|Ahead of print publication
Three-year evaluation of a novel, nonfluoroscopic, all-artificial model for EUS-guided biliary drainage training for the impact to practice: A prospective observational study (with videos)
Tanyaporn Chantarojanasiri1, Aroon Siripun2, Pradermchai Kongkam3, Nonthalee Pausawasdi4, Thawee Ratanachu-ek5
1 Thai Association for Gastrointestinal Endoscopy; Department of Internal Medicine, Division of Gastroenterology, Rajavithi Hospital, Bangkok, Thailand
2 Thai Association for Gastrointestinal Endoscopy; Department of Gastroenterology, Bangkok Hospital, Bangkok, Thailand
3 Thai Association for Gastrointestinal Endoscopy; Department of Medicine, Division of Gastroenterology, Excellence Center for Gastrointestinal Endoscopy, Faculty of Medicine, King Chulalongkorn Memorial Hospital, Chulalongkorn University, Thai Red Cross Society; Division of Hospital and Ambulatory Medicine, Department of Medicine, Faculty of Medicine, Pancreas Research Unit, and Tropical Medicine Cluster, Chulalongkorn University, Bangkok, Thailand
4 Thai Association for Gastrointestinal Endoscopy; Department of Medicine, Division of Gastroenterology, Faculty of Medicine, Siriraj Hospital, Mahidol University, Bangkok, Thailand
5 Thai Association for Gastrointestinal Endoscopy; Digestive Endoscopy Training Center, Rajavithi Hospital, Bangkok, Thailand
|Date of Submission||25-Oct-2021|
|Date of Acceptance||24-Apr-2022|
|Date of Web Publication||05-Oct-2022|
1873, GI Endoscopy Unit, 10th Floor Bhumisirimunkalanusorn Building, King Chulalongkorn Memorial Hospital, Rama 4 Road, Patumwan - 10500, Bangkok
Source of Support: None, Conflict of Interest: None
Background and Objectives: EUS-guided biliary drainage (EUS-BD) required a dedicated training. We developed and evaluated a nonfluoroscopic, all-artificial training model known as Thai Association for Gastrointestinal Endoscopy Model 2 (TAGE-2) for the training of EUS-guided hepaticogastrostomy (EUS-HGS) and EUS-guided choledochoduodenostomy (EUS-CDS). We hypothesize that trainers and trainees would appreciate the ease of the nonfluoroscopy model and increase their confidence to start their real procedures in humans. Materials and Methods: We prospectively evaluated the TAGE-2 launched in two international EUS hands-on workshops and have followed trainees for 3 years to see long-term outcomes. After completing the training procedure, the participants answered questionnaires to assess their immediate satisfaction of the models in and also the impact of these models on their clinical practice 3 years after the workshop. Results: A total of 28 participants used the EUS-HGS model and 45 participants used the EUS-CDS model. The EUS-HGS model was rated as excellent by 60% of beginners and 40% by experienced and the EUS-CDS model was rated as excellent by 62.5% of beginners and 57.2% of experienced. The majority of trainees (85.7%) have started the EUS-BD procedure in humans without additional training in other models. Conclusion: Our nonfluoroscopic, all-artificial model for EUS-BD training is convenient to be used with good-to-excellent satisfaction scored by the participants in most aspects. It can help the majority of trainees start their procedures in humans without additional training in other models.
Keywords: biliary drainage, EUS, evaluation, training model, Thai Association for Gastrointestinal Endoscopy Model 2
|How to cite this URL:|
Chantarojanasiri T, Siripun A, Kongkam P, Pausawasdi N, Ratanachu-ek T. Three-year evaluation of a novel, nonfluoroscopic, all-artificial model for EUS-guided biliary drainage training for the impact to practice: A prospective observational study (with videos). Endosc Ultrasound [Epub ahead of print] [cited 2022 Nov 28]. Available from: http://www.eusjournal.com/preprintarticle.asp?id=357880
| Introduction|| |
EUS-guided biliary drainage (EUS-BD) has been accepted as an alternative treatment for biliary obstruction after unsuccessful ERCP. Compared to conventional percutaneous drainage, EUS-BD delivers a higher rate of success, fewer complications, and a decreased need for re-intervention. Both ERCP and EUS procedures require specialized skills to achieve maximal benefit and to minimize procedural-related adverse events., General recommendations for those who begin the EUS procedure usually start with regular EUS-FNA followed by EUS-guided pseudocyst drainage under supervision. Compared to pseudocyst drainage, EUS-BD required more complicated multiple steps with a higher complication rate. Even in a large training center, the number of cases using EUS-BD is very limited, reported between 0.6% and 3.3% after failed ERCP., As a result, interventional endosonographers have limited opportunities to practice this procedure during their training period. Many hands-on models have been developed to help the trainees practice EUS-BD in each step, including maneuvering the scope and instruments. To simulate the echogenicity of human internal organs, most of the models require a combination of ex vivo specimens with a synthetic frame. Despite providing good echoendoscopic imaging, the use of ex vivo specimens requires special preparation and storage and, most importantly, raises concerns of zoonotic transmission., These models also need the use of fluoroscopy, which can limit portability and increase the risk of radiation exposure.,
We developed a new, all-artificial EUS-BD model which does not require fluoroscopy for guidewire visualization. This model has been practiced and evaluated in several international hands-on workshops in Thailand, India, Korea, and Spain by experienced endosonographers and trainees. The aim of this study was to assess the satisfaction of the workshop participants who have used these models.
| Materials and Methods|| |
Artificial EUS hands-on models were initially developed in 2016 for local workshops in Thailand. The initial model (Thai Association for Gastrointestinal Endoscopy Model 1 [TAGE-1]) was composed of gelatin and cornstarch for liver parenchyma and three-dimensional (3D) printed resin to replicate the bile duct [Figure 1]a. This provided good echogenicity and cauterizing ability [Figure 1]b, but this gelatin-starch model was not durable and needed to be refrigerated. The 3D-printed resin was also very costly. Moreover, the guidewire was not fully visible due to the limitations of the model design.
|Figure 1: (a) The gelatin starch model of EUS hepaticogastrostomy and EUS choledochoduodenostomy; (b) On the EUS, the echogenicity of the gelatin model is similar to the liver. The resin tube submerse in water appears as a black tubular structure similar to the intrahepatic bile duct. A needle puncture into the synthetic bile duct is seen in this figure|
Click here to view
To overcome the limitations of the TAGE-1, we developed a model which comprised the floral polyurethane foam (floral foam, Oasis®, Smithers-Oasis, Malaysia) as a substitute for gelatin-starch, and glass as a substitute for resin. For usability purposes, the polyurethane foam was submerged in water to eliminate air bubbles and then cut and reshaped. This model, known as the TAGE-2, was released in 2018. This model was expected to provide greater durability and reshape more easily.
Liver and bile duct model
For the EUS-guided hepaticogastrostomy (EUS-HGS) model, we cut the polyurethane foam and submerged in the water for 5–7 days to generate echogenicity as human liver parenchyma. The polyurethane foam was placed in the transparent acrylic box in the similar alignment to the usual position during EUS-HGS [Figure 2]a. The synthetic bile duct was made by inserting a 6-mm plastic tube (straw) into the model to create a tunnel simulating the left intrahepatic duct in segments 2 and 3 which connected to a Y-shape glass tube that simulated the common hepatic, common bile duct, and ampulla [Figure 2]b. In the EUS image, the polyurethane foam submersed in water has an echogenicity similar to the liver. The tunnels, which were filled with water, appeared as a tubular hypoechoic structure similar to biliary trees [Figure 2]c.
|Figure 2: The (a) Schematic of EUS-guided hepaticogastrostomy (EUS HGS) and the scope direction; (b) Thai Association for Gastrointestinal Endoscopy Model 2 (TAGE-2) for EUS-guided hepaticogastrostomy; (c) Endosonographic image of the EUS-guided hepaticogastrostomy model demonstrates the visualization of the polyurethane foam submerse in water, and a needle penetrating a tubular structure simulating intrahepatic bile duct. B2: Segment 2 intrahepatic bile duct; B3: Segment 3 intrahepatic bile duct; EUS-FNA: EUS fine-needle aspiration|
Click here to view
For the EUS-guided choledochoduodenostomy (EUS-CDS) model, duodenal bulb and perihepatic region that was the target area for the procedure were simulated [Figure 3]a. We cut the polyurethane foam to simulate the perihilar area and created a tract to simulate the portal vein and common bile duct. The synthetic bile duct was made into segments that represent a common bile duct, which was connected to a glass tube representing a common hepatic duct and cystic duct [Figure 3]b. A plastic tube was inserted into the polyurethane frame and water was injected to simulate the portal vein flow [Figure 3]c.
|Figure 3: (a) Schematic of EUS-guided choledochoduodenostomy and the scope direction; (b) Thai Association for Gastrointestinal Endoscopy Model 2 for EUS-guided choledochoduodenostomy; (c) Endosonographic image of the EUS-guided choledochoduodenostomy model demonstrates the visualization of the polyurethane foam submerse in water, and a needle penetrating a tubular structure simulating a common bile duct. Water was injected through the plastic tube inserted into a tunnel to simulate portal vein with positive Doppler signal|
Click here to view
Upper gastrointestinal model
For the EUS-HGS model, the polyurethane foam was shaped to represent the left lobe liver and then embedded in an acrylic frame that simulated the upper gastrointestinal tract. A part of the stomach cardia was created by a plastic frame covered by spun-bond polypropylene sheath at the interface between the synthetic stomach and the left lobe liver.
For the EUS-CDS model, the polyurethane foam was shaped to represent the soft tissue in the hepatoduodenal region. The duodenal bulb part was created using a plastic frame covered with spun-bond polypropylene sheath at the duodenal and bile duct interface similar to the EUS-HGS model.
In the TAGE-2, more than half of the bile duct is visible. We used a video projector to visualize the guidewire and instrument instead of the fluoroscopy. After needle puncturing, diluted methylene blue was used to represent the contrast injection.
Steps for EUS-hepaticogastrostomy and EUS- choledochoduodenostomy
The steps for EUS-HGS are as follows: identification of desired bile duct (preferably segment 3 left intrahepatic bile duct), bile duct puncture using 19G EUS-FNA needle, contrast injection, guidewire insertion, EUS-FNA needle removal, dilator insertion (coaxial cautery dilator preferred), tract dilation (balloon or dilation catheter), and stent insertion [Video 1 [Additional file 1]].
The steps for EUS-CDS are as follows: identification of puncture site, bile duct puncture, contrast injection, guidewire insertion, EUS-FNA needle removal, dilator insertion (coaxial cautery dilator preferred), tract dilation (balloon or dilation catheter), and stent insertion [Video 2 [Additional file 2]].
All the steps mentioned above can be demonstrated using our model. Three types of dilators (coaxial dilators, dilation catheter, and balloon dilator) and two types of stents (double-pigtail plastic stent and fully cover metallic tubular stent) were provided.
The EUS-HGS and EUS-CDS models were set up in multiple stations, and the procedures were performed by the trainees under supervision by one trainer in each station. Prior to the hands-on session, an experienced endosonographer (“experienced”) who was responsible for each station performed the procedure by themselves as a demonstration to the trainee (“beginner”). Equipment such as coaxial cautery dilator, dilation catheter, balloon dilator, double-pigtail plastic stents, and tubular fully covered metallic stents were selected according to each endoscopist's decision. After the hands-on session, both the experienced and the beginner endosonographers answered the questionnaire to access the satisfaction in 10 aspects in each step of EUS-BD. Each evaluation aspect was scored from 1 (poor), 2 (fair), 3 (good), and 4 (excellent) and was calculated separately for beginner and experienced endosonographers.
To evaluate the impact of this training model in real clinical practice, we sent the questionnaire to the beginner participants who attended the workshop 3 years after the workshop. The questionnaire included their EUS-HGS and EUS-CDS experience after the workshop, their experience of attending other workshops, their confidence of performing the procedure under supervision, and their evaluation of the usefulness of this model using the 5-scale rate from “least useful” to “most useful.”
A beginner endosonographer is defined as who has performed EUS-FNA of <50 cases. An experienced endosonographer who supervised each station is defined as who has performed EUS-FNA of more than 500 cases and experienced in EUS-BD.
The evaluation score for each aspect was expressed with the median response of participants. The difference between the beginners and the experienced endosonographers was compared using Fisher's exact test. P < 0.05 was considered statistically significant.
| Results|| |
The EUS-HGS model was evaluated by 28 participants which included 23 beginners and 5 experienced endosonographers. The median evaluation score for each aspect is displayed in [Table 1]. The scores were similar for most aspects generally ranging from good (3) to excellent (4). In the dilation step, either coaxial cautery dilator, dilation catheter, or dilation balloon was used in each case depending on the operators' selection. For example, only the dilation catheter was used by the experienced endosonographers, while all types of instruments were used by both beginners and experienced endosonographers. The overall score did show a small decrease from excellent (4) among beginners to good among experienced endosonographers (3). An excellent median score was given for needle visibility, dilation balloon, plastic stent insertion, and the overall rating.
|Table 1: Median evaluation score proportion and percentage of TAGE-2 for EUS-guided hepaticogastrostomy|
Click here to view
The EUS-CDS model was evaluated by 45 participants which included 36 beginners and 9 experienced endosonographers. The results of the evaluation score for TAGE-2 for choledochoduodenostomy are presented in [Table 2]. The scoring from endoscopists from all levels of experience was similar in most aspects. Experienced evaluations did show some minor variation in several areas including a slightly lower rating of bile duct identification and needle visibility in EUS-HGS, and lower rating in contrast injection, tract dilation, and stenting in EUS-CDS. However, these differences were not statistically significant. On the other hand, a significant scoring difference was seen (P = 0.02) in fluoroscopic image in EUS-CDS despite the same median score implying the difference in the scoring distribution. For overall rating for satisfaction, all participants rated these models as good to excellent. The EUS-HGS model was rated as excellent by 60% of beginners and 40% by experienced and the EUS-CDS model was rated as excellent by 62.5% of beginners and 57.2% of experienced. For the long-term evaluation, overall 17 participants in EUS-CDS model and 17 participants in EUS-HGS model answered the questionnaire. The result of the questionnaire is described in [Table 3]. The data from [Table 3] illustrate the long-term outcomes of trained trainees. It shows that around 41.2% of the trainees were able to actually perform the EUS-BD procedure after training through the workshop by this model and most of these trainees (85.7%), 6/7, can actually perform this procedure within 6 months after receiving the training without the need for additional training in other models. Most of the respondents showed that training by this model was very helpful for their practice.
|Table 2: Median evaluation score proportion and percentage of TAGE-2 for EUS-guided choledochoduodenostomy|
Click here to view
| Discussion|| |
We reported the evaluation of a novel, nonfluoroscopic, all-artificial model for EUS-BD made from polyurethane foam (floral foam). The model was built to be partly transparent to use the projector to represent the fluoroscopic image. Since it contains only synthetic materials, there is no concern about zoonotic transmission. The model itself is inexpensive and easy to transport and provides good practice for the endoscopic steps of EUS-BD.
Appropriate material representing soft tissue and liver parenchyma in these models should have similar echogenicity to the liver including firm consistency that can be punctured and dilated several times and easy to prepare. Ideally, the model should be more transparent, so that the projector can visualize the instrument as occurs in the real setting. We currently could not find material that possesses all these qualities. Polyurethanes are a copolymer that is synthesized from a polyol, a di- or multi-isocyanate, and a chain extender. It has been used as a scaffold for drug delivery and as a component in many industrial products such as the material for heat insulating, roofing, sealing and as the component of household furniture. The mechanical properties of polyurethane foam could be adjusted depending on the composition of hard and soft components so that it can be adjusted to mimic various organs such as bone or soft tissues. Polyurethane foam has also been used in the floral industry for its ability to retain water due to its porous structure and can prolong the flower's life by increasing water diffusion through the stem. After it has been submerged in water and saturated, it provides good echogenicity similar to the liver.
EUS-BD consists of many steps starting from selection of puncture site, guidewire manipulation, tract dilation, and stent placement. Each step needs skill, knowledge of endoscopic technique, and instrument manipulation which can be improved by practice in hands-on model. Many in vivo and in vitro models have been developed for practicing EUS-FNA,,, but not until 2015 was the ex vivo model for EUS-BD model launched. The initial model used porcine tissue to represent liver parenchyma and a 3D-printed polycarbonate tube to represent the bile duct. This model, the so-called “Mumbai EUS,” provided good radiographic and endosonographic images, but the need for fluoroscopy made it inconvenient for setting up outside of a medical institute. Moreover, the model needed to be freshly prepared just before the hands-on session. The use of porcine tissue also raised concerns of zoonotic transmission.
To evaluate our TAGE-2 quality, we launched it in 2018 during the international EUS hands-on workshop which included international EUS trainees and trainers. Our models received good-to-excellent scores in terms of visualization and maneuverability, especially providing good sensation for puncture and tract dilation. Most of the evaluation scores were similar between the beginner and the experienced. After the workshop, we sent the questionnaire regarding the usefulness of this model in real clinical practice. The interval between the workshop and the questionnaire was 3 years, which is considered long enough for the clinical evaluation. After the workshop, more than half of our participants did not perform the EUS-HGS or EUS-CDS after the training. For those who perform the procedure, the first procedure was performed within 6 months and majority of them started the procedure without additional training workshop. According to the evaluation, the model did provide more confidence in case they need to perform the procedure under supervision and most of them found these models useful in their clinical practice.
Our models provided a fully synthetic apparatus that is inexpensive, portable and does not require fluoroscopy. The current training model does have several drawbacks. First, the model preparation process is quite complicated and the lack of gastric or duodenal parts makes it less realistic. While our evaluation scores were generally good to excellent, our sample consisted primarily of less experienced endosonographers. Input from this initial evaluation and other tests will help guide an improved design incorporating materials that can better replicate the anatomy and allow enhanced visualization of the entire process. Although we do not have the direct comparison between this model with other available models, this model provides almost all steps required in the EUS-BD which is different from other models. Furthermore, our model is all-artificial and does not require fluoroscopy which is user-friendly for the training setup.
| Conclusion|| |
In summary, we presented a novel, nonfluoroscopic, all-artificial model for EUS-HGS and EUS-CDS which can be considered one of the tools for the training process for EUS-BD.
Financial support and sponsorship
Conflicts of interest
Pradermchai Kongkam, Nonthalee Pausawasdi and Thawee Ratanachu-ek are Editorial Board Members of the journal. This article was subject to the journal's standard procedures, with peer review handled independently of the editors and their research groups.
| References|| |
Teoh AY, Dhir V, Kida M, et al.
Consensus guidelines on the optimal management in interventional EUS procedures: Results from the Asian EUS group RAND/UCLA expert panel. Gut
Sharaiha RZ, Khan MA, Kamal F, et al.
Efficacy and safety of EUS-guided biliary drainage in comparison with percutaneous biliary drainage when ERCP fails: A systematic review and meta-analysis. Gastrointest Endosc
Ekkelenkamp VE, Koch AD, Rauws EA, et al
. Competence development in ERCP: The learning curve of novice trainees. Endoscopy
Wani S, Coté GA, Keswani R, et al.
Learning curves for EUS by using cumulative sum analysis: Implications for American Society for Gastrointestinal Endoscopy recommendations for training. Gastrointest Endosc
Saumoy M, Kahaleh M. Safety and complications of interventional endoscopic ultrasound. Clin Endosc
Ryozawa S, Fujita N, Irisawa A, et al
. Current status of interventional endoscopic ultrasound. Dig Endosc
Tonozuka R, Itoi T, Tsuchiya T, et al
. EUS-guided biliary drainage is infrequently used even in high-volume centers of interventional EUS. Gastrointest Endosc
Holt BA, Hawes R, Hasan M, et al.
Biliary drainage: Role of EUS guidance. Gastrointest Endosc
Hoshi K, Irisawa A, Shibukawa G, et al.
Validation of a realistic, simple, and inexpensive EUS-FNA training model using isolated porcine stomach. Endosc Int Open
Artifon EL, Cheng S, Nakadomari T, et al.
Ex vivos models to teaching therapeutic endoscopic ultrasound (T-EUS). Rev Gastroenterol Peru
Dhir V, Itoi T, Fockens P, et al.
Novel ex vivo
model for hands-on teaching of and training in EUS-guided biliary drainage: Creation of “Mumbai EUS” stereolithography/3D printing bile duct prototype (with videos). Gastrointest Endosc
Dhir V, Itoi T, Pausawasdi N, et al.
Evaluation of a novel, hybrid model (Mumbai EUS II) for stepwise teaching and training in EUS-guided biliary drainage and rendezvous procedures. Endosc Int Open
Chiono V, Sartori S, Calzone S, et al
. 6 – Synthetic biodegradable medical polyurethanes. In: Zhang X, editor. Science and Principles of Biodegradable and Bioresorbable Medical Polymers. Sawston, Cambridge: Woodhead Publishing; 2017. p. 189-216.
Nakai Y, Kogure H, Isayama H, et al. Endoscopic ultrasound-guided biliary drainage for unresectable hilar malignant biliary obstruction. Clin Endosc
Sorbi D, Vazquez-Sequeiros E, Wiersema MJ. A simple phantom for learning EUS-guided FNA. Gastrointest Endosc
Fritscher-Ravens A, Cuming T, Dhar S, et al.
Endoscopic ultrasound-guided fine needle aspiration training: Evaluation of a new porcine lymphadenopathy model for in vivo
hands-on teaching and training, and review of the literature. Endoscopy
Gonzalez JM, Cohen J, Gromski MA, et al
. Learning curve for endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) of pancreatic lesions in a novel ex-vivo
simulation model. Endosc Int Open
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3]