The Role of Large Language Models in Transforming Emergency Medicine: Scoping Review.

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.

Integrating Battlefield Documentation into Virtual Reality Medical Simulation Training: Virtual Battlefield Assisted Trauma Distributed Observation Kit (BATDOK).

INTRODUCTION: Clinical documentation is an essential component of the provision of medical care, enabling continuity of information across provider and site handoffs. This is particularly important in the combat casualty care setting when a single casualty may be treated by four or more or five completely disparate teams across the roles of care. The Battlefield Assisted Trauma Distributed Observation Kit (BATDOK) is a digital battlefield clinical documentation system developed by the Air Force Research Laboratory to address this need. To support the deployment of this tool, we integrated BATDOK into a commercially available virtual reality (VR) medical simulation platform used by the U.S. Air Force and Defense Health Agency personnel in order to provide an immersive simulation training experience which included battlefield documentation. METHODS: A multidisciplinary team consisting of medical educators, VR simulation engineers, emergency physicians and pararescuemen, and BATDOK developers first developed a specification for a virtual BATDOK capability, including a detailed listing of learning objectives, critical interfaces and task plans, and sensor integrations. These specifications were then implemented into the commercially available Virtual Advancement of Learning for Operational Readiness VR Medical Simulation System and underwent developmental testing and evaluation during pararescueman training exercises at the Air Force Special Operations Command Special Operations Center for Medical Integration and Development. RESULTS AND CONCLUSIONS: The BATDOK capability was successfully implemented within the VR Medical Simulation System. The capability consisted of a virtual tablet with replicated interfaces and capabilities based on the developed specifications. These capabilities included integrated point-of-care ultrasound capability, multi-patient management, vitals sign monitoring with sensor pairing and continuous monitoring, mechanism of injury documentation (including injury pattern documentation), intervention logging (including tourniquets, dressing, airways, lines, tubes and drains, splints, fluids, and medications), and event logging. The capability was found to be operational and in alignment with learning objectives and user acceptance goals.

Changes in low-acuity patient volume in an emergency department after launching a walk-in clinic.

Objective: Unscheduled low-acuity care options are on the rise and are often expected to reduce emergency department (ED) visits. We opened an ED-staffed walk-in clinic (WIC) as an alternative care location for low-acuity patients at a time when ED visits exceeded facility capacity and the impending flu season was anticipated to increase visits further, and we assessed whether low-acuity ED patient visits decreased after opening the WIC. Methods: In this retrospective cohort study, we compared patient and clinical visit characteristics of the ED and WIC patients and conducted interrupted time-series analyses to quantify the impact of the WIC on low-acuity ED patient visit volume and the trend. Results: There were 27,211 low-acuity ED visits (22.7% of total ED visits), and 7,058 patients seen in the WIC from February 26, 2018, to November 17, 2019. Low-acuity patient visits in the ED reduced significantly immediately after the WIC opened (P = 0.01). In the subsequent months, however, patient volume trended back to pre-WIC volumes such that there was no significant impact at 6, 9, or 12 months (P = 0.07). Had WIC patients been seen in the main ED, low-acuity volume would have been 27% of the total volume rather than the 22.7% that was observed. Conclusion: The WIC did not result in a sustained reduction in low-acuity patients in the main ED. However, it enabled emergency staff to see low-acuity patients in a lower resource setting during times when ED capacity was limited.

Effectiveness, safety, and efficiency of a drive-through care model as a response to the COVID-19 testing demand in the United States.

Objectives: Here we report the clinical performance of COVID-19 curbside screening with triage to a drive-through care pathway versus main emergency department (ED) care for ambulatory COVID-19 testing during a pandemic. Patients were evaluated from cars to prevent the demand for testing from spreading COVID-19 within the hospital. Methods: We examined the effectiveness of curbside screening to identify patients who would be tested during evaluation, patient flow from screening to care team evaluation and testing, and safety of drive-through care as 7-day ED revisits and 14-day hospital admissions. We also compared main ED efficiency versus drive-through care using ED length of stay (EDLOS). Standardized mean differences (SMD) >0.20 identify statistical significance. Results: Of 5931 ED patients seen, 2788 (47.0%) were walk-in patients. Of these patients, 1111 (39.8%) screened positive for potential COVID symptoms, of whom 708 (63.7%) were triaged to drive-through care (with 96.3% tested), and 403 (36.3%) triaged to the main ED (with 90.5% tested). The 1677 (60.2%) patients who screened negative were seen in the main ED, with 440 (26.2%) tested. Curbside screening sensitivity and specificity for predicting who ultimately received testing were 70.3% and 94.5%. Compared to the main ED, drive-through patients had fewer 7-day ED revisits (3.8% vs 12.5%, SMD = 0.321), fewer 14-day hospital readmissions (4.5% vs 15.6%, SMD = 0.37), and shorter EDLOS (0.56 vs 5.12 hours, SMD = 1.48). Conclusion: Curbside screening had high sensitivity, permitting early respiratory isolation precautions for most patients tested. Low ED revisit, hospital readmissions, and EDLOS suggest drive-through care, with appropriate screening, is safe and efficient for future respiratory illness pandemics.

Using a Real-Time Locating System to Evaluate the Impact of Telemedicine in an Emergency Department During COVID-19: Observational Study.

BACKGROUND: Telemedicine has been deployed by health care systems in response to the COVID-19 pandemic to enable health care workers to provide remote care for both outpatients and inpatients. Although it is reasonable to suspect telemedicine visits limit unnecessary personal contact and thus decrease the risk of infection transmission, the impact of the use of such technology on clinician workflows in the emergency department is unknown. OBJECTIVE: This study aimed to use a real-time locating system (RTLS) to evaluate the impact of a new telemedicine platform, which permitted clinicians located outside patient rooms to interact with patients who were under isolation precautions in the emergency department, on in-person interaction between health care workers and patients. METHODS: A pre-post analysis was conducted using a badge-based RTLS platform to collect movement data including entrances and duration of stay within patient rooms of the emergency department for nursing and physician staff. Movement data was captured between March 2, 2020, the date of the first patient screened for COVID-19 in the emergency department, and April 20, 2020. A new telemedicine platform was deployed on March 29, 2020. The number of entrances and duration of in-person interactions per patient encounter, adjusted for patient length of stay, were obtained for pre- and postimplementation phases and compared with t tests to determine statistical significance. RESULTS: There were 15,741 RTLS events linked to 2662 encounters for patients screened for COVID-19. There was no significant change in the number of in-person interactions between the pre- and postimplementation phases for both nurses (5.7 vs 7.0 entrances per patient, P=.07) and physicians (1.3 vs 1.5 entrances per patient, P=.12). Total duration of in-person interactions did not change (56.4 vs 55.2 minutes per patient, P=.74) despite significant increases in telemedicine videoconference frequency (0.6 vs 1.3 videoconferences per patient, P<.001 for change in daily average) and duration (4.3 vs 12.3 minutes per patient, P<.001 for change in daily average). CONCLUSIONS: Telemedicine was rapidly adopted with the intent of minimizing pathogen exposure to health care workers during the COVID-19 pandemic, yet RTLS movement data did not reveal significant changes for in-person interactions between staff and patients under investigation for COVID-19 infection. Additional research is needed to better understand how telemedicine technology may be better incorporated into emergency departments to improve workflows for frontline health care clinicians.

Emergency Department Access During COVID-19: Disparities in Utilization by Race/Ethnicity, Insurance, and Income.

INTRODUCTION: In March 2020, shelter-in-place orders were enacted to attenuate the spread of coronavirus 2019 (COVID-19). Emergency departments (EDs) experienced unexpected and dramatic decreases in patient volume, raising concerns about exacerbating health disparities. METHODS: We queried our electronic health record to describe the overall change in visits to a two-ED healthcare system in Northern California from March-June 2020 compared to 2019. We compared weekly absolute numbers and proportional change in visits focusing on race/ethnicity, insurance, household income, and acuity. We calculated the z-score to identify whether there was a statistically significant difference in proportions between 2020 and 2019. RESULTS: Overall ED volume declined 28% during the study period. The nadir of volume was 52% of 2019 levels and occurred five weeks after a shelter-in-place order was enacted. Patient demographics also shifted. By week 4 (April 5), the proportion of Hispanic patients decreased by 3.3 percentage points (pp) (P = 0.0053) compared to a 6.2 pp increase in White patients (P = 0.000005). The proportion of patients with commercial insurance increased by 11.6 pp, while Medicaid visits decreased by 9.5 pp (P < 0.00001) at the initiation of shelter-in-place orders. For patients from neighborhoods <300% federal poverty levels (FPL), visits were -3.8 pp (P = 0.000046) of baseline compared to +2.9 pp (P = 0.0044) for patients from ZIP codes at >400% FPL the week of the shelter-in-place order. Overall, 2020 evidenced a consistently elevated proportion of high-acuity Emergency Severity Index (ESI) level 1 patients compared to 2019. Increased acuity was also demonstrated by an increase in the admission rate, with a 10.8 pp increase from 2019. Although there was an increased proportion of high-acuity patients, the overall census was decreased. CONCLUSION: Our results demonstrate changing ED utilization patterns circa the shelter-in-place orders. Those from historically vulnerable populations such as Hispanics, those from lower socioeconomic areas, and Medicaid users presented at disproportionately lower rates and numbers than other groups. As the pandemic continues, hospitals should use operations data to monitor utilization patterns by demographic, in addition to clinical indicators. Messaging about availability of emergency care and other services should include vulnerable populations to avoid exacerbating healthcare disparities.

Protecting the emergency physician workforce during the coronavirus disease 2019 pandemic through precision scheduling at an academic tertiary care trauma center.

The coronavirus disease 2019 (COVID-19) pandemic created new emergency physician staffing challenges. Emergency physicians may be taken out of the workforce because of respiratory symptoms or pending severe acute respiratory syndrome coronavirus 2 test results. Vulnerable emergency physician populations with increased risk of serious disease and death from COVID-19 include physicians at older ages; those with chronic medical conditions, including cardiac and pulmonary diseases and immunosuppression; and potentially pregnancy. We present our approach to planning for staffing issues through precision scheduling. We describe the actions taken to protect our vulnerable physicians and maximize our physician coverage. Measures include optimizing workforce; increasing backup call system; adjusting shifts based on patient arrival times, volume, and surge predictions; minimizing exposure to COVID-19 and reduce personal protective equipment use through telemedicine, huddles, and, creating lower risk emergency department care areas; and standardizing intubations to limit exposure.

Patient Age, Race and Emergency Department Treatment Area Associated with "Topbox" Press Ganey Scores.

INTRODUCTION: Hospitals commonly use Press Ganey (PG) patient satisfaction surveys for benchmarking physician performance. PG scores range from 1 to 5, with 5 being the highest, which is known as the "topbox" score. Our objective was to identify patient and physician factors associated with topbox PG scores in the emergency department (ED). METHODS: We looked at PG surveys from January 2015-December 2017 at an academic, urban hospital with 78,000 ED visits each year. Outcomes were topbox scores for the questions: "Likelihood of your recommending our ED to others"; and "Courtesy of the doctor." We analyzed topbox scores using generalized estimating equation models clustered by physician and adjusted for patient and physician factors. Patient factors included age, gender, race, ethnicity, and ED area where patient was seen. The ED has four areas based on patient acuity: emergent; urgent; vertical (urgent but able to sit in a recliner rather than a gurney); and fast track (non-urgent). Physician factors included age, gender, race, ethnicity, and number of years at current institution. RESULTS: We analyzed a total of 3,038 surveys. For "Likelihood of your recommending our ED to others," topbox scores were more likely with increasing patient age (odds ratio [OR] 1.07; 95% confidence interval [CI], 1.03-1.12); less likely among female compared to male patients (OR 0.81; 95% CI, 0.70-0.93); less likely among Asian compared to White patients (OR 0.71; 95% CI, 0.60-0.83); and less likely in the urgent (OR 0.71; 95% CI, 0.54-0.93) and vertical areas (OR 0.71; 95% CI 0.53-0.95) compared to fast track. For "Courtesy of the doctor," topbox scores were more likely with increasing patient age (OR 1.1; CI, 1.06-1.14); less likely among Asian (OR 0.70; 95% CI, 0.58-0.84), Black (OR 0.66; 95% CI, 0.45-0.96), and Hispanic patients (OR 0.68; 95% CI, 0.55-0.83) compared to White patients; and less likely in urgent area (OR 0.69; 95% CI, 0.50-0.95) compared to fast track. CONCLUSION: Increasing patient age was associated with increased likelihood of topbox scores, while Asian patients, and urgent and vertical areas had decreased likelihood of topbox scores. We encourage hospitals that use PG topbox scores as financial incentives to understand the contribution of non-service factors to these scores.

A Custom-Developed Emergency Department Provider Electronic Documentation System Reduces Operational Efficiency.

STUDY OBJECTIVE: Electronic health record implementation can improve care, but may also adversely affect emergency department (ED) efficiency. We examine how a custom, ED provider, electronic documentation system (eDoc), which replaced paper documentation, affects operational performance. METHODS: We analyzed retrospective operational data for 1-year periods before and after eDoc implementation in a single ED. We computed daily operational statistics, reflecting 60,870 pre- and 59,337 postimplementation patient encounters. The prespecified primary outcome was daily mean length of stay; secondary outcomes were daily mean length of stay for admitted and discharged patients and daily mean arrival time to disposition for admitted patients. We used a prespecified multiple regression model to identify differences in outcomes while controlling for prespecified confounding variables. RESULTS: The unadjusted change in length of stay was 8.4 minutes; unadjusted changes in secondary outcomes were length of stay for admitted patients 11.4 minutes, length of stay for discharged patients 1.8 minutes, and time to disposition 1.8 minutes. With a prespecified regression analysis to control for variations in operational characteristics, there were significant increases in length of stay (6.3 minutes [95% confidence interval 3.5 to 9.1 minutes]) and length of stay for discharged patients (5.1 minutes [95% confidence interval 1.9 to 8.3 minutes]). There was no statistically significant change in length of stay for admitted patients or time to disposition. CONCLUSION: In our single-center study, the isolated implementation of eDoc was associated with increases in overall and discharge length of stay. Our findings suggest that a custom-designed electronic provider documentation may negatively affect ED throughput. Strategies to mitigate these effects, such as reducing documentation requirements or adding clinical staff, scribes, or voice recognition, would be a valuable area of future research.

Compliance with an ordinance requiring the use of personal flotation devices by children in public waterways.

INTRODUCTION: For children ages 1-14, 21.6% of drowning cases involve swimming, wading, or playing in natural bodies of water, such as rivers and lakes. Personal flotation devices (PFDs) are believed to be an effective prevention measure. We measure compliance with city and county ordinances, publicized but not actively enforced, requiring that PFDs be worn by children accessing public bodies of water in Sacramento County, California. METHODS: During June-August 2010, volunteers conducted 79 observation sessions at three popular local river beaches where PFDs were available for use at no cost. They recorded personal characteristics and PFD use for 1,727 children in or very near the water and believed to be 0-13 years of age (the age covered by the ordinances). We used logistic regression to quantify differences in use by subject characteristics and study site. RESULTS: The prevalence of PFD use was 29.9% overall, with large and significant differences by age: < 1, 55.6%; 1-4, 37.6%; 5-10, 29.4%; 10-13, 14.6%; P < 0.0001. Usage did not vary significantly by sex or race/ethnicity, and was somewhat higher at one study site (33.1%) than at the others (25.9% and 27.3%), P = 0.009. CONCLUSION: The combination of a statutory requirement and a cost-elimination strategy was associated with moderate rates of PFD use that were highest among young children.