAHL) on July 16, 2024. Articles in English discussing SRL from the perspective of general surgery or surgical subspecialty residents were included. Two team members independently reviewed abstracts, and six team members reviewed full texts in pairs with an initial calibration process and iterative conflict resolution. We characterized resident behavior by the three phases of Zimmerman’s cyclical model of self-regulation: forethought, performance, and self-reflection.
Results:
Twenty-three articles were included out of 4493 abstracts screened. For forethought: (1) Preoperative goal setting was utilized by residents to identify and communicate learning goals to faculty via email or educational timeouts; (2) Strategic preparation for cases included usage of various resources (textbooks, journal articles, videos, atlases, patient charts, self-made documents, mentor-guided simulation) to review steps of procedures and anatomy of relevance. For performance: (3) Mental rehearsal, or practicing motor tasks in one’s mind, was used for case preparation and for skill retention in the setting of infrequent procedures, especially by senior residents; (4) Help-seeking behaviors included eliciting the expertise of supervisors and near peers through discussions of intraoperative tasks or of procedural variations among faculty. For self-reflection: (5) Written, oral, or mental reflections of one’s performance and experiences helped residents iteratively identify key learning points, areas of improvement, and reactions to clinical events. The included figure lists these behaviors in detail.
Discussion:
Our scoping review identified specific strategies that surgical residents use to maximize SRL inside and outside of the operating room. The outlined SRL behaviors can help residents engage in practice-based learning and improvement by evaluating their current practices and considering new positive habits. This review can also provide guidance for incoming surgical residents.
(S049) FOSTERING GROWTH MINDSETS IN SURGICAL INTERNS: A MULTI-INSTITUTIONAL PILOT INTERVENTION
Michael Kochis, MD, EdM1, Rebecca Tang, MD1, James Warwick, MD2, Deanna Rothman, MD1, Dandan Chen, PhD1, Jonathan B Greer, MD1, Keith Baker, MD, PhD1, Roy Phitayakorn, MD, MHPE1, Angela A Guzzetta, MD2; 1Massachusetts General Hospital, 2University of Texas Southwestern
Background
Growth mindset (GM), the belief that one’s abilities can improve over time, is associated with enhanced learning, academic performance, and resilience. Most studies on GM in medical education to date are cross-sectional or exploratory. This study describes the development, implementation, and evaluation of a longitudinal growth mindset intervention for surgical interns.
Methods
This study occurs at two large urban academic general surgery residency programs during the 2024-25 academic year. We adapted an established GM intervention to the surgical intern experience. One workshop was held just prior to starting intern year, and a second was administered after a few months of clinical work. Both workshops were 90 minutes long and included a short didactic portion followed by interactive discussion. A validated GM questionnaire was administered before and after the first workshop, and again before the second workshop; differences and effect size were assessed with nonparametric tests for repeated measures. Free response comments on goals and barriers were inductively coded and thematically analyzed.
Results
Sixty-nine interns participated in the first workshop, and 36 (52.2%) participated in both. Over 85% somewhat or strongly agreed that the intervention was helpful. The mean GM score varied significantly across the three time points (Χ2[2]=30.75, p<0.001), increasing with large effect size after the first workshop (36.2 to 38.6 out of 42, Z=4.94, r=0.82, p<0.001) but falling to baseline levels by the time of the second workshop (34.8). Interns’ goals after the first workshop were primarily cognitive (reflecting on their learning, not comparing themselves to others); after the second, they referenced more behaviors (seeking feedback, supporting peers, setting goals) (Figure 1). Barriers to implementing GMs included personal (ego and discomfort of learning new things), interpersonal (unsupportive culture including judgments from other team members), and structural factors (balancing learning with heavy clinical demands, performance evaluations).
Conclusions
A brief GM intervention is well-regarded by surgical interns and improves GM scores with large effect size. However, exposure to the day-to-day challenges of intern year adversely impacts their mindsets and ability to enact GM-promoting strategies. Learner-focused interventions may benefit from additional institutional efforts to address interpersonal and structural barriers.
(S050) RETENTION AND OVERTRAINING: EVALUATING SKILL DECAY AND IMPROVEMENT IN SIMULATED ROBOTIC BOWEL ANASTOMOSIS
Ricardo E. Nunez-Rocha, MD1, Samy Castillo-Flores, MD1, Amr Al Abbas, MD1, Ryan Zeh, MD2, Andres Abreu, MD1, Sofia Garcés-Palacios, MD1, Angela Guzzetta, MD1, Daniel Scott, MD1, Herbert Zeh, MD1, Ganesh Sankaranarayanan, PhD1, Patricio M. Polanco, MD1; 1University fo Texas Southwestern Medical Center, 2University of Pittsburgh Medical Center
Background:
Skill retention and decay are crucial in simulated surgical training. While virtual reality studies have assessed decay, the effects in biotissue inanimate drills, especially following gaps in training, remain less understood. We examined both skill retention after training breaks and performance enhancement through continued training in general surgery residents proficient in simulated robotic bowel-anastomoses.
Methods:
We evaluated 163 simulated biotissue videos from 35 general surgery residents. Group 1 (G1) included residents who stopped training after reaching proficiency and returned for one session. Washout periods were categorized as short- (<30 days), mid-(30-90), and long-term (>90). Group 2 (G2) included residents who continued training beyond proficiency. Completion time and Objective Structured Assessment of Technical Skills (OSATS) scores were collected at the proficiency attempt for both groups, at the return session for G1, and at last attempt for G2. Paired t-tests were used for normally distributed data; otherwise, Wilcoxon or Kruskal-Wallis tests were applied. A linear regression model assessed predictors of skill decay, and receiver operating characteristic (ROC) analysis was performed.
Results:
Median time between sessions for G1 was 70 days (244-28). G1 showed no significant change in completion time between their proficiency attempt and return session, (32 vs 35.5 mins, p=0.14). Median OSATS score declined from 30 to 28 (p<0.001). Post hoc Kruskal-Wallis test showed significant skill decay for washout periods >90 days, with OSATS scores declining from 31 to 28 (p < 0.001). Linear regression linked washout periods >90 days to significant OSATS decline (β=-3.54, p=0.003). Initial OSATS score predicted final performance (β=0.57, p=0.04). For G2, time decreased from 30.5 mins to 25.5 (p=0.012) and OSATS scores improved from 29.5 to 34 (p<0.001) between proficiency and last attempt, indicating retention and improvement with continued sessions. ROC analysis identified a 98-day cutoff beyond which decay occurred (AUC = 0.61).
Conclusions:
Our robotic curriculum offers retainable skills. Residents who had a gap in training >90 days exhibited significant skill decay. Continued training beyond proficiency improved time and OSATS scores, underscoring the importance of continued practice in enhancing performance while reducing skill decay.
(S051) A SURVEY-BASED ASSESSMENT OF ATTENDING SURGEONS’ EXPERIENCES WITH INTERVENING AGAINST DISCRIMINATION TOWARD SURGICAL TRAINEES
Sarah Jung1, Natalia Iding1, Poojha Prabaharasundar1, Juan Garcia1, Amanda Cooper2, Sabrina E Sanchez3, Chelsey Dorsey4, Selwyn Rogers Jr.4, Elise Lawson1, Shannon Cannon1; 1University of Wisconsin-Madison, 2Penn State, 3Boston Medical Center, 4University of Chicago Medicine
Objectives: Discriminatory behaviors against surgical trainees have been documented in surgical learning environments. These experiences have major impacts on career satisfaction and progression, including attrition from training, which impacts patient care. Teaching support for attending surgeons typically focuses on areas such as clinical skill development, but rarely focuses on ways to be allies to trainees. Our objective was to explore attendings’ experiences of witnessing discrimination, knowledge of strategies for intervening when witnessing discrimination, intervention behaviors, and facilitators and barriers to such intervention.
Methods: A survey was developed to gather information on whether respondents have witnessed discrimination against surgical trainees, types of intervention behaviors, reasons for their decisions to intervene or not, responses to discriminatory scenarios, preferences for training in intervention strategies, experiences with training, and demographics. Pilot testing and cognitive interviews were conducted. Survey completion took approximately 15 minutes. The survey was distributed via email to attending surgeons across four institutions for a total of 355 survey invitations.
Results: We received 102 responses (29% response rate). Respondents witnessed discriminatory behaviors against trainees primarily from patients, but also from medical staff and other attendings (Figure 1). Most respondents reported that they were only somewhat or a little knowledgeable about strategies for intervening when discrimination is witnessed (66.7%) and only somewhat or a little confident in their ability to enact these strategies (66.7%). Reasons for not intervening included fears of intervening incorrectly or being viewed negatively (12%), increasing harm in the situation (8%), lack of support from leadership or the institution (7%), and fear of retaliation (6%). Only 30% of respondents reported engaging in education programs that explicitly taught intervention strategies, and participants expressed interest in interactive programs that cover multiple scenarios and contexts.
Conclusion: Education on intervening when witnessing discrimination against trainees is needed for attending surgeons, specifically in confident use of effective intervention behaviors. Based on these results, we will develop and pilot an upstander education program specifically adapted for the surgical environment to support knowledge and application of intervention strategies through an interactive program with a focus on enhancing self-efficacy and addressing fears that act as barriers to intervening.
(S052) OPTIMIZING INTRAOPERATIVE TEACHING EXCELLENCE THROUGHOUT THE ACADEMIC SURGEON’S CAREER TRAJECTORY: INSIGHTS AND FUTURE RECOMMENDATIONS GLEANED FROM FOCUSED WORKSHOPS
Leigh (Joanna) Giddens, MBA, Marisa Louridas, MD, Shady Ashamalla, MD, Abdollah Behzadi, MD, Carol Swallow, MD, Ashish Kumar, MD, Amin Madani, MD; University of Toronto
Background: Academic surgeons are responsible for training future generations of surgeons. Intraoperative teaching (IOT) poses unique challenges that differ along the career trajectory. Although surgeons report high satisfaction from published courses, evidence suggests minimal real-life impact on their IOT competence.
Purpose: To identify IOT needs and challenges in early, mid, and later stage academic surgeons to ultimately inform more impactful continuing professional development programs.
Methods: Early-stage academic surgeons (0 – 15 years in practice) were paired with senior surgeon mentors and attended half-day IOT workshops in 2022, 2023, and 2024. Attendees completed pre- and post-event questionnaires. Likert scales were used to evaluate self-perceived competence and confidence of their technical and nontechnical teaching skills. Individual development goals and obstacles were recorded as short answers. Descriptive statistics were applied to analyze Likert scales and identify common themes.
Findings: Over three years, 35 early-stage and 16 mid-to-senior stage surgeons participated in the workshops. Time management, granting autonomy, and communication were prevalent challenges in all groups. Early-stage surgeons showed greater improvement in comfort and confidence scores than their senior peers (x¯ difference 0.84 vs 0.35 respectively), especially in managing stressful IOT scenarios (x¯ difference 1.27). All participant groups gave positive feedback around the workshop value and format.
Discussion: Surgeons experience a continuum of change in IOT environments throughout their careers. Challenges are perceived as increasingly complex and critical. Academic surgeons move from acquiring to refining the teaching skills needed to navigate these scenarios but may resist change in later stages. Moreover, obligations to patient safety and learner experiences compete with individual confidence and fear, emotional regulation, impatience, and perfectionism to name a few.
Implications: Highlighting stage-specific IOT challenges may enhance IOT excellence and ensure the graduation of competent, confident surgeons. Future initiatives, such as structured coaching, can facilitate the transition into academic surgical practice as well as promote lifelong learning.
(S053) UTILIZATION OF ARTIFICIAL INTELLIGENCE TO EXPAND LEARNING FORMATS
Ayman Ali, MD1, Nina M Clark, MD, MS2, Dan Scheese, MD3, Jessica Millar, MD4, Kevin Kniery, MD5, Jason Bingham, MD6, Scott R Steele, MD7, John M McClellan, MD8, Patrick Georgoff, MD1; 1Duke University, 2University of Washington Department of Surgery, 3Virginia Commonwealth University Health System, 4University of Michigan Department of Surgery, 5UT Health Department of Surgery, 6Madigan Army Medical Center Department of Surgery, 7Cleveland Clinic Department of Colorectal Surgery, 8University of North Carolina Department of Surgery
Background: The American Board of Surgery In-Training Examination (ABSITE) is taken annually by surgical trainees, who must balance studying with clinical responsibilities. Podcasts are efficient and portable, and therefore are highly valued by surgical trainees. Similarly, flashcard programs that emphasize spaced repetition, a technique where material is repeated at increasing intervals to improve recall, are popular studying adjuncts. These tools have not been combined to facilitate studying across platforms. Artificial intelligence (AI) tools, particularly large language models (LLMs), have become widely accessible with many opportunities for surgical education. We hypothesized that LLMs could be applied to podcasts to rapidly create flashcards and summaries, thereby enhancing retention of content.
Methods: We used audio from a podcast series specific to ABSITE review with 34 episodes and over 15,000 listens per day. We used audio transcription software to transcribe each podcast episode, and then applied a commercially available LLM (ChatGPT-4) to summarize content and create flashcards. We then manually reviewed summaries and flashcards for errors. These resources were then integrated into a free mobile application for user review. The final product included the podcast, full podcast transcription, high-yield summary, and flashcards.
Results: The LLM was able to efficiently provide detailed summaries of the podcast episodes with minimal manual edits (<5 per podcast). Although we did not compare time to a fully manual approach, we estimate a 10-fold increase in efficiency. A total of 669 flashcards were created, although the model did not perform as well on flashcard content creation, with approximately 10% of flashcards removed during manual review for poor content. Integration into existing infrastructure was automated and scalable, with a pipeline established for future episodes and content.
Conclusion: There are a wide variety of learning styles, and learners benefit from different presentations of material. Here, we demonstrate that AI and LLMs can be used to efficiently and effectively expand surgical education content and existing study materials. Future work will emphasize prospective creation of study materials from other resources, including new audio, video, and written resources, and evaluating the impact of these tools on exam performance.