Advancing Surgical Education: A Comprehensive Systematic Review with Meta-Analysis and Novel Approach to Training Models for Local Skin Advancement Flaps

Performing local skin flaps is a challenging task that requires cognitive and technical skills to design flaps with proper orientation to avoid distorting normal anatomy. Junior trainees need adequate exposure to gain confidence and expertise in such procedures. This article systematically reviews the literature's different local skin advancement flap training models and describes a new, easy-to-use training model. A systematic review was performed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. PubMed, Cochrane, Web of Science, and Google Scholar databases were searched from their inception until August 2022 for articles about local skin advancement flap training models. The meta-analysis results were pooled across the studies using a random-effects model and presented as a weighted mean difference with a 95% confidence interval (95% CI). Out of 773 reviewed articles, 18 were included in the systematic review, and four reported enough data to be included in the meta-analysis. Rhomboid and Z-plasty flaps were the most commonly taught flaps by training models. The most commonly used training models were synthetic-based, followed by animal-based models. The training models significantly increased the trainees' confidence and expertise regarding local skin flap procedures (p<0.00001) for both domains. Training models, per our reported data, significantly improve the trainees' confidence and expertise in performing local skin advancement flap procedures; continuous efforts in developing and establishing new, simple-to-use, and effective training models are strongly encouraged to further improvement of surgical education and enhance the trainees' surgical skills.


Introduction And Background
In recent years, residency surgical training has become more challenging with increased working hours, increasing subspecializations, and centralizing surgical services [1]. The coronavirus disease-19 (COVID- 19) pandemic has further complicated practical surgical training due to these factors [2]. As a result, simulation and model-based surgical training have become increasingly popular [3].
Through simulation training, trainees can practice procedures efficiently and safely. The training can also provide valuable skills that can be transferred to the clinical setting and positively impact operative outcomes [4]. Local flaps are an essential technique in wound closure because they allow the mobilization of adjacent skin and subcutaneous tissue when direct closure is impossible [4,5]. Flap design and execution is a technically demanding process with cognitive and technical challenges, requiring the design of flaps appropriate to the local anatomy to avoid distortion [6]. It takes extensive exposure and practice for junior trainees to gain confidence and expertise in such procedures [7]. By using models, junior surgeons can practice, refine their skills, and gain confidence in a safe environment to efficiently perform and design local flaps, despite their technically challenging nature.
Many surgical simulators have been mentioned in the literature, but there have yet to be any models for training skin advancement local flaps other than synthetic materials or animals [8]. We aim to highlight all the available models and simulators for teaching local flaps in and to explore their impact on the trainees by conducting a meta-analysis. Furthermore, we describe a simple and affordable training model that trainees from all over the globe could utilize.

Systematic Review
This review was conducted by the International Prospective Register of Systematic Reviews (PRISMA) guidelines [9]. The study has been registered in PROSPERO, the International Prospective Register of Systematic Reviews, with the registration number 396851.
We conducted a systematic literature review using the four major databases, Google Scholar, PubMed, Cochrane, and Web of Science, searching for all the articles discussing the local advancement skin flap training models from the inception of the databases until August 2022.
The following keywords were utilized: "local advancement skin flaps training", "plastic surgery training models", "cost-effective local flap training models", "advanced flap teaching models", "cadaveric training models", "animal training models", and "computer-based training models". The references of the included articles were reviewed to ensure that all studies were included.

Studies Selection
The systematic review included all articles that discussed training or simulation models for local advancement skin flaps. We ensured quality control by excluding abstracts and letters published only as abstracts and using a reproducible search method. Articles written in languages other than English were not included. We also excluded models not explicitly designed for local flap simulations and articles that did not report relevant outcomes.
This review excluded articles that examined free flap simulation models. For this systematic review, no limitations were applied regarding publication year, publication status, or type of study. Disagreements regarding selection status were resolved through discussion. The senior author Hatan Mortada (HM) was consulted for a third and final opinion if no consensus could be reached.

Data Extraction
Data extraction was performed by all team members and further checked by the senior author for accuracy. The extracted variables include; country, authors, year of publication, flap type, simulation model, the purpose of developing the simulation model, evaluation method, and advantages and disadvantages, were obtained for the systematic review portion; while the questions assessing the confidence of the trainees before and after utilizing the training models, and the questions assessing their expertise, were extracted to perform the meta-analysis portion.

Level of Evidence Assessment
We assessed the selected studies for their level of evidence and recommendation using a modified educational Oxford Center for evidence-based medicine categorization system, where a level of recommendation of one is the greatest, and a level of recommendation of four is the lowest [10].

Statistical Analysis
All analyses were conducted using RevMan (version 5.4.1; Revman International, Inc., New York City, New York). Means and standard deviations of scoring for the questions assessing the trainee's confidence and expertise levels were extracted from included studies before and after the training models. A weighted mean difference with 95% confidence intervals (CIs) was pooled using a random-effects model. Forest plots were created to evaluate the results of pooling. A p-value less than 0.05 was considered significant; Heterogeneity between trials was assessed using the Higgin I2 test according to the Cochrane Handbook.

Easy-to-Use Training Model
We present five uses of this model in different local skin advancement flaps. The primary instruments needed for applying this training model are a piece of a surgical drape, forceps, suture forceps, scissors, blades, markers, and needle holders ( Figure 1). Any suture from 1-0 to 4-0 can be used for the suturing. With a thickness of 1-2 mm, a piece of a surgical drape (5x5 cm) was cut to size to fit into the hanger. Following the marking of the flap design with a marker, the blade was used to cut through the cloth, and the incision was rotated to aim at the flap, as shown in Figure 2. Using sutures, the flap is rotated and fixed accordingly (Figures 2-4).
Among all the studies included, eight articles discussed the study design. Two of the studies were randomized controlled trials [15,25], four were cross-sectional studies [7,13,16,17], one was a prospective cohort study [3], and one was a review article [12]. The studies were conducted in different parts of the world, including seven articles in the United States of America (USA), three articles in the United Kingdom (UK), three articles in Germany, and one article in Turkey, Sweden, Malaysia, and Japan. Figure 5 summarizes the PRISMA methodology for conducting a systematic review.

Study Characteristics
Across all studies, 324 candidates participated in different types of flap training. Four articles did not specify the number of candidates. The level of training and specialty of the candidates included in the different studies were diverse and ranged from medical students to faculty members. In total, four articles focused on medical students (n = 4), four articles on plastic surgery residents (n = 4), four articles on ear, nose, and throat (ENT) residents (n = 4), one article on dermatology residents (n = 1), and two articles mentioned surgical residents.
The three most frequently mentioned flaps were the rhomboid flap (mentioned four times), the Z-plasty (mentioned four times), and the Bilobed flap (mentioned three times). These flaps are the most commonly taught in surgical training programs.

Overview of Local Flap Training Models
In this study, we reviewed 18 training models for local flaps. Six of these were animal-based models, using cadaveric cattle digits, chicken thigh, pig head, porcine, rat's skin, and turkey thigh. Ten models were synthetic based, with four of them being 3D models and the rest utilizing materials such as Allevyn dressing (Smith & Nephew, London, United Kingdom), gelatin skin, or didactic materials. One study employed the use of human skin, and four articles focused on computer-assisted and virtual reality simulation, including mobile simulation applications or computer-based systems.
Regarding the evaluation methods, a variety of techniques were used. However, the most prevalent was the questionnaire-based evaluation method. Despite the diversity in the evaluation and progress assessment methods, most studies demonstrated a positive impact of the training models and a marked improvement in the participants' skills. Table 2 summarizes the advantages and disadvantages of each of the training models reviewed in this review. The costs of the training models were mentioned in five articles, while several other articles only mentioned "cost-effective" without specifying the exact amount. Costs ranged from 3.83 USD to 350 USD.  Replicates elasticity of natural skin NM Our suture pad simulates human tissue well. In flap workshops, this suture pad is a valuable teaching tool.
Hassan (2014) [22] Enables an understanding of 3D design and flap movement simulates the operation of face-like structures that have complicated 3D structures NM Using this elastic model, we taught residents and young doctors how to make several typical local flaps and to perform cheiloplasty. They could experience realistic simulated surgery and understand threedimensional movement of the flaps.

Denadai (2014) [23]
A thorough simulation of the advancement, rotation, and rhomboid flaps closure on a realistic one-layered 3D model of both genders Limited to rhomboid flaps, doesn't provide information on flap retention, only evaluated on one small sample of medical students from the same center.
This RCT showed that medical students mastered rhomboid flap regardless of the level of fidelity.

Meta-analysis Results
Trainee's confidence before and after utilizing training models: Yang et al., Ederer et al., Kite et al., and Denadai et al., reported data about the trainees' confidence in performing local skin flaps before and after utilizing the training models [3,7,16,24]. We pooled the mean score and the standard deviation of the participants in their study to investigate the impact of such training models on the trainees' confidence and found that utilizing a training model significantly increases the trainee's confidence (p<0.00001). Further details are depicted in the forest plot, Figure 6.

Quality Assessment and Risk of Bias
Following the Oxford Center for evidence-based medicine's modified classification system, where a level of recommendation of 1 represents the highest level of recommendation, and a level of recommendation of 4 represents the lowest level of recommendation [10], we graded the selected studies in terms of their level of evidence and recommendation. Ten articles were rated level 3, three were rated level 4, three were rated level 2b, and two could not be assessed due to insufficient data ( Table 1).

Discussion
Training on local flap techniques is essential for residents in plastic and reconstructive surgery [3]. These procedures are often complex and require a thorough understanding of anatomy, blood supply, and surgical technique. It is expected that residents who have been trained in local flap techniques will be better able to handle a variety of reconstructive cases and will be able to provide their patients with more effective care [8]. This, in turn, can lead to improved patient outcomes and a better overall experience for the patient.
Various obstacles can prevent new surgeons from learning and acquiring surgical skills, ranging from the challenging and stressful operating theatre environment to patient safety concerns and increased working hours, which limit the time available for bedside teaching [9]. Here is where the surgical training models come in handy. Even though these models cannot replace real clinical experience and bedside teaching, they facilitate the learning process. This article aims to provide a comprehensive review of flap simulation methods available and their impact to surgical trainees, as well as highlighting our skin advancement local flap training model, which is easy to use, accessible, and readily available. Our systematic review and metaanalysis included 773 articles across different databases; only 18 were included in the systematic review, and four were included in the meta-analysis, with 324 candidates participating in different flap training models.
Rhomboid and Z-plasty flaps were the most commonly taught flaps by those training models. The most commonly used training models were synthetic-based, followed by animal-based models.
In the literature, various training models have been described for teaching local flaps, including fresh human skin, animal skin, cadaveric skin, and 3D simulators [8]. This emphasizes the importance of involving simulation models in the teaching process. However, not all of these models are readily accessible to surgical trainees and junior surgeons. For example, plenty of abdominoplasties should be performed to provide adequate fresh human skin for training [7]. Conversely, cadaveric skin and 3D simulators are costly. Recently, various ethical issues have arisen regarding the use of animal skin for training [8].
Artificial intelligence (AI) has been increasingly employed in various aspects of medicine, including surgical training [27]. Although in its nascent stages, AI-based models hold significant potential in the domain of reconstructive surgery, especially in flap raising and insetting training [28]. Machine learning and deep learning algorithms, capable of learning from complex and extensive datasets, could assist in crafting realistic surgical simulations and providing objective, real-time feedback to residents [29]. However, literature on this specific application of AI in the local skin advancement flap training model is limited.
Incorporating AI into surgical education could revolutionize training approaches, creating a safer, more effective learning environment [29]. Advanced simulations could offer residents hands-on experience with flap design, raising, and insetting and tailor learning to individual skill levels. Yet, despite these promising prospects, there is a clear need for additional research to validate the effectiveness and feasibility of these models in day-to-day surgical training.
In this study, we shed light on a home-made, simple-to-use, easy-to-handle, and time-saving training model by which plastic surgery residents can improve their surgical skills in performing different types of local flaps, including classic unidirectional advancement flaps, V-Y advancement flaps, Z-plasty, rotational flaps, and bilobed flaps. The primary material used to create this model is easily accessible to any household. In addition, to surgical instruments that any surgical resident of any other discipline would causally own. The time required to build this model is only one to two minutes. However, our model has its limitations. The first problem is that this model needs more realism regarding elasticity and rigidity. Secondly, it lacks the multiple layers of natural skin. Several of the models described in this review are also at an early stage of development. These have been included because they are undoubtedly exciting and significant when indicating potential future directions for simulation training in local flap surgery. This article clearly illustrates the usefulness of this model despite its limitations. Our recommendation for future articles is to assess whether junior surgeons are satisfied with this training model and if any improvements can be made. This innovative, simple-to-use model meets most surgical residents' needs. However, further studies need to observe how plastic surgery residents, in particular, and surgical residents from other disciplines, in general, improve after completing skill courses using this model.

Conclusions
In our systematic analysis, most models described were evaluated in only a few cohorts. As a result, larger candidate sizes will be required, as well as standard assessment methods. The meta-analysis showed a significant improvement in the trainees' expertise and confidence levels following the training models' usage.
We also proposed a new model to be utilized by the trainees. An improvement in training in a complex area of surgery such as this will require further development and evaluation of promising high-fidelity models. A continuous effort in building and establishing innovative training models is strongly encouraged to improve further the trainees' confidence in performing complex procedures.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.