Bibliographic record
Abstract
FigureHave you ever felt that your gut instinct as an experienced emergency physician was just as accurate as some of the diagnostic tests you administer? Did you ever take a look at a patient, and think, “I know exactly what the lab findings are going to tell me?” It turns out you may be right — and now data back up that notion. An abstract presented at American College of Emergency Physicians' Scientific Assembly in October by researchers from the Mayo Clinic in Rochester, MN, confirmed that you should definitely listen to that voice in your gut. The study compared a “System 1” decision-making model (fast, information-limited, intuitive) with a “System 2” model (slow, information-heavy, cognitive) in the emergency department. “System 1” decision-making for acuity prediction (“sick” vs. “not sick”) had a sensitivity of 74% and a specificity of 83%, approximately as accurate as rapid-response diagnostic tests like the flu test. Jeffrey Wiswell, MD, one of the study's investigators and the lead author of a related paper (Am J Emerg Med 2013;31[10]:1448), said the team was inspired by Malcolm Gladwell's book Blink and the work of former experimental psychologist and diagnostic expert Pat Croskerry, MD, PhD, a professor of emergency medicine at Dalhousie University in Halifax, Nova Scotia, Canada. “I think we've all got the sense that we know when someone comes in who's not doing well vs. someone who looks OK, and has come in for something relatively minor,” Dr. Wiswell said. “With EDs getting more busy, there's always the drive to improve efficiency and decrease length of stay, and we wanted to study whether or not you could safely use that initial gut impression to speed things along in your work flow.” Dr. Wiswell and his colleagues, including Daniel Cabrera, MD, an assistant professor of emergency medicine, conducted a month-long prospective observational study of emergency physicians and residents in the Mayo's tertiary-care ED, which has some 73,000 patient visits annually. Physicians were given less than a minute to observe each patient, along with access to vital signs and basic demographic data. They were then asked to predict whether the patient was “sick” or “not sick.” “What we asked them to assess was whether or not this person was really ill to the point where if you don't act now, there could be some sort of long-term harm,” Dr. Wiswell explained. “We wanted to keep it really simple and assess that impression in a way that's difficult to do with simulations or scenarios on paper.” Two independent clinical investigators blinded to the on-the-spot assessments used published criteria to reach a gold standard of whether the patients were, in fact, “sick,” and their findings were compared with the rapid ratings. The study included 662 observations of 289 different patients; providers had a sensitivity of 73.9% and a specificity of 83.3% in accurately determining the acuity of the patients they observed. Residents' gut instincts appeared just about as fine-tuned as attendings'. “I think this is the first study I've seen of gut instinct diagnostic accuracy for all comers in the ED,” said Gabe Wilson, MD, the medical director of the ED at CHRISTUS Santa Rosa Hospital in Westover Hills, TX, and a regional director for EmCare overseeing five of the CHRISTUS Santa Rosa emergency departments. “There are some studies on detecting pulmonary embolism and risk stratifying based on gut instinct vs. very strict criteria like the Wells' criteria. In those studies as well, experienced attendings' accuracy was similar to rule-based criteria. I don't think this issue is addressed very well in residency, how much we should go with our instinct, and how much further we have to delve.” But the sensitivity and specificity, while just as good as a rapid-response flu test, still left something to be desired in the low 70s and 80s, Dr. Wilson added. “This means we are good at knowing when we need to attend to patients right away, but we can't stop there. We still need to attend to our differentials, go down our checklists, and do our testing.” Dr. Wiswell agreed. “If your prediction is right four out of five times, that's still not good enough,” he said. “With pushback nationally to cut costs and perhaps reduce diagnostic testing and diagnostic imaging, this study validates the general strength of the gut instinct and still documents the need for those other tools. A lot of times you can make a decision based on the initial gestalt, but a significant minority of the time you need to have additional input that won't come in the first few minutes but is really going to change your management.” Nonetheless, he said, trusting the power of the initial gut instinct could be useful in ED triage planning. “One way this could be helpful is that if you think somebody is truly ill, you could get things started more quickly, in that first minute — getting an ICU bed ready, calling a surgeon, getting some intensivists and other downstream providers involved,” said Dr. Wiswell. “That could be particularly helpful in an era where we have been boarding a lot of people in the ED, and length of stay is so long.” Access the linksin EMN by reading this on our website or in our free iPad app, both available atwww.EM-News.com. Comments?Write to us at[email protected].
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) | 0.001 | 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) | 0.006 | 0.000 |
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; a candidate call from one teacher head, not a consensus.
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".