A Novel 3D-Printed Mouse Model for Surgical Training: Multicenter Construct, Face, and Content Validation Study

Ethical review.

This study was reviewed and approved by the University of Arizona Institutional Review Board, which complies with the Federalwide Assurance with the Office for Human Research Protections. The project was deemed exempt from full Institutional Review Board review, with a risk level of minimal risk to participants. The study did not involve animal subjects, and no IACUC approval was required. All research procedures were conducted following the approved protocol, the Belmont Report, and applicable regulations outlined in 45 CFR 46.111 and 21 CFR Part 50. The study adhered to ethical standards for human subject research, ensuring participant safety and confidentiality throughout the project.

Simulator design and setup.

The midlaparotomy mouse model simulator used in this study was conceptualized and developed by Dr. Celdran at the University Animal Care Department of the University of Arizona (Tucson, AZ). The simulator was designed based on a mouse cadaver fixed in dorsal recumbency (the standard surgical position for midline laparotomy procedures), maintaining 1:1 scale and external morphologic fidelity. For this validation study, early-stage prototypes were produced using fused deposition modeling with standard biocompatible thermoplastic (polylactic acid) using a P1P model printer (Bambu Labs, Austin, TX). A commonly available synthetic material (0.2032-mm-thick nitrile gloves) was used to simulate the structural composition (skin and muscular layers) of a mouse abdominal wall. The inorganic layers were held in place with interchangeable custom-fit inserts that allow for the replacement of the layers and repeated practice. The design of the “3R Mouse” simulator (Figure 1A) tested in this study has been submitted for full patent protection with the US Patent and Trademark Office.

View Figure

• Download as Powerpoint
• Download Figure

Study design.

Each of the collaborating centers recruited 6 inexperienced and 3 expert subjects for a total of 45 participants. Inexperienced subjects were defined for this study as trainees requiring rodent surgery training who had never performed any surgical procedure. The expert subjects were defined for this study as experienced and proficient rodent surgeons with over 50 rodent surgeries performed during the 3 mo before the simulator assessment. The subjects were assigned to each group based on the above-mentioned criteria and independently of their educational or professional background.

A comprehensive and standardized approach was used to harmonize the data collection and testing procedure among participating institutions. All enrolled subjects attended a standardized orientation/training session in which the same presentation and aiding materials were used. Each center’s study coordinator, a laboratory animal veterinarian with extensive rodent surgical training experience, explained the study objectives and methodology, enrolled participants, provided standardized guidance on proper surgical instrument and needle/suture handling, and identical testing materials. The standardized presentation also included step-by-step video demonstrations of all the standardized tasks to be performed by the participants as part of the validation study.

Construct validity assessment.

To evaluate the construct validity of the 3R Mouse as a simulator for surgical training, we aimed to determine (1) whether it is effective for training basic rodent surgical skills, (2) whether it helps the user to acquire key surgical competencies, and (3) whether it can distinguish between different skill levels (for example, inexperienced compared with expert). To do this, all enrolled subjects completed the same 3 tasks during each simulator iteration. To perform the tasks, the simulators were secured in place with tape and covered with a translucent plastic draping (Press ‘n Seal; SC Johnson Professional, Racine WI) to replicate the standardized conditions typically encountered in rodent aseptic surgery (Figure 1B). All subjects were provided with the same standardized materials (Figure 2) to complete all tasks in every iteration.

View Figure

• Download as Powerpoint
• Download Figure

Task 1 involved creating a window through the transparent draping, making a 20-mm-long incision through the premarked simulated skin layer, placing and fixing the provided skin retractors, and making a 15-mm-long incision through the simulated muscular layer along a premarked line. Task 2 involved closing the simulated muscular layer with a single, continuous, simple running suture. Task 3 involved closing the simulated skin layer using simple interrupted surgeon’s knots.

The same video demonstration used during the standardized orientation training session, along with identical step-by-step printed reference material, was available at all testing facilities for the subjects to consult during and between task exercises. Subjects in the inexperienced group completed a total of 5 iterations (performing all 3 tasks on each iteration) within a maximum period of 10 d, but completion times and standardized images were only collected for the first, third, and fifth iterations. Subjects in the expert group completed only 3 iterations within an equal 10-d period, with completion times and standardized images collected for all 3 iterations.

The study coordinator at each participating center recorded the total time taken to complete each task, the total time needed to fully complete all 3 tasks, and standardized photographs once each task was completed (Figure 3).

View Figure

• Download as Powerpoint
• Download Figure

The standardized photographs from tasks 2 and 3 were scored using preset criteria (Table 1). Each task received a quality score, and these individual scores were then combined to calculate an overall quality score for each simulator iteration. Once all standardized images were collected, the image folders were deidentified and renamed with computer-generated aleatory alphanumeric codes to ensure that the scorer remained unaware of the participant’s location, iteration, or group. Scoring was then performed in a double-blind manner across all institutions. Each study coordinator was randomly assigned to score images from 3 other sites. The final quality score for each category was calculated as the average of the 3 independent scores for each image.

Construct validity was assessed by analyzing both task completion times, evaluated for each individual task as well as cumulatively, and quality scores, which were similarly assessed per task and in aggregate. Quality scores were based on blind evaluations of posttask photographs.

Face and content validity assessment.

To evaluate the face and content validity of the 3R mouse simulator, all participants filled out a post hoc questionnaire survey that included basic de-identified demographic data (that is dominant hand, level of education, current position, and experience working with laboratory animals), 3 questions to rate the perceived level of realism and accuracy of the simulators (face validity), 4 questions to assess the usefulness of the simulator as a tool for acquiring basic rodent surgical skills (content validity), and 5 additional questions were added to gauge the perceived value and impact of simulation-based rodent surgical training (perceived utility, emotional comfort, skill acquisition, and endorsement of the 3R Mouse simulator). A 5-level Likert scale was used to rate the quality of each of the listed aspects.

Statistical analysis.

Statistical analysis was conducted using R software (R Core Team, 2023), version 4.4.1. The analysis consisted of linear mixed-effects models to evaluate the effects of 2 fixed variables, group (experienced and inexperienced) and iteration, as well as their interaction, on the response variables time and quality across the different tasks performed (task 1, task 2, and task 3) and each iteration as a whole (total). The models also included location and individual as a random effect to account for variability between locations while focusing on the primary variables of interest. Linear models were run using the function lmer (package:lme4).¹⁷

Participant demographics.

A total of 44 participants were included in the study: 29 in the inexperienced group and 15 in the expert group. One individual initially allocated to the inexperienced group was excluded after it was determined that, although limited, that individual had prior surgical experience, thus not meeting the inclusion criteria.

All participants completed the demographic questionnaire. Among the final cohort, 41 (93.1%) were right-handed, 2 (4.5%) were ambidextrous, and 1 (2.3%) was left-handed. In the expert group, 14 (93.3%) participants were right-handed, and 1 (6.7%) was ambidextrous. In the inexperienced group, 27 (93.1%) were right-handed, 1 (3.4%) was ambidextrous, and 1 (3.4%) was left-handed.

Regarding the education level, within the expert group, 5 (33.3%) participants held a doctoral degree (PhD), 5 (33.3%) held a professional medical degree (DVM, MD, or DDS), 2 (13.3%) held a master’s degree, 1 (6.7%) had an associate degree, and 2 (13.3%) reported a high school diploma as their highest level of education.

In the inexperienced group, 7 (24.1%) participants held a doctoral degree (PhD), 2 (6.9%) held a professional medical degree (DVM, MD, or DDS), 12 (41.4%) held a master’s degree, 2 (6.9%) held a bachelor’s degree, 2 (6.9%) had completed “some college” with no degree, and 4 (13.8%) reported a high school diploma as their highest educational attainment.

In relation to their current employment position, the participants in the expert group included 1 (6.7%) doctoral student, 3 (20.0%) postdoctoral researchers, 5 (33.3%) principal investigators, 4 (26.7%) research technicians, and 2 (13.3%) veterinarians. The inexperienced group included 11 (37.9%) doctoral students, 7 (24.1%) postdoctoral researchers, 1 (3.4%) master’s student, 5 (17.2%) research technicians, 4 (13.8%) veterinary technicians, and 1 (3.4%) classified under “other.”

Finally, in regards to the overall experience working with animals in research, in the expert group 12 (80.0%) participants reported more than 5 y of experience, 2 (13.3%) had 3 to 5 y of experience, and 1 (6.7%) had 1 to 3 y of experience. In the inexperienced group, 8 (27.6%) participants reported more than 5 y of experience, 5 (17.2%) had 3 to 5 y, 9 (31.0%) had 1 to 3 y, 4 (13.8%) had less than 1 y, and 3 (10.3%) reported no prior experience with research animals. A detailed survey response distribution for all the demographic data obtained is reported in Table 2.

Construct validity assessment.

Construct validity was assessed by analyzing both task completion times and quality scores, which were derived from blinded evaluations of post-task photographs. The marginal R² values, representing the proportion of variance explained by fixed effects (group and iteration), ranged from 0.042 to 0.518. The conditional R² values, which account for variance explained by both fixed and random effects (including individual participant differences and location variability), ranged from 0.572 to 0.911.

Total time to complete all tasks.

The expert group completed all tasks across 3 iterations in an average time of 19:43 ± 3:37 min, while the inexperienced group averaged 31:60 ± 7:11 min. On the first iteration, the expert group completed all tasks significantly faster than the inexperienced group (21:17 ± 3:01 compared with 40:02 ± 9:45 min, P < 0.05). By the third iteration, both groups showed improvement; however, the expert group continued to perform significantly faster (18:03 ± 3:25 compared with 29:10 ± 5:37 min, P < 0.05) (Figure 4). By the fifth iteration, the inexperienced group’s time, although still longer, was not statistically different from the expert group’s time in the first iteration.

View Figure

• Download as Powerpoint
• Download Figure

Within-group comparisons revealed that the inexperienced group significantly reduced their total task time between the first (40:02 ± 9:45 min), third (29:10 ± 5:37 min), and fifth (26:47 ± 6:10 min) iterations (P < 0.05) (Figure 5A). The expert group showed consistent but nonsignificant improvement from the first to the third iteration (from 21:17 to 18:03 min) (Figure 5B).

View Figure

• Download as Powerpoint
• Download Figure

Time to complete individual tasks.

The inexperienced group demonstrated continuous improvement across repetitions. For task 1, the average completion time decreased from 4:07 ± 1:47 min on the first iteration to 3:20 ± 1:30 min on the third and 2:60 ± 1:23 min on the fifth. This reduction reached statistical significance only when comparing the first and fifth iterations. For task 2, the completion time decreased from 16:46 ± 4:28 min in the first iteration to 11:27 ± 2:41 min (third iteration) and 10:17 ± 2:52 min (fifth iteration), with both reductions reaching statistical significance (P < 0.05). Similarly, for task 3, the time dropped from 19:09 ± 5:22 min in the first iteration to 14:46 ± 3:17 min (third iteration) and 13:30 ± 3:48 min (fifth iteration), also showing statistically significant improvement (P < 0.05) (Figure 5C).

The expert group’s task times remained relatively stable across iterations. For task 1, the average time decreased from 3:39 ± 1:27 min in the first iteration to 3:04 ± 1:07 min in the second and 2:48 ± 1:03 min in the third. The time recorded for task 1 on the third iteration was significantly different from the first (P < 0.05). For task 2, the average time was 8:05 ± 1:53 min in the first iteration, 7:03 ± 1:54 min in the second, and 6:10 ± 1:23 min in the third. For task 3, the expert group recorded times of 9:33 ± 1:22 min, 9:43 ± 2:38 min, and 9:06 ± 1:49 min for the first, second, and third iterations, respectively. None of these differences were statistically significant (Figure 5D).

Total quality scores.

Regarding overall quality, the inexperienced group averaged 17.6 ± 2.6 points across all 3 iterations, while the expert group achieved an average score of 21.7 ± 1.3 points. On the first iteration, the expert group’s quality score was significantly higher than that of the inexperienced group (20.5 ± 1.6 compared with 14.2 ± 2.5, P < 0.05). By the third iteration, both groups showed improvement, but the expert group continued to perform significantly better (22.7 ± 0.8 compared with 18.1 ± 2.9, P < 0.05) (Figure 6). By the fifth iteration, the total quality score of the inexperienced group (20.5 ± 2.3) was not statistically different from the expert group’s score in their first iteration (20.5 ± 1.6).

View Figure

• Download as Powerpoint
• Download Figure

The inexperienced group exhibited a statistically significant increase in average quality scores from the first iteration (14.2 ± 2.5 points) to both the third (18.1 ± 2.9 points) and fifth (20.5 ± 2.3 points) iterations (P < 0.05; Figure 7A). The expert group also showed an increase in average scores across iterations, from 20.5 ± 1.6 points on the first iteration to 21.8 ± 1.4 points on the second and 22.7 ± 0.8 points on the third. However, these changes only reach statistical significance when comparing the first and third iterations (Figure 7B).

View Figure

• Download as Powerpoint
• Download Figure

Quality scores for individual tasks.

The inexperienced group showed consistent improvement in quality scores for both scored tasks across repetitions. For task 2, the average quality score increased from 6.3 ± 1.5 points in the first iteration to 8.8 ± 1.5 points in the third and 9.7 ± 1.6 points in the fifth iteration, with both increases being statistically significant compared with the first iteration (P < 0.05). For task 3, the quality score improved from 8.1 ± 1.5 points in the first iteration to 9.2 ± 1.9 points in the third and 10.8 ± 1.3 points in the fifth iteration. All improvements, when compared with the first iteration, reached statistical significance (P < 0.05) (Figure 7C).

The expert group’s quality scores remained stable across iterations for both scored tasks. For task 2, the average quality score was 9.5 ± 1.0 points in the first iteration, 10.6 ± 1.1 points in the second, and 11.1 ± 0.8 points in the third iteration, with a significant increase observed between the first and third iterations (P < 0.05). For task 3, the quality score was 11.1 ± 1.1 points in the first iteration, remaining consistent at 11.1 ± 0.7 points in the second and 11.6 ± 0.6 points in the third iteration, with none of these differences reaching statistical significance (Figure 7D).

Face and content validity assessment.

For face validity, participants answered 3 questions related to the simulator’s realism and surface credibility. The first question inquired about the accuracy of the anatomic size and position. Among experts, 53.3% rated it as “very accurate” and 40.0% as “extremely accurate,” while only one participant (6.7%) rated it “moderately accurate.” The inexperienced group showed similar perceptions, with 51.7% selecting very accurate, 44.8% extremely accurate, and one individual (3.4%) reporting it as moderately accurate.

The second question asked the participants to rate the accuracy of the tissue feel and haptic feedback. Ratings for tissue feel and haptic feedback were more varied. Among experts, 46.7% rated it very accurate, 46.7% moderately accurate, and 6.7% “not accurate at all.” In the inexperienced group, 44.8% found it moderately accurate, 34.5% very accurate, 13.8% “slightly accurate,” and 6.9% extremely accurate.

The third question was centered on assessing the overall anatomic realism of the simulator. Here, 53.3% of experts rated the simulator as very accurate, 13.3% as extremely accurate, and 33.3% as moderately accurate. Inexperienced participants responded similarly: 58.6% selected very accurate, 34.5% moderately accurate, and 6.9% extremely accurate.

Content validity was assessed with 4 questions addressing the simulator’s perceived educational value. First, the participants were asked about the usefulness of the simulator for learning general rodent surgery fundamentals. Among experts, 53.3% rated the simulator as “extremely useful,” 26.7% as “very useful,” and 20.0% as “moderately useful.” Inexperienced participants responded similarly, with 48.3% rating it extremely useful, 48.3% very useful, and 3.4% moderately useful. When asked about the usefulness of the simulator for learning instrument and tissue handling skills, expert ratings were 53.3% extremely useful, 40.0% very useful, and 6.7% moderately useful. The inexperienced group responded even more favorably, with 75.9% selecting extremely useful, 20.7% very useful, and 3.4% moderately useful.

The participants also judged favorably the usefulness of the simulator for the acquisition of basic suturing skills. Most participants found the simulator helpful for suturing practice: 66.7% of experts and 79.3% of inexperienced users rated it extremely useful, while 33.3% and 20.7%, respectively, rated it very useful. In regard to the overall usefulness of rodent surgery training, 53.3% of experts and 65.5% of inexperienced participants rated the simulator as extremely useful, while 40.0% and 34.5%, respectively, rated it very useful. Only one expert (6.7%) considered it moderately useful.

The complete distribution of the participant responses to the Likert-scale statements assessing face (accuracy) and content (usefulness) validity of the simulator is reported in Table 3.

Perceived value and impact of simulation-based training.

Participants were surveyed regarding their perspective on the value and impact of simulation as part of rodent surgical training. In terms of perceived utility, when asked whether ex vivo simulation should be a mandatory first step, 29 (65.9%) strongly agreed, 9 (20.5%) somewhat agreed, 2 (4.5%) neither agreed nor disagreed, 2 (4.5%) somewhat disagreed, and 2 (4.5%) strongly disagreed.

Regarding the emotional comfort and perceived pressure while training, a majority of participants (32; 72.7%) strongly agreed that they felt less pressured using artificial simulators, 7 (15.9%) somewhat agreed, and 5 (11.4%) neither agreed nor disagreed. When asked whether a variety of realistic models is needed to gain surgical skills and competencies, 23 (52.3%) strongly agreed, 16 (36.4%) somewhat agreed, 4 (9.1%) neither agreed nor disagreed, and 1 (2.3%) somewhat disagreed.

In terms of confidence gained from simulator training, 19 (43.8%) participants strongly agreed that they felt confident to perform skin and muscle sutures after using the simulator, 16 (36.4%) somewhat agreed, 5 (11.4%) neither agreed nor disagreed, and 4 (9.1%) somewhat disagreed. Finally, all participants endorsed the simulator, indicating that they would recommend the simulator to peers, with 38 (86.4%) strongly agreeing and 6 (13.6%) somewhat agreeing.

The complete distribution of the participant responses to Likert-scale statements assessing attitudes toward simulation-based training is reported in Table 4.

Validation process and results.

The validation process of a surgical simulator typically involves assessing its face, content, and construct validity using a combination of subjective and objective measures across participants with varying levels of expertise.^30–32

Face validity.

The face validity of the 3R-Mouse simulator was supported by structured postuse surveys centered on the level of realism of the simulator. The simulator received consistently high ratings across both groups for anatomic accuracy, size, positioning, and haptic feedback. Over 90% of respondents in each group rated the anatomic size and positioning as very or extremely accurate. Although the evaluations of tissue feel and haptic feedback were slightly more variable, the expert feedback on haptic realism also ranged from moderately to very accurate, with only a single negative rating, suggesting general agreement on the simulator’s realism. Despite the overall positive feedback, the face validity results highlight opportunities to improve the materials used to replicate muscle and skin layers. The prototype tested in this study used simple, low-cost materials (for example, nitrile gloves) mainly for accessibility and because of the tear-prone nature of the nitrile when applying excessive pressure on the suture. This was perceived as a positive quality by the trainers, since it encourages a gentle approach to tissue handling and suture tension in the trainees. Nevertheless, based on the face validity results obtained, we are currently exploring alternative materials that better mimic the visual and elastic properties of mouse tissues.

Content validity.

Content validity was evaluated based on participants’ perceptions of the simulator’s usefulness across various training domains. Both expert and inexperienced participants consistently rated the simulator as highly beneficial for learning fundamental rodent surgical skills. Over 90% of respondents in each group rated it as either very or extremely useful for acquiring general surgical competencies, with similarly high ratings for its role in developing instrument and tissue handling skills. These favorable ratings, although inherently subjective, are substantiated by the objective improvements documented in the construct validity assessment, which supports the credibility of the simulator’s educational value.

Inexperienced users were particularly enthusiastic, with more than 75% rating the simulator as extremely useful, reflecting the simulator’s strong perceived impact at the early stages of surgical training and the high receptiveness of the trainees to using simulations during the early stages of training. However, it is important to acknowledge that content validity ideally relies more on evaluations from experts since they are expected to possess a deeper understanding of the required competencies. In this context, the more tempered endorsement from the expert group (only 53% rated the simulator extremely useful) likely reflects a more accurate and realistic representation of its strengths and limitations.

The high rating received from both groups as a suturing practice training tool highlights the simulator’s value for task-specific skill development, and the similarly high rating on the overall usefulness for rodent surgical training reinforces the relevance of the 3R-Mouse simulator as a significant step forward in ethically grounded, accessible, and effective early-stage rodent surgical training.

Construct validity.

Construct validity was established through an objective assessment of participant performance over repeated simulator use, based on task completion times and blinded quality scores. The inexperienced participants’ significant improvements in both quantitative (time) and qualitative (quality score) measures demonstrate progressive skill acquisition and support the simulator’s utility as an effective tool for early-stage training.

It is worth highlighting that the average scores obtained by the inexperienced participants on the fifth iteration reached the level of the results obtained by the experts in their first use of the simulator (both values were not significantly different for either the time or the quality scores). This improvement underscores the value of the simulator as a training tool and hints at the level of iteration at which the inexperienced trainees start reaching a learning plateau using the simulator. While these results are highly encouraging, we acknowledge that further work is needed to assess if the skills gained with the simulator adequately transfer to procedures performed on live animals.

The expert group demonstrated significantly shorter completion times and consistently higher quality scores (Figures 4 and 6), which supports the simulator’s construct validity through its ability to distinguish between experience levels. While some performance gains were observed among experts, statistically significant improvements were limited to comparisons between the first and third iterations in only 2 separate occasions (Figures 5 and 7). The absence of generalized significant changes in expert performance suggests that the simulator represents a limited challenge for experienced rodent surgeons. Taking into account their established technical proficiency, we believe that the isolated improvements observed are likely attributable to a brief adaptation to the simulator’s synthetic materials rather than true skill acquisition.

Survey insights on the perceived value and impact of simulation-based training.

Beyond the technical validation metrics, this study offers valuable insights into participants’ perceptions of simulation-based rodent surgical training. As experienced trainers, the authors of this study were not surprised to corroborate that both expert and inexperienced users generally rated the simulator as extremely useful for acquiring basic rodent surgery skills. These findings are consistent with prior reports demonstrating broad trainee support for simulation-based learning in laboratory animal science.³³

The majority of participants strongly agreed that ex vivo simulation should be a mandatory first step in surgical training, which, in our opinion, is a loud call to action that adds empirical support for the integration of simulation approaches into rodent surgery training programs.

A balanced interpretation of these findings requires considering both the subjective nature of self-reported usefulness ratings and the differences detected between groups. The high enthusiasm among inexperienced participants is encouraging and underscores the strong receptivity to simulation-based training. In contrast, the slightly lower ratings from experts may reflect a more critical perspective or perhaps an acknowledgment of the simulator’s limitations in fully replicating the complexity and intricacies of real in vivo surgery. Although live animal models remain the gold standard for advanced surgical training (that is, microsurgical techniques), evidence supports that prior ex vivo practice effectively consolidates foundational surgical skills and enhances overall training outcomes.³⁴

Limitations.

This study has several limitations to acknowledge. First, the number of participants (particularly in the expert group) was limited, which may constrain the generalizability of the statistical comparisons. Another limitation is that although the quality assessments were double-blinded and based on a standardized scoring rubric, they remain inherently subjective. To mitigate individual bias, 3 independent instructors evaluated each image, and the average score was used for analysis. While this approach improves reliability, the development and implementation of automated or algorithm-based scoring systems would optimize consistency and objectivity in future studies.

Another notable limitation is the inconsistency in training session formats. While some participants completed their tasks individually, others did it as part of a group setting. In such environments, the potential peer pressure may have influenced performance metrics. Prior research has demonstrated that stress can impact performance during surgical simulations,³⁵ so peer-induced stress may have introduced variability in our outcomes. Future studies should consider standardizing the simulator training/testing environment, ideally, as individualized sessions, to better isolate true performance and eliminate social influences.

Finally, the conditional R² values obtained, ranging from 0.572 to 0.911, indicate that a substantial portion of the variance in performance can be attributed to both fixed and random effects, including participant- and site-specific factors. Although this reflects the overall robustness of our statistical models, it also highlights the influence of variables unrelated to the simulator itself, which limits the ability to fully isolate its effect on training outcomes.

Conclusion.

Overall, this study supports the validity of the 3R-Mouse simulator as a reliable tool for early-stage rodent surgical training. High face validity ratings, significant performance improvements among novice users, and clear differentiation between skill levels collectively demonstrate the simulator’s educational effectiveness. Its repeatability, accessibility, and alignment with the 3Rs principles position it as a scalable and ethical alternative to traditional training approaches.

The simulator offers a repeatable, accessible, and low-risk environment for foundational surgical skill development, and its potential application in competency-based assessments addresses a critical need for standardization across training programs. Essential skills such as proper instrument handling, different basic suture patterns practice, and principles of aseptic techniques can be acquired and their competency assessed without the use of any animals.

The authors firmly believe that embracing advanced training methods, such as simulation, not only enhances and standardizes educational outcomes and user confidence but also plays a vital role in advancing the 3Rs principles.

As simulation technologies and synthetic models continue to evolve, they promise increasingly realistic and effective training experiences. Conducting well-designed studies to evaluate their validity and impact is essential to ensure their broader adoption and to shape a future where high-quality surgical training and animal welfare go hand in hand.