Bibliographic record
Abstract
simulation, trauma: simulation, traumaMonitors that aren't visible to the team member managing the trauma patient's airway. A porter coming from the blood bank who isn't where the ED nurse expects him to be. Scalpels, chest tubes, and central lines that are out of place. In the chaotic, high-stakes environment of the emergency department, seemingly minor miscues like these aren't just inconveniences, they are latent safety threats (LSTs), increasing the potential for errors in patient care and putting unnecessary stress on the clinical team. Simulated emergency cases practiced in the environment of a simulation lab can help clinicians hone procedural skills, but they don't replicate the actual environment of your ED and its unique systems and structures. “The sim lab is great for practice—you can't ‘shoot free throws’ over and over in a workplace environment,” said Andrew Petrosoniak, MD, an emergency physician and the trauma team leader at St. Michael's Hospital and an assistant professor of medicine at the University of Toronto. “But if your goal is to figure out how you work within a space, how you optimally design it, and identify and overcome LSTs that lie beneath the surface, you can't figure that out very well in a sim lab,” he said. “High-performance ED and trauma teams are typically not making mistakes because they don't know things, but because of systemic problems: They don't have the right equipment, it's not optimized ergonomically, the workspace is not designed properly, and so on. These things contribute to poor outcomes and are often incorrectly attributed to human error.” In situ simulation—mock emergency cases staged in the actual ED or trauma bay, with a full on-site team—has been used for more than a decade to address safety threats and optimize teamwork training in the ED, and the literature suggests that it is successful. (BMJ Qual Saf 2013;22[6]:468; http://bit.ly/2MMYMFE.) Now, Dr. Petrosoniak and his colleagues at St. Michael's have developed a fine-tuned approach to in situ simulation called TRUST (trauma resuscitation using in situ simulation team training), which employs several unique strategies to optimize the benefits realized from each exercise. (BMJ Open 2016;6[11]:e013683; http://bit.ly/2Ms0teJ.) Each simulation using the TRUST model is recorded using multiple GoPro cameras placed at different angles. The resulting footage is analyzed not only by the clinical team but also by human factors experts who use a framework analysis methodology to review and code the videos. The team also developed a tracing tool—much like you might see football commentators using to trace the movements of the offensive line—that documents the movements of the ED team throughout the simulation. (BMJ Simulation and Technology Enhanced Learning 16 March 2018; http://bit.ly/2BbZFGe.) “All this may seem simple, but so far, there has been nothing in health care like it,” said Dr. Petrosoniak. “All of this put together with our method of analysis allows us to identify issues, quantify them more precisely, and trace them back to their origin.” Safety Problems After publishing their study protocol for TRUST, Dr. Petrosoniak and his colleagues reported on the findings from their series of 12 unannounced in situ simulations. (CJEM 2018;20[1]:132.) They identified 893 LSTs from those 12 simulations, which the clinicians and human factors experts assigned to eight themes, subcategorized into 43 codes. The themes included situational awareness, staff safety, mental model alignment, team and individual responsibility, team resources, equipment considerations, workplace environment, and clinical protocols. “A perfect example of how the TRUST system worked to identify LSTs is within our massive transfusion protocol that rapidly delivers blood to the ED or trauma bay,” Dr. Petrosoniak said. “Most institutions have them, and just like any protocol they make things more efficient, but they can also be clunky depending on how well they are thought out.” The simulation identified, for example, that nurses had to make two phone calls to launch the protocol—one to the blood bank and one to locating who would call a porter to come to the trauma bay, pick up the blood requisition, and go to the blood bank to retrieve it. “Ideally, you should be making one phone call,” Dr. Petrosoniak said. “So we shifted to a one-call system in which the nurse calls locating the porter, and then is automatically forwarded to the blood bank. The first time we ran that protocol, the TRUST system identified another problem: the call dropped. We found that you couldn't do a forward on the phone lines in our hospital, so we had the system fixed to allow that.” Fixes like these reduced the overall time to blood delivery in real trauma patients by 2.5 minutes, he said. “It was taking us too long before—probably about 11 minutes from the time of activation. Now, we're at an average of nine minutes, which is the recommended time, and many cases are less than that. It's a big win.” Common and Dangerous Michael Lauria, MD, a critical care flight paramedic and former U.S. Air Force pararescueman who is now a resident in emergency medicine at the University of New Mexico, said TRUST is the first in situ simulation system he has seen that involves a system for prioritizing and rating the LSTs by how common and how dangerous they are. “To me, that's particularly valuable because practically speaking, when you do these simulations, you can generate a long list of issues that may be dangerous or problematic, and you can't address it all at once,” said Dr. Lauria, who writes frequently about training to improve emergency personnel performance under stress. “By using this system, you can develop a more quantitative way of ranking those threats. In addition, involving people who are trained in human factors and systems processes is very useful. Even if you study human factors as a clinician, you're not going to be as adept in some of these methodologies as engineers with advanced degrees in human factors science.” In situ simulations like TRUST hone in on the differences between what these human factors experts would describe as “work as imagined vs. work as done,” said Mary Patterson, MD, MEd, an associate dean and the Lou Oberndorf Professor in Healthcare Technology in the Center for Experiential Learning and Simulation in the Department of Emergency Medicine at the University of Florida. “In emergency medicine, we are masters of the workaround,” she said. “But if we have to employ a workaround, that is usually an indicator of a systems problem. The delta between ‘work as imagined’ and ‘work as done’ is where we find these systems issues—whether it's that policies and procedures do not reflect work as it's actually done, or sufficient resources are not available under time pressure. “Maybe, for example, the ‘standard operating procedures’ are based on staffing that is rarely available in the real world or equipment that is not always on hand,” she said. “Simulations like TRUST give you the benefit of being able to see how the work is actually done.” There are drawbacks, of course, the most obvious one being that TRUST or any other in situ simulation by its very nature must take place in an ED or trauma unit that may, at any time, be called upon to respond to a real-life urgent case. “High-performing organizations that are most likely to employ these protocols to improve their care are by definition the kind of organizations that see a lot of trauma,” said Dr. Lauria. “If you're working in a busy ED, it's hard to stop and do a simulation in the middle of the day, interrupting everyone's workflow and utilizing equipment that may need to be in place at a moment's notice for a real trauma.”FigureThis may mean that simulation needs to be done in the off hours—either early in the morning or the middle of the night, which generally tend to be less busy times for the ED. “We've done in situ simulations at 2, 3, or 4 a.m.,” said Dr. Patterson. “Of course, you may also want to do one at high noon, which for a typical ED is around 9 p.m., to understand the adaptive capacity of the system even when it's at high volume, but it must be carefully done. We create ‘no-go’ rules, under which a planned simulation will be cancelled, such as when a certain number of people are already in the ED.” Getting Buy-In In situ simulation, especially when it involves the kind of technological and personnel resources deployed in TRUST, also has to have significant buy-in from the institution and potential participants. “There has to be a fair amount of up-front investment in terms of helping the institution's leadership and the affected teams understand what the purpose is,” she added. Dr. Petrosoniak and the team at St. Michael's continue to employ the TRUST protocol to improve emergency and trauma care at their institution. It is particularly useful when new procedures, systems, or equipment are put into place, he said. St. Michael's recently began utilizing the Resuscitative Endovascular Balloon Occlusion of the Aorta (REBOA) procedure, which involves occluding the aorta using an endovascular balloon to control hemorrhage in emergent situations, for example. “We just got approval from Health Canada for this device, which has been used for a couple of years in the United States, and piloted it using TRUST in our trauma bay for the first time in August,” he said. “We had five trauma surgeons and an entire trauma team there for a full-scale simulation.” The hospital has also employed TRUST to design its new emergency department and trauma bay. “The ED had already been built, but we were able to test the space using video capture and make changes before going live,” Dr. Petrosoniak said. “For example, we're piloting a new portable resuscitation tower because we're in a space that's about four times the size of our old space and we need to rapidly get equipment like intubation and chest tubes to the bedside quickly. We used to put it all in one spot, but now because of the size of the new space, we have two six-foot-tall resus towers on wheels that get rolled to wherever they are needed. We piloted them in simulation and made modifications. In those simulations, we also realized that we had poor sight lines to certain rooms, so we established a policy to place sicker patients in the rooms with optimal sight lines.” The TRUST team was able to be involved in the design of the trauma bay earlier on, prior to construction. “We did a sequential simulation approach where we started with tabletop simulations,” Dr. Petrosoniak said. “Using the tracing tool we employ in TRUST, we were able to track the movement of the entire trauma team. We realized that the radius allocated by architects for movement around the patient beds was too small, so they modified the space. This is just one example of many data-driven decisions we've been able to make based on simulation.”
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.104 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; both teacher heads agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".