ABSTRACT
BACKGROUND: Artificial intelligence (AI), more specifically large language models (LLMs), holds significant potential in revolutionizing emergency care delivery by optimizing clinical workflows and enhancing the quality of decision-making. Although enthusiasm for integrating LLMs into emergency medicine (EM) is growing, the existing literature is characterized by a disparate collection of individual studies, conceptual analyses, and preliminary implementations. Given these complexities and gaps in understanding, a cohesive framework is needed to comprehend the existing body of knowledge on the application of LLMs in EM. OBJECTIVE: Given the absence of a comprehensive framework for exploring the roles of LLMs in EM, this scoping review aims to systematically map the existing literature on LLMs' potential applications within EM and identify directions for future research. Addressing this gap will allow for informed advancements in the field. METHODS: Using PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) criteria, we searched Ovid MEDLINE, Embase, Web of Science, and Google Scholar for papers published between January 2018 and August 2023 that discussed LLMs' use in EM. We excluded other forms of AI. A total of 1994 unique titles and abstracts were screened, and each full-text paper was independently reviewed by 2 authors. Data were abstracted independently, and 5 authors performed a collaborative quantitative and qualitative synthesis of the data. RESULTS: A total of 43 papers were included. Studies were predominantly from 2022 to 2023 and conducted in the United States and China. We uncovered four major themes: (1) clinical decision-making and support was highlighted as a pivotal area, with LLMs playing a substantial role in enhancing patient care, notably through their application in real-time triage, allowing early recognition of patient urgency; (2) efficiency, workflow, and information management demonstrated the capacity of LLMs to significantly boost operational efficiency, particularly through the automation of patient record synthesis, which could reduce administrative burden and enhance patient-centric care; (3) risks, ethics, and transparency were identified as areas of concern, especially regarding the reliability of LLMs' outputs, and specific studies highlighted the challenges of ensuring unbiased decision-making amidst potentially flawed training data sets, stressing the importance of thorough validation and ethical oversight; and (4) education and communication possibilities included LLMs' capacity to enrich medical training, such as through using simulated patient interactions that enhance communication skills. CONCLUSIONS: LLMs have the potential to fundamentally transform EM, enhancing clinical decision-making, optimizing workflows, and improving patient outcomes. This review sets the stage for future advancements by identifying key research areas: prospective validation of LLM applications, establishing standards for responsible use, understanding provider and patient perceptions, and improving physicians' AI literacy. Effective integration of LLMs into EM will require collaborative efforts and thorough evaluation to ensure these technologies can be safely and effectively applied.
ABSTRACT
Background: Delayed intervention for ST-segment elevation myocardial infarction (STEMI) is associated with higher mortality. The association of door-to-ECG (D2E) with clinical outcomes has not been directly explored in a contemporary US-based population. Methods: This was a three-year, 10-center, retrospective cohort study of ED-diagnosed patients with STEMI comparing mortality between those who received timely (<10 min) vs. untimely (>10 min) diagnostic ECG. Among survivors, we explored left ventricular ejection fraction (LVEF) dysfunction during the STEMI encounter and recovery upon post-discharge follow-up. Results: Mortality was lower among those who received a timely ECG where one-week mortality was 5% (21/420) vs. 10.2% (26/256) among those with untimely ECGs (p = 0.016), and in-hospital mortality was 6.0% (25/420) vs. 10.9% (28/256) (p = 0.028). Data to compare change in LVEF metrics were available in only 24% of patients during the STEMI encounter and 46.5% on discharge follow-up. Conclusions: D2E within 10 min may be associated with a 50% reduction in mortality among ED STEMI patients. LVEF dysfunction is the primary resultant morbidity among STEMI survivors but was infrequently assessed despite low LVEF being an indication for survival-improving therapy. It will be difficult to assess the impact of STEMI care interventions without more consistent LVEF assessment.
ABSTRACT
We compared four methods to screen emergency department (ED) patients for an early electrocardiogram (ECG) to diagnose ST-elevation myocardial infarction (STEMI) in a 5-year retrospective cohort through observed practice, objective application of screening protocol criteria, a predictive model, and a model augmenting human practice. We measured screening performance by sensitivity, missed acute coronary syndrome (ACS) and STEMI, and the number of ECGs required. Our cohort of 279,132 ED visits included 1397 patients who had a diagnosis of ACS. We found that screening by observed practice augmented with the model delivered the highest sensitivity for detecting ACS (92.9%, 95%CI: 91.4-94.2%) and showed little variation across sex, race, ethnicity, language, and age, demonstrating equity. Although it missed a few cases of ACS (7.6%) and STEMI (4.4%), it did require ECGs on an additional 11.1% of patients compared to current practice. Screening by protocol performed the worst, underdiagnosing young, Black, Native American, Alaskan or Hawaiian/Pacific Islander, and Hispanic patients. Thus, adding a predictive model to augment human practice improved the detection of ACS and STEMI and did so most equitably across the groups. Hence, combining human and model screening--rather than relying on either alone--may maximize ACS screening performance and equity.
ABSTRACT
BACKGROUND: Innovation and entrepreneurship training are increasingly recognized as being important in medical education. However, the lack of faculty comfort with the instruction of these concepts as well as limited scholarly recognition for this work has limited the implementation of curricula focused on these skills. Furthermore, this lack of familiarity limits the inclusion of practicing physicians in health care innovation, where their experience is valuable. Hackathons are intense innovation competitions that use gamification principles to increase comfort with creative thinking, problem-solving, and interpersonal collaboration, but they require further exploration in medical innovation. OBJECTIVE: To address this, we aimed to design, implement, and evaluate a health care hackathon with 2 main goals: to improve emergency physician familiarity with the principles of health care innovation and entrepreneurship and to develop innovative solutions to 3 discrete problems facing emergency medicine physicians and patients. METHODS: We used previously described practices for conducting hackathons to develop and implement our hackathon (HackED!). We partnered with the American College of Emergency Physicians, the Stanford School of Biodesign, and the Institute of Design at Stanford (d.school) to lend institutional support and expertise in health care innovation to our event. We determined a location, time frame, and logistics for the competition and settled on 3 use cases for teams to work on. We planned to explore the learning experience of participants within a pragmatic paradigm and complete an abductive thematic analysis using data from a variety of sources. RESULTS: HackED! took place from October 1-3, 2022. In all, 3 teams developed novel solutions to each of the use cases. Our investigation into the educational experience of participants suggested that the event was valuable and uncovered themes suggesting that the learning experience could be understood within a framework from entrepreneurship education not previously described in relation to hackathons. CONCLUSIONS: Health care hackathons appear to be a viable method of increasing physician experience with innovation and entrepreneurship principles and addressing complex problems in health care. Hackathons should be considered as part of educational programs that focus on these concepts.
ABSTRACT
OBJECTIVE: To evaluate the association of the MyProstateScore (MPS) urine test on the decision to undergo biopsy in men referred for prostate biopsy in urology practice. METHODS: MPS testing was offered as an alternative to immediate biopsy in men referred to the University of Michigan for prostate biopsy from October 2013 through October 2016. The primary endpoint was the decision to perform biopsy. The proportion of patients undergoing biopsy was compared to predicted risk scores from the Prostate Cancer Prevention Trial risk calculator (PCPTrc). Analyses were stratified by the use of multiparametric magnetic resonance imaging (mpMRI). The associations of PCPTrc, MPS, and mpMRI with the decision to undergo biopsy were explored in a multivariable logistic regression model. RESULTS: Of 248 patients, 134 (54%) proceeded to prostate biopsy. MPS was significantly higher in biopsied patients (median 29 vs14, P < .001). The use of biopsy was strongly associated with MPS, with biopsy rates of 26%, 38%, 58%, 90%, and 85% in the first through fifth quintiles, respectively (P < .001). MPS association with biopsy persisted upon stratification by mpMRI. On multivariable analysis, MPS was strongly associated with the decision to undergo biopsy when modeled as both a continuous (odds ratio [OR] 1.05, 95%; confidence interval [CI] 1.04-1.08; <.001) and binary (OR 7.76, 95%; CI 4.14-14.5; P < .001) variable. CONCLUSION: Many patients (46%) undergoing clinical MPS testing as an alternative to immediate prostate biopsy were able to avoid biopsy. Increasing MPS was strongly associated with biopsy rates. These findings were robust to use of mpMRI.
