Development and application of a swine model for training ultrasonography-guided central venous access

Introduction: simulation based teaching is a powerful tool in medical education, allowing hands on practice under a controlled environment and with repeated maneuvers. Central venous access venipuncture is one of the most frequent procedures carried out in the hospital setting, due to its various clinical indications and, when performed with the help of ultrasonography, the risk of adverse events is minimized. Aim: to develop, to describe and to test a porcine model that simulates the central venous access puncture aided by ultrasonography. Method: a low cost porcine model was developed to train medical students and residents on central venous access guided by ultrasonography. Both students and medical residents underwent a theoretical training regarding the model, followed by a hands-on training session. Afterwards, the participants assessed the model by answering a questionnaire. Results: there were 51 participants. The average score regarding the similarity between the model and the human anatomy was 9.15. When the characteristics were separately assessed, the mean scores regarding the similarity of the vessels, anatomic disposition and ultrasonographic characteristics as well as the venipuncture were, respectively, 9.27; 9.31; 9.54 and 8.86. Conclusion: The model was approved and considered appropriate for the training of central venous venipuncture by all the participants. Furthermore, it is a low cost, simple and reproducible model, that presents high similarity with the human anatomy. Therefore, it may be used as an aid to train people on ultrasonography guided central venous access.


INTRODUCTION
S imulation-based teaching has become popular in training the skills of professionals in several areas, being a powerful learning tool in medicine. Simulation allows the practice of procedures in a controlled environment, in which error becomes an opportunity to improve learning, conferring greater autonomy to students and reducing risks to the patient [1][2][3][4] .
Central venous access is one of the most commonly performed medical procedures in clinical practice, approximately 200,000 of these procedures being performed annually in England 5 , and five million in the USA 6 . There indications for a central venous access are infusion of medications, parenteral nutrition, dialysis therapy, invasive monitoring in critically ill patients, and others 7,8 . Despite being considered a low complexity procedure that every doctor must be qualified to perform, it is not free of complications, the most common being infections associated with catheters, arterial puncture, pneumothorax, hematomas and venous thrombosis.
The number of punctures to obtain success is directly proportional to the rate of complications [7][8][9] , whose incidence increases six-fold when three or more attempts are required 10 . As for the puncture sites, there is a preference for the internal jugular and subclavian veins, due to the higher rates of complications of central venous access in the femoral vein 7 .
Traditionally, central venous access has always been performed considering anatomical references.
However, in the last 20 years, many studies have shown the benefits of using ultrasonography to guide the venipuncture. When aided by the ultrasonography, the puncture is performed with visualization of the vessels through the sonographic image, reducing the number of necessary punctures and the complication rates [5][6][7][8][9][10][11][12][13] . This fact is observed independently of the chosen puncture site (internal jugular, subclavian or femoral), although most of the studies are for internal jugular vein 11  Another factor that helps to reduce the number of complications is training and experience with the procedure. Professionals who have performed more than 50 central venous access have lower rates of complications than those who carried out less 6,14 .
Thus, training in simulators is an attractive modality to guarantee dexterity to the professional before exposure to  to apply it to students, and to assess teaching of the procedure and the resemblance to reality when using it.

METHODS
The project was developed in the State

Porcine model
We used porcine parts in the manufacturing of the models. We acquired them in a meat-marketing establishment duly regulated by the competent health surveillance institutions, and they were composed of skin, subcutaneous tissue and muscle of the animal's abdominal region (pancetta). Each model was made with two parts, placed on top of each other, fixed to a rigid wooden surface, and interspersed with a Penrose drain number two (mean diameter 12mm) for simulation of the internal jugular vein, and a 20 french Kehr drain for simulation of the common carotid artery. We sought to maintain the same anatomical relationship between the in vivo structures -internal jugular vein lateral and slightly anterior to the common carotid artery ( Figure 1).
The Kehr drain was filled with a red dye artificially colored solution with for blood simulation, connected to a two-way connector equipment and coupled to a 10 mL syringe, which simulated arterial pulse by means of manual, intermittent pushing of the  plunger. The Penrose drain was also connected to a twoway connector equipment, one of the ways was coupled to a 60 mL syringe, and the other, to a blood transfusion equipment and a 1,000 mL saline bottle. All the system was filled with the same solution for blood simulation.
We used the 60 mL syringe to suck the drain solution for simulation of hypovolemic states and also to retrieve any air bubbles, while the connection to the bottle of solution was needed for replacing the volume lost by puncture holes, keeping the drain always full. In this way, we could use the same drain several times. (Figure 2).
The HUOP Laboratory of Medical Skills provided all materials used but the porcine parts. There were no incurred costs, since they were materials with validity dates expired for use in clinical practice.

RESULTS
The research had 51 participants. Of these, 43 were medical graduates, and eight, resident doctors. As prior experience, most of the participants had performed between zero and less than five central accesses, and only eight had experience in performing the procedure with ultrasonography (Table 1).

Data analysis
We loaded the data to Microsoft Excel® spreadsheets and analyzed them by means and percentage.  and reduces risks to patients, in addition to providing the correction of technical errors in a controlled environment, which allows for multiple iterations [1][2][3][4] . Moreover, that have been major advances in the development of models that simulate situations of actual medical practice 21 .
Even so, many teaching centers are still based on the old methodology, mainly due to the lack of resources 6,20 . One way to get around this issue is the development of lowcost, handmade simulators, as described in this work.
We developed a low-cost, simulation model of  (Table 2). Considering only those participants who had experience with the eco-guided procedure, the average score of the overall similarity was 8.62, and the similarity scores of the vessels image, anatomical arrangement of structures, sonographic characteristics, and puncture were, respectively, 9.25, 9.00, 9.25, and 8.25 (Table 2).
All participants approved the use of the model in learning eco-guided central access puncture and judged the training in the simulated experimental model as useful before performing the procedure in patients.

DISCUSSION
Classically, learning of procedures was based on the teaching model "see one, do one, teach one".
However, in recent years, there has been increasing need of prior training in simulators before acting in real clinical situations 6,20 . This boosts the confidence of practitioners had mean reality similarity scores below 70 15 . Other models described also used chicken parts, vascular prostheses 14 or ballistic gelatin as a base 18 , but the similarity with reality was not evaluated.