Basic Requirements of Gastroenterologists to Treat Upper Gastrointestinal Bleeding: Competency and Sedation Issues

This article reviews the components of adequate training required for a gastroenterologist to treat upper gastrointestinal bleeding (UGIB). The current status of endoscopic simulators is critically reviewed to determine whether these should be part of the UGIB armamentarium in the training of individuals and whether credentialing could be accomplished through this method of instruction. Finally, the author discusses the appropriate use of sedation in patients with UGIB.

Endoscopic skills are generally developed and honed in the endoscopy suite. The hands-on training that a gastroenterology trainee receives is based on several factors: the skill level of supervisors, the quality of mentors, the patient population, and the trainee’s ability to meld the technical and cognitive aspects of training into a level of acceptable competence. Occasionally, the trainee’s hands-on time may be limited because of the difficulty of the procedure, financial implications of increased procedure time, patient discomfort, patient preference, or the urgency of the case. These challenging aspects are eliminated with the use of simulators.

Simulator training is an important aspect in other industries such as the airline industry, where a pilot goes through hours of training on simulators as well as flying alongside as a copilot. In medicine, the majority of training occurs “on the job.” This approach usually has implications that can involve patient care and even safety.

Upper gastrointestinal endoscopy at a time of gastrointestinal bleeding has many associated factors that occur before or during the endoscopy that amplify the difficulty of the training environment. Cardiopulmonary instability, aspiration, and hepatic encephalopathy are but a few hurdles. Complications such as perforation or worsening the bleeding may also arise during treatment of bleeding lesions. The risk of rebleeding without treatment may be high. Therefore, it is extremely important that the trainees are adequately trained and competent in treating upper gastrointestinal bleeding. All training programs require that a supervising physician be present for cases that involve fellows, but the comfort level of teaching physicians varies drastically, as does the technical expertise of the fellow being trained. Endoscopic simulators may resolve many of these issues if a trainee is allowed to learn the equipment and techniques of hemostasis under a calm, controlled environment, consequently becoming better equipped to handle the real case.

Endoscopic hemostasis is one of the most challenging and sometimes nerve-racking techniques we perform as gastroenterologists, so fellows should be better prepared to handle this usually high-stress and difficult procedure by training on a simulator prior to the real-life case. There are many in vivo, ex vivo, and mechanical simulators available for the training of fellows, some of which are specifically designed to teach techniques of hemostasis.

The earliest endoluminal gastrointestinal endoscopic simulators were first used in the late 1960s and early 1970s. The early plastic mechanical simulators, such as the Erlangen plastic mannequin, were used to train for upper endoscopic examination with a flexible endoscope. Although the realism of tissue elasticity and resistance is difficult to simulate in mechanical models, their simplicity can be useful in training novices. Several different plastic mechanical simulators can still be purchased. Live animal models, also known as in vivo models, can provide the realism that mechanical simulators cannot, but the orientation of organs may sometimes be different. Juvenile pigs are considered ideal for endoscopy simulation. The drawbacks of using animal models include ethical concerns, expense, and the infrastructure and resources required for storage, setup, and disposal.

Hybrids of these two types of models are the composite and explanted animal organ simulators, also known as ex vivo models, which are made using both plastic parts and animal organs. The Erlangen Active Simulator for Interventional Endoscopy (EASIE) was the first model to have a realistic depiction of spurting blood for therapeutic endoscopy training. There are also more lightweight models called the Erlangen compactEASIE and the Endo X Trainer (Medical Innovations International, Rochester, MN, USA). Similar to the live animal models, the composite and explanted animal organ simulators give the trainee a more realistic feel than mechanical simulators, but there is a lengthier preparation time and the tissue must be properly disposed of and/or replaced after a limited number of uses. In comparing the in vivo and ex vivo models, the in vivo model provides better tactile simulation but the ex vivo models do not tend to have organ orientation issues.

The most complex models are the virtual-reality or computerized simulators, which use a videographic tool technology that displays computerized images based on the endoscope’s real-time movement by the user. Computerized models can offer the greatest variety of training opportunities. In comparison with the in vivo or ex vivo simulators, the computerized simulators eliminate the lengthier setup and do not require regular replacement/disposal of animal organs. Computerized simulators currently available on the market include the GI-Bronch Mentor (Simbionix, Cleveland, OH, USA) and the CAE Healthcare Accutouch (CAE Healthcare, Montreal, Canada). In addition, other prototypes exist that are not yet on the market.

On a wheeled trolley, the GI-Bronch Mentor is a plastic mannequin with a mouth and nose for upper endoscopy or bronchoscopy and an anus for lower endoscopy. The mannequin is equipped with sensors for haptic feedback to the user and a liquid-crystal display (LCD) screen on a movable arm. The tip of the endoscope contains sensors as opposed to an actual lens. The sensors generate a dynamic endoscopic view based on movement by the user. Multiple modules are available for this model with different levels of difficulty of anatomy and procedure complexity, including endoscopic ultrasonography. The model also records an evaluation of the user’s performance.

The CAE Healthcare Accutouch is similar to the GI-Bronch Mentor in terms of the sensors that provide feedback and the LCD on a movable arm, all on a trolley with wheels. However, this model also simulates a patient’s vital signs as well as responses to the administration of sedation and to pain.

Training

It is important for fellows to learn all the available tools in the treatment armamentarium of upper gastrointestinal bleeding. The skills required for endoscopic hemostasis are injection, coagulation, and hemoclip application.

There is ample evidence that suggests combination therapy (epinephrine injection followed by coaptive coagulation or hemoclip placement) for treatment of upper gastrointestinal bleeding from ulcers has a better outcome than single modality. Therefore, treatment usually involves injection therapy followed by coaptive coagulation or hemoclip placement. Multipolar probes include the Microvasive Gold Probe (Boston Scientific, Natick, MA, USA), the BICAP (Circon ACMI, Stamford, CN, USA), and the Heater Probe (Olympus Corporation, Lake Success, NY, USA). Multipolar probes have alternating arrays of negative and positive electrodes at the tip, through which electricity passes and causes heating of the contacted tissue, therefore achieving hemostasis. The benefit of multipolar probes is that once tissue has been desiccated, its resistance to further coagulation is increased. Consequently, it avoids deep tissue injury and perforation. While the Heater Probe transfers heat across the ceramic tip, it can cause deeper penetration of the tissue and is not limited by tissue desiccation, water, or resistance, causing the risks of perforation to be higher. Therefore, it is very important to know that the depth of coagulation is increased when using slower coagulation, lower energy, larger probes, and tamponade by making a firm contact with the tissue being treated. Most of these coagulation methods combine injection of dilute epinephrine (1:10,000 or 1:20,000 diluted in saline) before thermal treatment. Another important factor in the treatment of a bleeding lesion is to have adequate visualization during treatment by washing the lesion. There is a device that combines injection with a multipolar probe (Microvasive Injection Gold Probe; Boston Scientific) and allows for treatment of lesions by an injection needle that is used to inject epinephrine, which then retracts to allow the multipolar probe to coagulate the lesion. Another advantage of this device is that it has an irrigation port, which allows washing of the lesion to assess for adequate treatment effect and avoids pulling off the eschar formed by coagulation.

There are various hemoclips commercially available, including QuickClip2 (Olympus Corporation), Resolution clip (Microvasive Endoscopy, Boston Scientific) and TriClip (Cook Medical, Bloomington, IN, USA). These devices are preloaded inside a catheter that is advanced to the lesion and positioned by opening the hemoclip, closing, and deploying onto the bleeding stigma in the ulcer base. TriClip also has an irrigation port that allows washing of the lesion before clip placement. The Resolution clip has reopening capability prior to deployment, which may allow for better placement of the clip.

As it is important for fellows to become familiar with these devices, training via simulators is safer than real-life practice. As more sophisticated simulators are being developed and used, the next logical question to ask is, “is this training beneficial?”

Based on several reported studies that have looked at computer simulators and their impact, training appears to have a positive impact on technical skills of gastroenterology fellows. Some studies have also reported that computer simulators used for upper endoscopy have been able to differentiate between a novice and expert endoscopist. One of the studies conducted used the compactEASIE, and involved 37 gastroenterology fellows from 9 training programs in New York City. It was a randomized controlled study comparing effects of intensive 7-month, hands-on training in endoscopic hemostatic techniques. The techniques evaluated included manual skills, injection and electrocoagulation, hemoclip application, and variceal ligation. Twenty-eight fellows were randomized into group A, which received purely clinical training at their own hospitals, and the same number into group B, trained by experienced tutors over 3 full-day workshops over a 7 month period in addition to the training received at their own hospital. Baseline endoscopic experience of the fellows in the two groups was similar. The tutors and an evaluator, who was blinded to the method of training, performed the final evaluations. Blinded evaluation results of the study of 10 of 14 fellows in group A and 13 of 14 fellows in group B revealed that the only statistically significant skill improvement was for hemoclip application in group B, 1.0 versus 7.6 ( P <.001). If unblinded tutor evaluation was included in the analysis, the injection and electrocoagulation skill was also improved statistically in group B compared with group A. Furthermore, evaluation of performance of endoscopic hemostatic procedures performed at their home institution during the study period revealed that group B fellows achieved hemostasis at a rate of 100% versus group A at a median rate of 87%. A median complication rate for fellows in group A was 11% versus 0% in group B.

Another study of general upper endoscopy training enrolled 28 internal medicine residents who had not yet received endoscopy training. This study did not evaluate colonoscopies or the control of bleeding or hemostasis. Only patients who would not be undergoing sedation were selected, and were not told about the trainee’s training status. The residents were randomly split into two groups: Group C, who received only conventional training, and Group S, who trained on a simulator prior to conventional training. The study considered the first 10 endoscopies for patient discomfort, the time it took to get to the duodenum, technical accuracy, and total endoscopy time. Seven residents from each group continued their endoscopy training and were reevaluated at this advanced training stage on another 10 endoscopies after completing 50 successful endoscopies. The GI Mentor virtual endoscopy simulator was used, which is a gastrointestinal mannequin through which an endoscope with sensors on the tip and shaft provide 3-dimensional real-time pictures. There is also a force-feedback module that creates a realistic feeling of resistance when the walls of the gastrointestinal tract are touched.

Comparisons were made between the groups during the first 10 endoscopies and the advanced training stage, the 51st through 60th endoscopies. Patient discomfort was determined by an evaluation completed by the patients, who were not aware of whether or not the trainee had received simulator training. There did not seem to be a major difference in patient discomfort between the groups for the either set of 10 endoscopies evaluated. As one would expect, patients’ comfort levels did improve as the endoscopists performed more procedures regardless of whether or not they trained on a simulator. There was a significant difference in the time it took to get to the duodenum between the groups, both at the initial 10 endoscopies and at the advanced training stage. Group S, the group who had received simulator training in addition to conventional training, was much faster than Group C, who had only the conventional training in both instances. As for technical accuracy, there was a difference in the first 10 endoscopies between the groups but this difference was no longer evident at the advanced training stage. Group S required less assistance, had a better intubation ratio, and found pyloric passage and retroflexion of the endoscope to be easier. Similar to patient discomfort, technical accuracy improved as the endoscopists performed more procedures. However, Group S still showed significantly better technical accuracy at the advanced training stage. Similar to the technical accuracy, the total endoscopy time was significantly better for Group S at the initial 10 endoscopies as well as the advanced training stage. Again, as one would expect, the times improved as the endoscopists performed more procedures. The significant differences between the groups at the advanced training stage showed that there is a positive long-term impact to having had virtual simulator training in addition to conventional training.

Another study looked at a longer supervised training period of up to 160 investigations, and showed an even longer positive impact of having the simulator training.

A randomized, blinded, controlled study was conducted with 28 eligible participants to determine the impact of knowledge-based teaching in comparison with skills-based training in 4 therapeutic endoscopic procedures: control of nonvariceal upper gastrointestinal bleeding, polypectomy, stricture dilation, and percutaneous endoscopic gastrostomy (PEG) tube insertion. Group 1 received 4 40-minute lectures and then underwent an initial assessment, whereas Group 2 underwent the initial assessment first and then received the same 4 40-minute lectures. Both groups were then split into subjects and controls. Controls took a reassessment without further training; whereas the subjects received one half-day (4 30-minute sessions) of hands-on training first and then took a reassessment. The knowledge-based assessment was evaluated using a multiple-choice questionnaire. An expert observer and endoscopy assistant, both blinded to the group allocations, assessed the procedural skills on the simulators (a modified mechanical upper gastrointestinal phantom and ex vivo models).

The effect of lectures before or after the initial assessment of the two groups was not significant. The effects of hands-on training showed a significant difference between subjects and controls as well as between the initial baseline assessments and reassessments for 3 of the 4 procedures. Insertion of a PEG tube showed no significant difference. Therefore, the study concluded that hands-on training does significantly improve execution of controlling nonvariceal upper gastrointestinal bleeding, polypectomy, and stricture dilation.

Another randomized controlled trial involving 37 novice gastrointestinal fellows concluded that the skills of the subject group that received intense training on hemostatic techniques via simulator improved significantly over the baseline. In comparison with the control group, the area with significant difference was hemoclip application. A randomized clinical trial of 22 novice gastrointestinal fellows concluded that those trained on the simulator required less assistance and were able to perform more complete examinations.

Sedation and Gastrointestinal Training

Little research has been done on the topic of sedation in training for endoscopic procedures. Guidelines for the didactic practical aspects of training have been outlined but these have not been validated, and training under an anesthesiologist has not been required for sedation training in gastroenterology fellowship programs. The training and sedation training guidelines by the American Society of Gastrointestinal Endoscopy (ASGE) recommend that fellows should learn to provide patients with adequate preprocedure sedation education. One should obtain appropriate preprocedural history and conduct a physical examination. Sedation continuum must involve management of deeper unintended levels of sedation. All of this must be accomplished by acquiring thorough knowledge of pharmacology, cardiopulmonary physiology, and pharyngeal anatomy. Airway management skills must involve learning head tilt-jaw thrust maneuver, nasopharyngeal airway, oropharyngeal airway, and bag-mask ventilation. Physiologic monitoring must comprise interpretation of pulse oximetry, electrocardiography, and capnography, along with appropriate use of oxygen. The Human Patient Simulator (METI Inc, Sarasota, FL, USA) is a simulator specifically designed for training in anesthesia, and respiratory and critical care. This simulator has pupils that dilate and constrict in response to light, thumbs that twitch in response to a peripheral nerve simulator, automatic recognition in response to administered drugs and drug dosages, variable lung compliance and airways resistance, and automatic response to needle decompression of a tension pneumothorax, chest tube drainage, and pericardiocentesis, but there is no dedicated sedation module available with this simulator. Major drawbacks of simulator training for gastrointestinal sedation are expense, inadequate facilities, and lack of faculty teachers and technicians. Major barriers were identified as lack of free time for training, lack of training opportunities, and financial consequences of missing work. Sedation training in endoscopy needs to involve didactic training, airway workshops, simulator training, preceptorship, competency-based instruction, and maintenance of competency. There are many challenges and unanswered questions in sedation training such as high cost, development of competency-based education, and who will lead the education (anesthesiology vs gastroenterology).

Only gold members can continue reading. Log In or Register to continue