Development and validation of a Dor fundoplication simulator Authors: Sadia Tasnim, Ravi Dhamija, Monisha Sudarshan, Daniel Raymond, Siva Raja, Sudish Murthy, James Yun, Dean Schraufnagel Published: 2026 ABSTRACT ObjectiveOur objective was to create and validate a cost-effective Dor fundoplication simulator for surgical trainees.MethodA scale-appropriate replica of the distal esophagus, proximal stomach, and diaphragmatic hiatus was created using felt. Ten experienced foregut surgeons evaluated it from 1 (not realistic) to 5 (highly realistic). From May 1, 2023, to November 30, 2023, 56 trainees performed Dor fundoplication before and after watching a focused teaching video, which was scored based on standard criteria. Postgraduate level 3 and under were considered Junior trainees. A median score of more than 4 (of 5) was considered the face and content validity benchmark. Construct validity was evaluated using the Wilcoxon rank-sum test.ResultsThe observed median [interquartile range] of the domains were 5 [4-5] (relevance), 5 [4-5] (ability to perform the task), 4.5 [4-5] (value), 4 [4-5] (physical attributes), 4 [4-5] (realism of experience), and 4 [4-4.25] (realism of materials) satisfying face and content validity. Seniors (n = 28) scored significantly higher on both objective and subjective measures for both video and specimen scores for both attempts (P < .001), satisfying construct validity. Trainees reported higher confidence in performing the assigned task (P < .001) on the second attempt, with seniors reporting higher confidence levels than juniors (n = 28). The overall simulation was rated as a median of 5 [4-5] by trainees.ConclusionsThis model satisfies all 3 benchmark criteria required to be considered validated. The global findings and comments indicate that the simulator is valuable as a Dor fundoplication teaching tool. Video score comparison. P value = .0005 (pre-scores) and P value = .0007 (post-scores). Central MessageWe have developed a medium-fidelity, low-cost, and easily replicable surgical simulator for Dor fundoplication that meets all benchmarks of robust validation and serves as a valuable teaching tool. PerspectiveSimulations range from virtual reality modules, dry lab models, or wet labs using cadavers or live anesthetized animals. Although wet labs are considered the gold standard, they are not a great repetitive practice platform. Commercial stimulators are costly and beyond the capability of most trainees to purchase. Thus, we developed and validated a medium-fidelity, low-cost Dor fundoplication model. Video score comparison. P value = .0005 (pre-scores) and P value = .0007 (post-scores). Central Message We have developed a medium-fidelity, low-cost, and easily replicable surgical simulator for Dor fundoplication that meets all benchmarks of robust validation and serves as a valuable teaching tool. Perspective Simulations range from virtual reality modules, dry lab models, or wet labs using cadavers or live anesthetized animals. Although wet labs are considered the gold standard, they are not a great repetitive practice platform. Commercial stimulators are costly and beyond the capability of most trainees to purchase. Thus, we developed and validated a medium-fidelity, low-cost Dor fundoplication model. Surgical training has shifted from “see one, do one, teach one” in the operating room to simulations to minimize patient risk. Better simulation training reduces intraoperative timing and cost. It also improves efficiency and increases the confidence of both attending surgeons and trainees. Simulation training is associated with improved technical proficiency and operative performance, which may translate to improved patient care. Simulations vary from prebuilt virtual reality modules, dry lab models, or wet labs using cadavers or live anesthetized animals.1,2 Although wet labs are more anatomically accurate, they are expensive and limited by time and resources. On the other hand, virtual reality modules do not always provide the same realism of the experience. Dry labs are a cost-effective, easy to set-up alternative that can provide a repetitive practice platform in a controlled setting.3 However, training on a dry lab model is limited by how realistic and reliable the model is. A dry lab model should be properly vetted and validated by the standard validation criteria to ensure its efficacy before routine use as a surgical teaching tool. We have constructed a cost-effective, easily replicable simulation model for Dor fundoplication, a routine procedure performed by general thoracic surgeons due to the current paucity of such models. Our goal is to validate this model and develop a fundoplication simulation curriculum for our surgical trainees. Materials and Methods Model Development A scale-appropriate replica of the distal esophagus, proximal stomach, and diaphragmatic hiatus was created using 2 pieces of felt ($5) taped together (Figure 1). The Heller myotomy was marked on the distal esophagus as a landmark. The model was fixed on a magnetic board ($30) using magnetic clips ($10). The model was then assessed by 10 experienced foregut surgeons and further modified based on their expert feedback. The model was stuffed with cotton wool to give it a 3-dimensional appearance. The final model was then used by the trainees for their assigned task.Figure 1Dor fundoplication model. Dor fundoplication model. Model Validation Criteria Three validation criteria should be fulfilled for a validated teaching tool4,5:1.Face Validity, which is defined as the degree to which the simulated environment represents reality. For example, does the simulated tissue look or feel real?2.Content Validity, which is defined as a measure of the usefulness of a training tool from the perspective of an expert user. For example, is the simulated task one that surgical trainees need to master?3.Construct Validity, which is defined as the ability of a learning tool to differentiate between different levels of trainees. For example, can the simulated task differentiate between the performance of an intern and a fellow? 1.Face Validity, which is defined as the degree to which the simulated environment represents reality. For example, does the simulated tissue look or feel real? Face Validity, which is defined as the degree to which the simulated environment represents reality. For example, does the simulated tissue look or feel real? 2.Content Validity, which is defined as a measure of the usefulness of a training tool from the perspective of an expert user. For example, is the simulated task one that surgical trainees need to master? Content Validity, which is defined as a measure of the usefulness of a training tool from the perspective of an expert user. For example, is the simulated task one that surgical trainees need to master? 3.Construct Validity, which is defined as the ability of a learning tool to differentiate between different levels of trainees. For example, can the simulated task differentiate between the performance of an intern and a fellow? Construct Validity, which is defined as the ability of a learning tool to differentiate between different levels of trainees. For example, can the simulated task differentiate between the performance of an intern and a fellow? Survey Validation Criteria 1.Item measure correlation assesses the degree to which the scores on an item are consistent with the average scores of the remaining items. A positive point–measure correlation indicates that the particular item contributed meaningful information to the construct measured by the survey as a whole.2.Internal consistency reflects agreement in participants' ratings to the extent that the items in the survey are interrelated or whether they measure the same construct consistently. 1.Item measure correlation assesses the degree to which the scores on an item are consistent with the average scores of the remaining items. A positive point–measure correlation indicates that the particular item contributed meaningful information to the construct measured by the survey as a whole. Item measure correlation assesses the degree to which the scores on an item are consistent with the average scores of the remaining items. A positive point–measure correlation indicates that the particular item contributed meaningful information to the construct measured by the survey as a whole. 2.Internal consistency reflects agreement in participants' ratings to the extent that the items in the survey are interrelated or whether they measure the same construct consistently. Internal consistency reflects agreement in participants' ratings to the extent that the items in the survey are interrelated or whether they measure the same construct consistently. Study Design The study was approved by the Institutional Review Board of the Cleveland Clinic (IRB #23-513 on May 24, 2023, with a waiver of subject consent. Face and Content Validity Ten experienced foregut surgeons from multiple institutions evaluated the model based on 6 domains, namely, physical attributes, realism of materials and experience, ability to perform tasks, values as teaching and testing tools, and relevance to clinical practice, using a modified version of a previously validated survey.6 The modified survey contained twenty 5-point rating scales (1 = not realistic to 5 = highly realistic), and one 4-point global rating scale (1 = requires a lot of improvement, 4 = no improvement needed). A median score of more than 4 (of 5) was considered the face and content validity benchmark. Construct Validity After final modifications were made to the model, 56 trainees (cardiothoracic fellows, integrated cardiothoracic surgery residents, general surgery residents, and medical students) performed a robotic Dor fundoplication before and after watching a focused teaching video and completed pre- and postconfidence surveys with an optional practice session between 2 attempts. All participants had watched an operative video of the 3-stitch Dor technique before the first attempt.7 Medical students had an additional session of familiarizing themselves with the robotic console. Postgraduate level 3 and under were considered Junior trainees. Postgraduate level 4 and up were considered Senior trainees. The 2 attempts were video recorded, and trainees were scored for the specimen and video based on the Objective Structured Assessment of Technical Skills criteria (outlined in Appendix E1) by D.S.8 To minimize potential bias, the evaluator was blinded to both the identity of the participants and the attempt number corresponding to each video at the time of scoring. Construct validity was achieved by comparing the Juniors’ video and specimen scores with those of the seniors for each attempt. Statistical Analysis All analyses were performed using R Studio (version 4.2.3). Continuous variables are summarized as median [25th, 75th percentile]. Categorical variables are summarized by frequencies and percentages. Comparative analysis for continuous variables was conducted using the Wilcoxon rank-sum test. Item measure correlation was assessed using Pearson's point measure correlation. Internal consistency was measured using Cronbach's alpha. Values 0.90 or greater show high internal consistency. Results Study Participants Of the 10 experts, 6 were thoracic surgeons, 2 were minimally invasive general surgeons, and 2 were cardiothoracic surgeons with community practice (Table E1). Of the 56 trainees, there were 10 medical students, 32 general surgery residents, 10 cardiothoracic integrated residents, and 4 fellows. Only 4 trainees had more than 20 hours of robotic experience; the majority had less than 5 hours of experience (Figure E1). Only 1 trainee had performed 50 Dor fundoplications before, and 45 had never performed one. Model Validation Face and content validity The observed median [interquartile range] of the domains were 5 [4-5] (relevance), 5 [4-5] (ability to perform the task), 4.5 [4-5] (value), 4 [4-5] (physical attributes), 4 [4-5] (realism of experience), and 4 [4-4.25] (realism of materials) (Table 1). Global opinion ratings were 3 [3-3.75], indicating that overall, experts believed the new Dor fundoplication model could be considered for teaching but could be improved slightly. These results support the face and content validity.Table 1Expert rating of the Dor modelCharacteristics of the Dor modelN = 10 (median [IQR])Face validity Physical attributes (1-5)4 [4-5] a. Size of the model4 [4-4.75] b. Landmark visualization, angle of His4.5 [4-5] c. Landmark visualization, fundus size4 [4-4] Realism of materials (1-5)4 [4-4.25] a. Overall impression4 [4-4] b. Tissue quality4 [3.25-4.75] Realism of experience (1-5)4 [4-5] a. Amount of pressure needed to place a suture through the material4 [4-5] b. Realism of anatomy4.5 [4-5] c. Does the simulation represent real a fundoplication?4.5 [4-5] Ability to perform tasks (1-5)5 [4-5] a. Placement of suture5 [4.25-5] b. Closure of angle of His4.5 [4-5] c. Folding the fundus4.5 [4-5] d. Proper securing of fundus on the hiatus4.5 [4-5]Content validity Value as a training tool (1-5)4.5 [4-5] a. As a training tool5 [4.25-5] b. As a testing tool4 [4-4.75] Relevance to practice (1-5)5 [4-5]Global (1-4)3 [3-3.75] Table 1Expert rating of the Dor modelCharacteristics of the Dor modelN = 10 (median [IQR])Face validity Physical attributes (1-5)4 [4-5] a. Size of the model4 [4-4.75] b. Landmark visualization, angle of His4.5 [4-5] c. Landmark visualization, fundus size4 [4-4] Realism of materials (1-5)4 [4-4.25] a. Overall impression4 [4-4] b. Tissue quality4 [3.25-4.75] Realism of experience (1-5)4 [4-5] a. Amount of pressure needed to place a suture through the material4 [4-5] b. Realism of anatomy4.5 [4-5] c. Does the simulation represent real a fundoplication?4.5 [4-5] Ability to perform tasks (1-5)5 [4-5] a. Placement of suture5 [4.25-5] b. Closure of angle of His4.5 [4-5] c. Folding the fundus4.5 [4-5] d. Proper securing of fundus on the hiatus4.5 [4-5]Content validity Value as a training tool (1-5)4.5 [4-5] a. As a training tool5 [4.25-5] b. As a testing tool4 [4-4.75] Relevance to practice (1-5)5 [4-5]Global (1-4)3 [3-3.75] Expert rating of the Dor model Construct validity Seniors (n = 28) scored significantly higher than juniors (n = 28) on both objective and subjective measures for both specimen and video scores for both attempts (P < .001), satisfying construct validity (Figures 2 and 3 and Figures E2 and E3, respectively). When the medical students are excluded, specimen and video scores for both attempts (P < .05), satisfying construct validity (Figures E4 and E5, respectively).Figure 2Specimen/model score.Figure 3Video score. Specimen/model score. Video score. Survey Validation Item measure correlation All domains correlated positively (Pearson correlation coefficient range 0.25-0.88), satisfying good item measure correlation (Table 2).Table 2Item measure correlation and internal consistency of domainsCharacteristics of the Dor modelItem measure correlationInternal consistency coefficientFace validity Physical attributes a. Size of the model0.7750.883 b. Landmark visualization, angle of His0.4810.891 c. Landmark visualization, fundus size0.6230.887 Realism of materials a. Overall impression0.6450.886 b. Tissue quality0.2630.905 Realism of experience a. Amount of pressure needed to place a suture through the material0.6210.887 b. Realism of anatomy0.8830.878 c. Does the simulation represent real a fundoplication?0.6610.885 Ability to perform tasks a. Placement of suture0.2770.897 b. Closure of angle of His0.4040.894 c. Folding the fundus0.2540.899 d. Proper securing of fundus on the hiatus0.6780.884Content validity Value as a training tool a. As a training tool0.5670.889 b. As a testing tool0.6680.884 Relevance to practice0.6270.887Global0.7050.883 Table 2Item measure correlation and internal consistency of domainsCharacteristics of the Dor modelItem measure correlationInternal consistency coefficientFace validity Physical attributes a. Size of the model0.7750.883 b. Landmark visualization, angle of His0.4810.891 c. Landmark visualization, fundus size0.6230.887 Realism of materials a. Overall impression0.6450.886 b. Tissue quality0.2630.905 Realism of experience a. Amount of pressure needed to place a suture through the material0.6210.887 b. Realism of anatomy0.8830.878 c. Does the simulation represent real a fundoplication?0.6610.885 Ability to perform tasks a. Placement of suture0.2770.897 b. Closure of angle of His0.4040.894 c. Folding the fundus0.2540.899 d. Proper securing of fundus on the hiatus0.6780.884Content validity Value as a training tool a. As a training tool0.5670.889 b. As a testing tool0.6680.884 Relevance to practice0.6270.887Global0.7050.883 Item measure correlation and internal consistency of domains Internal consistency Cronbach's alpha demonstrated excellent internal consistency for all domains (range, 0.87-0.90) (Table 2). Confidence Surveys The survey response rate was 100%. Trainees reported higher confidence in performing the assigned task (P < .001) on the second attempt, with seniors reporting higher confidence levels than juniors (n = 28) (Table E2, Table E3, Table E4). Most trainees had difficulty comprehending the second and third stitches of the 3 stitch Dor techniques; however, seniors were more comfortable in executing the steps after watching the directed teaching video between the 2 attempts. Expert Feedback Experts found this to be an “[e]xcellent teaching tool to practice the steps of the procedure and review the concepts.” It “[h]elps to take out [the] learning curve from OR to outside.” However, they do not recommend this as a testing tool due to the low-moderate fidelity of the model. They reported that the model is good for Dor and Toupet fundoplication, but not for Nissen fundoplication, and needs more scale-adjusted modification and realism of experience. Trainee Feedback The overall simulation was rated as a median of 5 [4-5] by trainees. Trainees reported feeling more confident in operative tasks after the simulation. All of them wanted similar dry labs to be available to them (Table E5). Many have requested to use this model to practice before their thoracic surgery rotations. Discussion Principal Findings We have developed a medium-fidelity, low-cost surgical simulator for Dor fundoplication, which meets all the benchmarks of robust validation. It is easy to replicate and can be used in a preoperative training setting, mainly for junior trainees. Significance of Surgical Simulators With increased interest in surgical education research, simulator development has become a burning topic in graduate medical education. Many such endeavors are undertaken in different disciplines.6,9 Thoracic surgery is no different.10 Although high-fidelity cadaveric or live-anesthetized animal model wet labs are considered the gold standard for such simulations, they present their own challenges. Some training centers do not have the finances or facilities to have routine, easily available wet labs. Live-anesthetized animal labs also bring about an ethical dilemma. On the other hand, computerized robotic simulators have the advantage of providing statistical feedback, but tasks such as troubleshooting are better taught in dry and wet lab settings. Need for Dry Labs The main purpose of a simulator is to be a practice tool.11 There has been an increasing push for low to medium-fidelity dry labs as a repetitive practice platform for trainees to master operative steps before performing them for the first time in patients. This is to complement, not substitute, the wet labs. Cheap, readily available household materials are being used to facilitate easy replication of models and widespread adaptability.12 For example, tissue paper rolls have been used to construct a tracheal model for emergency cricothyroidectomy at a major tertiary care hospital.13 With that in mind, we created a fundoplication model from felt costing only $2 per model. Comparatively, commercial fundoplication models are available from SimuLab, but at a much higher cost.14 Instruction Delivery Low-fidelity simulations paired with an instructional video (Video 1) with auditory and text step-by-step instructions can have high retention and good educational value for junior residents.15 Residents demonstrated a better understanding and execution of tasks with the struggle-first teaching method compared with the teach-first method.16 Thus, we allowed the residents to struggle first with the first attempt before providing them with an instructional video. This was to assess the pretesting effect. Furthermore, when trainees practice on dry labs followed by a culminating wet lab training, they feel more prepared for complex surgeries.2 Learning Curve It is not uncommon for a trainee to have the opportunity to attempt a Dor fundoplication once or twice on a cadaver or a high-fidelity model, but they would rarely have multiple attempts. Repetition, particularly proficiency-based practice, is the key to successful operative performance.17,18 One study reported that the learning curve for Nissen fundoplication is more than 20 cases.19 Dry labs can help in reducing the steepness of the learning curve. By integrating monitored simulation into training programs, the number of cases required to master a complex operation can be reduced.4 We observed a decrease in scores among some trainees, particularly the junior cohort, who were less familiar with the procedure. One possible explanation is the cognitive load associated with learning and immediately applying a new technical skill. Because specimen grading was weighted heavily on time, trainees who were focused on executing the newly learned steps may have required more time, resulting in lower scores. In addition, during video grading, many trainees experienced difficulty with the placement of the second stitch, particularly without the assistance of a fourth robotic arm or an assistant, which may have affected their performance scores. Finally, trainees participated in the simulation outside of protected educational time and had to step away from clinical responsibilities to complete the sessions. In some instances, participants needed to return to clinical duties earlier than anticipated, which may have impacted performance. The majority of trainees who elected to participate in the additional practice session were medical students and junior trainees who were less familiar with the robotic platform. These practice sessions were formative in nature and were not recorded or formally scored. Given that the additional practice was brief and primarily focused on familiarization with the robotic interface rather than deliberate, structured skill refinement, we do not anticipate that it substantially influenced immediate postvideo performance. Strengths and Weaknesses We chose to incorporate the robotic platform in this study to provide additional educational value—particularly for junior trainees—by increasing familiarity with robotic instrumentation, console ergonomics, and suturing techniques. Nonetheless, the core technical steps can readily be adapted for use in an open setting or within a standard laparoscopic box trainer, allowing broader implementation even in resource-limited environments. Our study has several limitations regarding its interpretation and application. The first limitation is regarding the number of trainees recruited. Initially, 56 may appear as a small sample size, making the generalizability of the study limited. However, based on our power analysis, only 30 recruits were sufficient to show statistical significance. But because of the increasing popularity of the study among trainees, we almost doubled our sample size. We had to close the study and turn away interested subjects due to the timeline of the project. We were only able to recruit trainees from the Cleveland Clinic main campus, which might limit the diversity of the training backgrounds for volunteers. We also had trainees in our program who had previous experience of completing advanced postgraduate training elsewhere. Such trainees were grouped in the Senior class based on the number of years they were out from their medical school graduation. In our study design, we defined the junior (novice) group as postgraduate year (PGY) 3 and below to reflect the typical structure of training at our institution, where exposure to advanced minimally invasive and robotic procedures is generally limited before PGY 4. As such, trainees at the PGY 3 level and below have comparatively less experience with these techniques and were grouped to represent a novice cohort. The senior group, consisting of PGY 4 and above trainees as well as advanced MIS and thoracic surgery fellows, reflects individuals with substantially greater operative exposure, thereby providing an appropriate comparator group for assessing construct validity. At the time this study was conducted, objective operative performance data were not readily accessible. The intuitive platform–based evaluation application was not yet widely implemented in routine clinical assessment, and many trainees were not enrolled in the system. In addition, formal clinical evaluations constitute protected trainee information, and we did not obtain Institutional Review Board approval to access or analyze those data. Finally, several of the senior trainees involved in this study have since graduated, limiting our ability to retrospectively assess operative performance outcomes. We agree that assessing the impact of simulator training on operative performance represents an important next step, and this will be the focus of future studies. We had originally intended to include multiple assessors for the video specimen. However, this was not feasible due to a combination of logistical constraints, including the time-intensive nature of detailed video review, limited availability of trained expert raters familiar with the OSAT rubric, and institutional limitations related to data access and transfer during a transition period because the principal investigator and co-investigators relocated. The model also has certain limitations. Although the cost of production is extremely low ($2 per model), the material is not perfect for replicating the stomach. Experts have commented that the material is much stiffer than the stomach. The fundus on the original model was large, which was modified to scale on revision. The esophagus and fundus were stuffed to give them a 3-dimensional appearance. Mobilization of the fundus was not a focus of the simulation. The tool is better for juniors to learn the steps and practice compared with seniors. The model indeed needed some improvement but was deemed a good enough tool for teaching. Conclusions and Future Directions This model satisfies all 3 benchmark criteria required to be considered validated. The model is cheap and easily reproducible, and has all the major anatomic representations needed for training. It can be easily replicated in resource-limited settings, even by trainees themselves. The global findings and comments indicate that the simulator is valuable as a Dor fundoplication teaching tool. We are currently building a teaching curriculum with this model to train our surgical trainees to practice before coming to the operating room. Future prospective studies specifically designed to evaluate transfer validity in centers with high volumes of fundoplication procedures will be valuable. Conflict of Interest Statement The authors reported no conflicts of interest. The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they may have a conflict of interest. The editors and reviewers of this article have no conflicts of interest.