Relevance of a simulation model to microsurgery for military surgical residents

Microsurgical training is an asset for military orthopedic surgeons who frequently treat hand or nerve injuries in external operations. The objective of this study was to evaluate a microvascular surgery simulation model intended to prepare residents prior to their enrolment in conventional degree training. Materials and Methods An experimental study was conducted to evaluate technical progress and satisfaction of military surgical residents using a model based on Japanese noodles with four tests of increasing diculty. Objective endpoints included: instrument handling, distribution and quality of stitches, duration of anastomoses and responses to the Structured Assesment of MicrosurgerySkill (SAMS) self-assessment questionnaire were also analyzed. Results


Introduction
The teaching of microsurgery is still based on animal models, which is the only way to faithfully replicate real operating conditions, in particular the di culties of dissection and manipulation of the walls of micro-vessels [1]. On the other hand, initial training in the handling of instrument and anastomosis techniques can be carried out by means of simulation. Many models have been developed and evaluated in recent years with the aim of limiting the use of animal models due to ethical and nancial reasons [1][2][3].
Such simulation models could be useful for the initial and continuing education of military orthopedic surgeons, who are frequently called upon to repair vessels or nerves in external operations [1,4,5]. In the French Army Health Service, the validation of a university degree in microsurgical techniques is thus part of the initial training of orthopedic residents [5]. Prior initiation and regular training on a simulation model would allow them to better develop their technical capacities, dexterity and self-control [6]. Such a model could also be used to maintain the skills of senior surgeons who do not have regular practice of microsurgery in their daily activity in their country.
We have therefore developed a microsurgical simulation model for military surgical residents that complies with the 3R principle: Reduce the number of animals in experiments, Re ne the methodology used and Replace animal models [7]. The objective of this study was to evaluate the relevance of this model. Our hypothesis was that the majority of participants would rapidly improve their performance without using the animal model.

Study population
An experimental study was conducted among surgical residents assigned to the French Military Health Service Academy, Ecole du Val de Grâce, Paris. Included were residents from all specialties who had not yet training in microsurgery. Excluded were residents who had already been enrolled in a university degree in microsurgery.

Experimental protocol
The aim was to simulate the performance of various types of microvascular anastomoses over four training sessions. The simulation model was based on Japanese noodles of the konnyakushirataki type as described by Prunières et al. [8]. Sutures were performed using standard microsurgical instruments (needle driver, dissecting forceps, double and single clamp) using surgical magni ers with 3.5x magni cation (YEARSUN, China) and 10.0 nylon thread.
Each session followed the same protocol and was supervised by a quali ed microsurgery instructor: 1. The participant adjusted his magni ers after explanation and demonstration by the instructor (Fig. 1A); 2. Each noodle was pierced with an intravenous catheter to provide visible and uniform light [8]; 3. The anastomosis was described by the instructor with the help of an explanatory diagram, then the suture was started after internal understanding of the objectives of the session; 4. The amount of time taken to perform the anastomosis was measured and the handling of the instruments evaluated by the monitor.
At the end of each session, the instructor checked the permeability and quality of the anastomosis while the participant completed a self-assessment questionnaire. Table I details the content of each of the sessions which consisted of various anastomosis of increasing di culty: end-to-end anastomosis, end-to-side anastomosis and bypass with double end-to-end anastomosis ( Fig. 1B) [9,10].

Evaluation parameters
The anastomosis patency and the existence of leaks were assessed by injecting physiological saline through a catheter inserted proximally to the model [8].
The quality of the anastomosis was assessed from the inside after a longitudinal opening of the noodle (Fig. 1C). The assessment was based on an adaptation of the rating scale described by Chan et al. [11] according to three criteria scored from 1 to 5: instrument handling, repair of stitches and quality of stitches. The criteria for assessing the quality of the stitches were symmetry of the grip of the two edges, the tightness of the knots, the occurrence of a wall tear or a trans xing grip ( Table 2).
The duration of the anastomosis procedure was also evaluated. It corresponded to the time elapsed between the positioning of the noodle on the double clamp and the section of thread during the last stitch. The amount of time was split in two during session 4 which included two anastomoses.
The subjective end-of-session self-assessment was performed using the SAMS (Structured Assesment of Microsurgery Skill) questionnaire, which has a high inter-rater reliability [11]. Fourteen criteria scored from 1 to 5 were assessed, with a maximum score of 70.
The primary endpoint was the comparison of scores obtained between the rst and last session. The secondary endpoint included: 1) an independent assessment of each criterion in our rating scale; 2) a comparison of the participant's self-assigned scores; 3) a comparison of the length of completion time between anastomoses during the resident's advancement.

Statistical analysis
Data were collected using Excel software (Microsoft Corp., Redmond, WA, USA) to calculate means and standard deviations. Statistical analysis was performed using R software [12]. A p value <0.05 was considered statistically signi cant.

Results
Nine residents with an average age of 28 (range: 24-35) completed all four sessions. There were six orthopedic residents, two visceral surgery residents and an ophthalmology resident. Their advancement within their six-year training course was uneven: two were in their rst year, three in their second year and three in their fourth year.
The anastomosis were permeable in all cases, with constant leaks more or less important depending on the quality of the anastomosis. Their quality was signi cantly increased between the rst and the last session overall and for each of the three criteria evaluated ( Table 3). The average length of procedure for complete anastomosis was 92 minutes (range: 60-115 minutes) during the rst session versus 52 minutes (range: 40-70 minutes) during the last session. There was a signi cant decrease in operation duration over the sessions (p < 0.001, Fig. 2).
The average SAMS score was 37.5 37.5 (range: 28-45) during the rst session compared to 47.5 (range: 37-58) during the last session. There was a signi cant increase in this score over the sessions (p < 0.001, Fig. 3). Twelve out of 14 criteria were signi cantly increased, but two had only a statistical tendency to increase: "atraumatic use of the needle" (p = 0.069) and "section of the thread" (p = 0.052).

Discussion
Numerous studies have been published in recent years in the eld of experimental microsurgery in order to test various simulation models and alternative means of magni cation under the microscope [1-3, 8, 11, 13]. The goal is always the same: to facilitate the transmission of microsurgical techniques while limiting costs and the use of animals. An ideal training model is a simple, reproducible and easily accessible tool that also complies with regulations [7]. To our knowledge, this study is the rst to report the use of a microsurgical simulation model for military surgeons in training. Although many of them will not have daily practice of microsurgery, they will have to use these skills occasionally within forward surgical structures where hand wounds are frequent [5]. This work thus meets the need for training and regular practice of microsurgery for French military orthopedic surgeons.
With the suggested model, there is a signi cant improvement in the technical performance of residents over the sessions, both in terms of quality of anastomoses and in terms of speed of execution. These objective results, in association with a signi cant increase in the average SAMS score, provide reliable evidence of improvement of surgical skills by the participants [11]. The presence of an instructor and the constant increase in technical di culty probably explain this rapid progression over the course of the sessions. This type of companionship is similar to that of traditional university training which has proven its e ciency [9,14]. In fact, our method is based on various studies in order to recommend a model adapted to the speci c needs of military surgeons in terms of initial and continuing training [1,8,13].
The use of surgical magni ers seemed obvious to us because it is often the only means of optical magni cation available in external operations. They have been used to perform digital revascularization and nerve repair in many local injuries in forward surgical structures [5]. Magni ers are also used by experienced microsurgeons for microvascular anastomoses and nerve repairs in their daily practice [15]. They are an easily accessible tool, outside of any university or hospital structure, relatively inexpensive, and adapted to the Japanese noodle model with a diameter of about 1mm. We haveshown that they can reliably perform anastomoses on the rat abdominal aorta, whose diameter (also about 1 mm) is comparable to that of digital arteries [1]. The konnyakushirataki noodle model was selected because of its availability, its very low cost and its close structure to the animal model [8]. However, although this noodle is suitable for learning how to handle instruments, it does not take into account the di culties of suturing on real vessels. The assessment of patency and leakage could not be reliably assessed in this study. Pruniers et al. [8] point out that this model does not reproduce the arterial spasm of thrombosis, nor the platelet deposit that seals the anastomosis in the rst few minutes. Unlike them, we did not use a dye in the saline, which could have facilitated the detection of some leaks. The other defects of this model lie in the noodle's low resistance and high porosity. As Pruniers et al.
[8], we observed a large number of wall tears in the rst sessions. However, parietal lesions became less frequent over the course of the sessions, indicating a progression of the criterion "atraumatic tissue manipulation". For all these reasons, we did not perform the Acland patency test [16], which we felt was hazardous and irrelevant. We were therefore forced to limit our objective assessment to instrument handling and the quality of the stitches. Despite these drawbacks, we nd that this model is still suitable for microsurgical initiation before the animal model, and that it is an interesting alternative to more elaborate or more expensive models [17][18][19].
The other limitations of this study are related to the small sample size and its purely descriptive nature, without comparison to other simulation procedures, or to the reference training model of the rat. In spite of these limitations, our model allowed students to become familiar with the handling of microsurgical instruments in order to prepare them before transitioning on to the animal model, which remains mandatory for the moment. Its main advantage is easy access: any surgeon with magnifying glasses and a set of microsurgical instruments can use it to train alone in all circumstances, including during eld deployments.

Conclusion
The proposed microsurgical simulation model appears to be relevant for the initial training of military surgical residents. It allows a fast technical advancement which is an excellent groundwork, and even a useful counterpart to classical training on animals. This simple and readily available tool could also be applied to the continuing education of military orthopedic surgeons who only occasional practice microsurgery during deployments. Figure 1 Photographs showing the working position of the participants (A), the performance of a lateral anastomosis (B) and the control of the anastomosis by the instructor at the end of the session (C). SAMS score through sessions [11]