The traditional mode of using live animal experiments has long been employed in the teaching of surgical operations. However, the issues such as animal ethical controversies, high costs associated with animal breeding, anesthesia, and operating room construction, as well as biosafety risks, are becoming increasingly disturbing. There is a growing need for alternative approaches ^{[1, 4]}. For example, in the case of using dogs for teaching purposes, a single session can cost approximately $500 (including anesthesia, feeding, and postoperative care), and the ethical approval process for animal experiments can take as long as 3–6 months in some of the Chinese institutes, which severely limiting the efficiency of teaching ^[1]. In some countries, the use of live animals for surgical education is even prohibited. To reduce the use of live animals, alternative solutions such as virtual reality (VR) and synthetic models have been widely explored ^[7]. However, VR technology fails to replicate key physiological feedbacks in open surgery, such as tissue elasticity and dynamic bleeding. Synthetic models (e.g. silicone organs), is cost-effective, but they are significantly different from real tissues, making it difficult for trainees to master delicate operations (e.g. controlling the tension of intestinal anastomosis) ^[6]. The augmented reality liver resection simulator developed by Strickland et al.^[5] still cannot overcome the authenticity deficit, further highlighting the limitations of virtual training in complex surgeries. Taking the above scenarios into consideration, there is an urgent need in the field of surgical techniques trainingfor an alternative that can avoid ethical controversies, reduce costs, and provide realistic tissue feedback.

Ex vivo organ models, with their anatomical authenticity, cost-effectiveness, and ethical friendliness, are gradually emerging as a new option for surgical education. These models utilize discarded organs from slaughtered animals (e.g. pigs used for meat) and simulate blood circulation through perfusion techniques (with a flow rate of 5–10 mL/min), enabling the organs to exhibit mechanical properties close to those of living organisms during training (e.g. intestinal peristalsis and vascular pulsation). Research by Wang et al.^[1] has shown that ex vivo porcine liver model platform, the Smagister, is as effective as live animals in laparoscopic surgical training, with a cost reduction of 60%. However, their study was limited to the training of specialist physicians and did not address the basic skill training of undergraduate students. More importantly, the existing literature rarely discusses the role of ex vivo organs in the development of non-technical skills, such as ethical concerns, teamwork, and dynamic physiological judgment. Therefore, there is an urgent need for systematic research to verify the comprehensive value of ex vivo organs in undergraduate surgical education, especially in balancing ethical costs, operational convenience, and the efficacy of surgical skills practice.

This study compares the teaching efficacy, ethical issues, and operational advantages and disadvantages of the Smagister ex vivo organ perfusion system and live animals in undergraduate surgical skill training through a randomized controlled trial. The study aims to provide empirical evidence for optimizing undergraduate surgical training programs with Smagister and laying the foundation for future curriculum design.

The control group used healthy Beagle dogs. Anesthesia was induced with an intravenous injection of sodium pentobarbital at 30 mg/kg, with a maintenance dose of 5 mg/kg/h. The dogs were fixed on the operating table, with real-time monitoring of heart rate (ECG) and oxygen saturation (SpO₂). The surgical procedures included gastrostomy, gastric perforation repair, small intestine resection, and intestinal anastomosis. Postoperatively, euthanasia was performed by a licensed veterinarian with an intravenous injection of potassium chloride solution at a dose of 1 mmol/kg.

Both groups followed a unified teaching syllabus (Fig. 1):

In phase 1, the intervention group utilized Smatristers for animal surgical teaching, while the control group employed Beagle dogs. Accomplishment of gastrostomy, gastric perforation repair, small intestine resection, and intestinal anastomosis on Smagister/live models. Participant trainees exchange the roles of chief surgeon and assistant, with each procedure limited to 30 minutes.In phase 2, two groups exchanged the teaching media( Fig. 5).

A total of 20 participants were involved. There were no statistically significant differences between the two groups in terms of age, gender, handedness, and baseline scores of surgical skill assessments( Table 1).

The internal consistency reliability of the OSATS assessment tool was acceptable, with a Cronbach's alpha value of 0.78. The inter-rater reliability was also satisfactory, with a Pearson's r value of 0.76. Figure 2 indicated the differences of basic skill scores,OSATS scores and BTS scores between the interventionand control groups during the two phases of the advanced operating training courses. In phase 1, the intervention group had significantly higher OSATS scores when they practice on Smagister compared to the control group when they practiced on the live animal (23.1 ± 4.5 vs. 19.5 ± 2.5, P = 0.05), with the main advantages being in instrument handling (4.2 ± 0.8 vs. 3.5 ± 0.6) and hemostasis (4.0 ± 0.7 vs. 3.2 ± 0.5). However, the intervention group received lower scores than the control group in BTS scores in phase 2 (81.4 ± 4.6 vs 86.2 ± 4.5, P = 0.029), suggesting that live animals may not necessarily be an advantage in strengthening the practice of basic skills.. Both groups had significantly higher scores in phase 1 of advanced training courses than in the basic skill training courses. The crossover design analysis revealed that the learning effect (Cohen’s d = 0.63) made a significant contribution to the improvement of BTS (P < 0.01).

A questionnaire on animal ethical distress was administered to the participating students, after completing phase 1 of the advanced training courses. A total of 20 questionnaires were distributed and all were returned. As shown in Table 3, there was no difference regarding ethical distress between the two groups (Fig. 3). Gender, growing environment, and pet ownership experience had no significant influence on the ethical distress scores (p > 0.05), as shown in Table 2. However, 95% of the students (19/20) believed Smagister system were "more socially ethical" because they avoided the use of live animals and aligned with the concept of resource recycling.

A questionnaire concerning teaching efficiency was administered after phase 2 of the advanced training courses. A total of 20 questionnaires were distributed and all were returned. As shown in Fig. 4, the Smagister system stood out in terms of operational convenience and safety, including preparation time, biosafety, and surgical field exposure, as follows,

90% of the students (18/20) believed that Smagister required shorter preparation time (10 minutes vs. 1 hour), which is conducive to improving practice efficiency.

80% of the students (16/20) believed that Smagister provided lower risk of blood contact and reduced burden of operational harm.

80% of the students (16/20) found that the Smagister surgical field was more stable and convenient.

In contrast, live animals had advantages in dynamic situation simulation, including teamwork and tissue viability judgement, as follows,

60% of the students (12/20) believed that live animal experiments required more teamwork (anesthesia, life monitoring), which better simulate the real clinical scnarios. .

80% of the students (16/20) more acknowledged the authenticity of the temperature, elasticity, and bleeding response of live tissues.

Regarding the influence of intraoperative bleeding and the psychological tension brought about by surgical operations, the students' opinions were inconsistent.

Surgery techniques training on animals holds irreplaceable value in modern surgery education systems. It effectively cultivates medical students' surgical skills and decision-making abilities by providing realistic tissue feedback and clinical scenario simulation. In Chinese medical schools, dogs are widely used for undergraduate surgical education. According to the questionnaire results, all students believe that live animal surgery course for basic surgical operations is indispensable or irreplaceable by simulators, and 90% of students think that the current animal surgery teaching time should be increased. However, traditional animal experiments face challenges in accessibility, ethical restrictions, and operational repeatability, with the strengthening of ethical standards and the promotion of the 3R principles (replacement, reduction, and refinement)^{[4, 2]}. Ex vivo organs, with their highly realistic anatomical structures, cost-effectiveness, and ethical acceptability^[2], are gradually becoming a necessary supplement to live animal experiments.

Our study evaluated the surgical skills, as well as the psychological factors of students when they were trained on Smagister or live animals. BTS improved in both groups without significant differences. This indicates that the Smagister ex vivo organ training system can achieve similar teaching effect regarding basic skills to live animals, which consistent with the findings of Izawa et al.^[1]. OSATS scores, representing the quality of the overall surgical process, were 3.6 score higher (close to significance) in intervention group after phase 1 of advanced training course, which indicated the advantages of Smagister for beginners..Conceivable reasons might be that the Smagister system allows the trainees to reduce the nervousness on the management of bleeding, thereby enabling them to effectively concentrate on fundamental procedural steps. The broader surgical field and larger organs, while maintaining realistic tissue feedback with tissue humidity and mechanical properties under non-life-support conditions, making Smagister more beneficial for beginners than live animals^{[2, 9]}. Gromski et al. recommend that beginners should complete at least nine ex vivo surgical procedures before stepping into live animal or supervised human colorectal ESD training^[10]. OSATS also includes assessments of clinical judgement and collaborative skills.^{[2, 2]}. Our results demonstrate the advantages of Smagister in training comprehensive abilities.

However, our study reveals that the advantages of Smagister are significantly context-dependent. Incomplex operations requiring dynamic teamwork (e.g. anesthesia coordination, vital sign monitoring) and physiological judgment (e.g. bleeding assessment), live animals remain irreplaceable. In phase 2 of advanced training courses, the control group demonstrated significantly higher BTS scores compared to the intervention group (86.2 ± 4.5 vs. 81.4 ± 4.6, P = 0.029).

We conducted a survey on ethical cognition and guilt of students concerning animal experiments after the course. The results showed no statistically significant differences between the groups in all indicators. Univariate analysis of variance revealed no significant differences in guilt scores among students with different genders, upbringing environments, or pet-keeping experiences (P > 0.05). In terms of ethical cognition of animal experiments, both groups scored highly (intervention group 16 ± 1.8 vs. control group 15.7 ± 2.5, P = 0.7 ), indicating that students have a good understanding and recognition of the ethical principles of animal experiments, regardless of the teaching method used. This may be related to the widespread strengthening of ethics courses in modern medical education^[1]. Regarding guilt experience, both groups scored moderately (intervention group 13.8 ± 4.9 vs. control group 14.5 ± 3.8, P = 0.7), with no significant differences. This finding is particularly noteworthy, as it suggests that even with the adoption of the supposedly more "humane" teaching method, students' guilt towards animal experiments did not significantly decrease. One plausible interpretation is that students' ethical discomfort may reflect broader concerns about animal welfare rather than being tied to particular intervention procedures^{[2, 12]}. Current simulation technologies may lack sufficient realism to meaningfully address trainees' ethical discomfort.

Smagister system demonstrated clear advantages in terms of operational convenience and biosafety without contacting the live animals, especially when the beginners are not proficient in surgical operations. Moreover, due to its standardized organ samples, it can achieve a more homogeneous training effect. Smagister system also provided a certain degree of surgical stress simulation, and the two groups of the student's opinions are not significantly different. However, our survey suggested that from the undergraduate students’ perspectives, live animal experiments remain indispensable despite all the advantages of Smagister, including team work, and items requiring indentification of dynamic physiological responses, such as tissue viability and tissue elasticity assessment, which is crucial for developing trainees' clinical judgment skills ^[3].

For the critical stage of skill shaping for undergraduate students, unlike previous studies focusing on specialized surgeons^{[2, 2, 9]}, Smagister has its unique adaptability. The operating environment is more stable (no bleeding interference), and the surgical field exposure is more adequate. We can prioritize the use of Smagister in elementary courses, and put the limited live animal resources on advanced teaching, thereby achieving a low-cost, high-efficiency teaching effect.The study has some limitations. Sample size is small and it is a single-center study. Multi-center, large-sample studies, especially students with different cultural backgrounds, are needed. The study assessed immediate skill performance and has not been able to track the skill retention rate of trainees during their clinical internships.

Our findings establish the idea that Smagister ex vivo platform is educationally non-inferior to live animal models. Smagister exhibited advantages in economic feasibility, ethical acceptability, and biosafety compliance. This evidence supports its adoption as a promising training platform in undergraduate surgical curricula. Smagister might achieve more complex training objectives if incorporated with advanced simulation systems.