Generative Artificial Intelligence for Chest Radiograph Interpretation in the Emergency Department.

Importance: Multimodal generative artificial intelligence (AI) methodologies have the potential to optimize emergency department care by producing draft radiology reports from input images. Objective: To evaluate the accuracy and quality of AI-generated chest radiograph interpretations in the emergency department setting. Design, Setting, and Participants: This was a retrospective diagnostic study of 500 randomly sampled emergency department encounters at a tertiary care institution including chest radiographs interpreted by both a teleradiology service and on-site attending radiologist from January 2022 to January 2023. An AI interpretation was generated for each radiograph. The 3 radiograph interpretations were each rated in duplicate by 6 emergency department physicians using a 5-point Likert scale. Main Outcomes and Measures: The primary outcome was any difference in Likert scores between radiologist, AI, and teleradiology reports, using a cumulative link mixed model. Secondary analyses compared the probability of each report type containing no clinically significant discrepancy with further stratification by finding presence, using a logistic mixed-effects model. Physician comments on discrepancies were recorded. Results: A total of 500 ED studies were included from 500 unique patients with a mean (SD) age of 53.3 (21.6) years; 282 patients (56.4%) were female. There was a significant association of report type with ratings, with post hoc tests revealing significantly greater scores for AI (mean [SE] score, 3.22 [0.34]; P < .001) and radiologist (mean [SE] score, 3.34 [0.34]; P < .001) reports compared with teleradiology (mean [SE] score, 2.74 [0.34]) reports. AI and radiologist reports were not significantly different. On secondary analysis, there was no difference in the probability of no clinically significant discrepancy between the 3 report types. Further stratification of reports by presence of cardiomegaly, pulmonary edema, pleural effusion, infiltrate, pneumothorax, and support devices also yielded no difference in the probability of containing no clinically significant discrepancy between the report types. Conclusions and Relevance: In a representative sample of emergency department chest radiographs, results suggest that the generative AI model produced reports of similar clinical accuracy and textual quality to radiologist reports while providing higher textual quality than teleradiologist reports. Implementation of the model in the clinical workflow could enable timely alerts to life-threatening pathology while aiding imaging interpretation and documentation.

Impact of the COVID pandemic on emergency department CT utilization: where do we go from here?

PURPOSE: To analyze the impact of the coronavirus disease (COVID) pandemic on emergency department (ED) computed tomography (CT) utilization. METHODS: A retrospective observational study was conducted assessing seven hospitals' ED imaging volumes between Jan. 6, 2019, and Feb. 27, 2021. Weekly CT utilization is reported as CTs ordered per 100 ED visits. Utilization was ascertained in aggregate and by body area. Interrupted time series analysis was performed to assess significance of utilization change. Prespecified sensitivity analysis was performed for influenza-like or COVID-like illness (ILI/CLI). RESULTS: Weekly ED CT utilization increased from 35.9 CTs per 100 visits (95% confidence interval [95% CI] 35.8-36.1) to 41.8 per 100 visits (95% CI 41.7-42.0) in pre- and post-pandemic periods. Weekly ED CT chest utilization increased immediately following the pandemic declaration (+ 0.52 chest CTs per 100 ED visits, 95% CI 0.01-1.03, p < 0.05) and compared to pre-pandemic period (+ 0.02 per 100 ED visits, 95% CI 0.02-0.05, p < 0.02). For both CT abdomen/pelvis and CT head, there was neither an immediate effect (+ 0.34 CT-AP per 100 ED visits, 95% CI - 0.74 to 1.44, p = 0.89; - 0.42 CT-H per 100 ED visits, 95% CI - 1.53 to 0.70, p = 0.46) nor a change in weekly CT utilization (+ 0.03 CT-AP per 100 ED visits, 95% CI - 0.01 to 0.05, p = 0.09; + 0.03 CT-H per 100 ED visits, 95% CI - 0.01 to 0.06, p = 0.10).  CONCLUSION: These data may help formulate future strategies for resource utilization and imaging operations as we envision a future with COVID and other federal mandates affecting imaging utilization and appropriateness.

Integrating Google Trends Search Engine Query Data Into Adult Emergency Department Volume Forecasting: Infodemiology Study.

Background: The search for health information from web-based resources raises opportunities to inform the service operations of health care systems. Google Trends search query data have been used to study public health topics, such as seasonal influenza, suicide, and prescription drug abuse; however, there is a paucity of literature using Google Trends data to improve emergency department patient-volume forecasting. Objective: We assessed the ability of Google Trends search query data to improve the performance of adult emergency department daily volume prediction models. Methods: Google Trends search query data related to chief complaints and health care facilities were collected from Chicago, Illinois (July 2015 to June 2017). We calculated correlations between Google Trends search query data and emergency department daily patient volumes from a tertiary care adult hospital in Chicago. A baseline multiple linear regression model of emergency department daily volume with traditional predictors was augmented with Google Trends search query data; model performance was measured using mean absolute error and mean absolute percentage error. Results: There were substantial correlations between emergency department daily volume and Google Trends "hospital" (r=0.54), combined terms (r=0.50), and "Northwestern Memorial Hospital" (r=0.34) search query data. The final Google Trends data-augmented model included the predictors Combined 3-day moving average and Hospital 3-day moving average and performed better (mean absolute percentage error 6.42%) than the final baseline model (mean absolute percentage error 6.67%)-an improvement of 3.1%. Conclusions: The incorporation of Google Trends search query data into an adult tertiary care hospital emergency department daily volume prediction model modestly improved model performance. Further development of advanced models with comprehensive search query terms and complementary data sources may improve prediction performance and could be an avenue for further research.

Rapid Emergency Department Physical Space Modifications for COVID-19: Keeping Patients and Health Care Workers Safe.

The COVID-19 pandemic has placed significant strain on emergency departments (EDs) that were not designed to care for many patients who may be highly contagious. This report outlines how a busy urban ED was adapted to prepare for COVID-19 via 3 primary interventions: (1) creating an open-air care space in the ambulance bay to cohort, triage, and rapidly test patients with suspected COVID-19, (2) quickly constructing temporary doors on all open treatment rooms, and (3) adapting and expanding the waiting room. This description serves as a model by which other EDs can repurpose their own care spaces to help ensure safety of their patients and health care workers.

Use of Incident Command System for Disaster Preparedness: A Model for an Emergency Department COVID-19 Response.

The COVID-19 pandemic has placed unprecedented demands on health systems, where hospitals have become overwhelmed with patients amidst limited resources. Disaster response and resource allocation during such crises present multiple challenges. A breakdown in communication and organization can lead to unnecessary disruptions and adverse events. The Federal Emergency Management Agency (FEMA) promotes the use of an incident command system (ICS) model during large-scale disasters, and we hope that an institutional disaster plan and ICS will help to mitigate these lapses. In this article, we describe the alignment of an emergency department (ED) specific Forward Command structure with the hospital ICS and address the challenges specific to the ED. Key components of this ICS include a hospital-wide incident command or Joint Operations Center (JOC) and an ED Forward Command. This type of structure leads to a shared mental model with division of responsibilities that allows institutional adaptations to changing environments and maintenance of specific roles for optimal coordination and communication. We present this as a model that can be applied to other hospital EDs around the country to help structure the response to the COVID-19 pandemic while remaining generalizable to other disaster situations.

The Emergent Evaluation and Treatment of Shoulder, Clavicle, and Humerus Injuries.

Shoulder pain is a common presentation in the emergency department. The list of differential diagnoses is broad. This article summarizes common diagnoses of shoulder pain, including bony, infectious, and connective tissue pathologies and their proper treatment. It also reviews which shoulder pain conditions are emergency diagnoses and need immediate treatment and which diagnoses need conservative management and outpatient follow-up.

A Quality Improvement Initiative Featuring Peer-Comparison Prescribing Feedback Reduces Emergency Department Opioid Prescribing.

BACKGROUND: Opioid prescribing in the United States nearly tripled from 1999 to 2015, and opioid overdose deaths doubled in the same time frame. Emergency departments (EDs) may play a pivotal role in the opioid epidemic as a source of first-time opioid exposure; however, many prescribers are generally unaware of their prescribing behaviors relative to their peers. METHODS: All 117 ED prescribers at an urban academic medical center were provided with regular feedback on individual rates of opioid prescribing relative to their de-identified peers. To evaluate the effect of this intervention on the departmental rate of opioid prescribing, a statistical process control (SPC) chart was created to identify special cause variation, and an interrupted time series analysis was conducted to evaluate the immediate effect of the intervention and any change in the postintervention trend due to the intervention. RESULTS: The aggregate opioid prescribing rate in the preintervention period was 8.6% (95% confidence interval [CI]: 8.3%-8.9%), while the aggregate postintervention prescribing rate was 5.8% (95% CI: 5.5%-6.1%). The SPC chart revealed special cause variation in both the pre- and postintervention periods, with an overall downtrend of opioid prescribing rates across the evaluation period and flattening of rates in the final four blocks. Interrupted time series analysis demonstrated a significant immediate downward effect of the intervention and a nonsignificant additional decrease in postintervention trend. CONCLUSION: Implementation of peer-comparison opioid prescribing feedback was associated with a significant immediate reduction in the rate of ED discharge opioid prescribing.

The care of VIPs in the emergency department: Triage, treatment and ethics.

People identified as Very Important Persons (VIPs) often present or are referred to the Emergency Department (ED). Celebrities are a small subset of this group, but many others are included. Triage of these patients, including occasional prioritization, creates practical and ethical challenges. Treatment also provides challenges with the risks of over testing, overtreatment, over consultation, and over or under admission to the hospital. This article presents a practical and ethical framework for addressing the care of VIPs in the ED.

Medical Care at a Large Vertical Running Event.

OBJECTIVE: Vertical running events, during which participants race up the stairwells of skyscrapers, are becoming increasingly popular. Such events have unique and specific operational and clinical considerations for event medical directors, but descriptions of the medical care provided at these events are lacking. We sought to perform a descriptive analysis of the medical care delivered at a single, large vertical running event. METHODS: A retrospective chart review of medical encounters at a large vertical running event from 2011-2017 was performed. Participants competed in either the full course (94 stories) or half course (54 stories); potential patients also included observers. Medical staffing included a main medical station at the finish line, medical way stations along the routes (within stairwells), and medical response teams. Descriptive statistics were used for analysis. RESULTS: During the study period, a total of 23,920 participants completed the event, with 84.6% participating in the full course. Medical staff treated 150 unique patients during 154 medical encounters (0.6% treatment rate). The median age of patients was 36 (IQR 27, 43), and 40.3% were male. Most encounters (66.4%) occurred at the finish line main medical area. Of medical encounters occurring along the race routes, 56.1% of encounters occurred before the halfway point in the full course. Encounters were clustered around medical way stations along the half course. The most common chief complaints were gastrointestinal (27.3%), respiratory (25.3%), syncope/near-syncope (24.7%), trauma (12.3%), and chest pain (10.4%). One cardiac arrest was observed. The most frequent interventions were oral fluids or food (40.3%), respiratory care (18.2%), and minor trauma care (12.3%). An electrocardiogram (ECG) was obtained in 10.4% of encounters, and intravenous fluids were started on 1.9% of patients. Eleven patients (7.3% of treated patients and 0.05% of all participants) were transported by ambulance. CONCLUSIONS: Medical encounters during vertical running events, the majority of which are not life-threatening, mainly occur at the finish line but can occur at any point along the route. Understanding the nature and location of medical encounters along a vertical running event route can help inform event medical directors supervising care at these increasingly popular events.

Impact of Discharge Anticoagulation Education by Emergency Department Pharmacists at a Tertiary Academic Medical Center.

BACKGROUND: Although pharmacists commonly provide patient education and help manage high-risk anticoagulant medications in inpatient and outpatient settings, the evidence for these interventions in the emergency department (ED) is less established, especially in the era of direct-acting oral anticoagulants. In 2013, a formal program was initiated whereby patients discharged with a new prescription for any anticoagulant receive education from an ED pharmacist when on-site. In addition, they received follow-up phone calls from an ED pharmacist within 72 hours of discharge. OBJECTIVE: We sought to identify the impact of pharmacist education, defined as the need for intervention on callback, versus physician and nursing-driven discharge measures on patient understanding and appropriate use of anticoagulant medications. METHODS: A single-center retrospective analysis included patients discharged from the ED between May 2013 and May 2016 with a new anticoagulant prescription. Electronic callback records were reviewed to assess patients' adherence and understanding of discharge instructions as well as for an anticoagulant-related hospital readmission within 90 days. RESULTS: One hundred seventy-four patients were evaluated in a per protocol analysis. Patients who did not receive pharmacist education prior to discharge required an increased need for intervention during callback versus those who did (36.4% vs. 12.9%, p = 0.0005) related to adherence, inappropriate administration, and continued use of interacting medications or supplements, among other concerns. In addition, patients who had not received pharmacist counseling were more likely to be readmitted to a hospital or return to the ED within 90 days after their initial visit for an anticoagulation-related problem versus patients who had (12.12% vs. 1.85%, p = 0.0069). CONCLUSION: Discharge education by ED pharmacists leads to improved patient understanding and appropriate use of anticoagulants.

Effectiveness of Resident Physicians as Triage Liaison Providers in an Academic Emergency Department.

INTRODUCTION: Emergency department (ED) crowding is associated with detrimental effects on ED quality of care. Triage liaison providers (TLP) have been used to mitigate the effects of crowding. Prior studies have evaluated attending physicians and advanced practice providers as TLPs, with limited data evaluating resident physicians as TLPs. This study compares operational performance outcomes between resident and attending physicians as TLPs. METHODS: This retrospective cohort study compared aggregate operational performance at an urban, academic ED during pre- and post-TLP periods. The primary outcome was defined as cost-effectiveness based upon return on investment (ROI). Secondary outcomes were defined as differences in median ED length of stay (LOS), median door-to-provider (DTP) time, proportion of left without being seen (LWBS), and proportion of "very good" overall patient satisfaction scores. RESULTS: Annual profit generated for physician-based collections through LWBS capture (after deducting respective salary costs) equated to a gain (ROI: 54%) for resident TLPs and a loss (ROI: -31%) for attending TLPs. Accounting for hospital-based collections made both profitable, with gains for resident TLPs (ROI: 317%) and for attending TLPs (ROI: 86%). Median DTP time for resident TLPs was significantly lower (p<0.0001) than attending or historical control. Proportion of "very good" patient satisfaction scores and LWBS was improved for both resident and attending TLPs over historical control. Overall median LOS was not significantly different. CONCLUSION: Resident and attending TLPs improved DTP time, patient satisfaction, and LWBS rates. Both resident and attending TLPs are cost effective, with residents having a more favorable financial profile.

Large-volume hypertonic saline therapy in endurance athlete with exercise-associated hyponatremic encephalopathy.

BACKGROUND: Small-volume boluses of intravenous hypertonic saline are the recommended therapy for exercise-associated hyponatremic encephalopathy (EAHE). Failure to properly diagnose and treat EAHE has been associated with significant morbidity and death. To prevent this, current consensus statement guidelines recommend up to three 100-mL boluses of 3% NaCl spaced at 10-min intervals to correct symptoms. Due to lack of evidence, however, guidelines are vague regarding the maximal volume that can be safely administered in a given time period beyond these initial boluses. OBJECTIVES: This case report will review the underlying pathophysiology, clinical presentation, diagnosis, and management of EAHE in a patient refractory to initial treatment. CASE REPORT: We report a case of EAHE in an experienced marathon runner requiring large-volume infusion (950 mL) of 3% NaCl therapy for resolution of symptoms without any adverse events. CONCLUSION: Although further research is needed, this case may provide helpful information for acute care and sports medicine physicians who encounter patients with EAHE refractory to initial therapy.

Prior altitude experience of climbers attempting to summit Aconcagua.

Aconcagua (6962 m) is one of the seven summits and the highest mountain outside of Asia. Climbers of varying experience are drawn to its nontechnical route. Our objective was to detail the prior altitude experience of climbers attempting to summit Aconcagua. We asked all climbers on the normal route of Aconcagua to complete questionnaires with demographics and prior high altitude experience while acclimatizing at Plaza de Mulas base camp during 9 nonconsecutive days in January 2009. 127 volunteers from 22 countries were enrolled. Mean age was 39.8 years and 88.2% were male. Median altitude at place of residence was 200 m (IQR: 30, 700). Median previous maximum altitude reached was 5895 m (IQR: 5365, 6150). 7.1% of climbers had never been above 4000 m. Median previous maximum sleeping altitude was 4800 m (IQR: 4300, 5486). 12.6% of climbers had never slept above 4000 m. Climbers who performed acclimatization treks spent a mean of 3.6 (2.5, 4.7) days at>3000 m in the previous 2 months. However, 50.4% of climbers performed no acclimatization treks. Although the majority of mountaineers who attempt Aconcagua have prior high altitude experience, a substantial minority has never been above 4000 m.

Enhancing community disaster resilience through mass sporting events.

Disaster response requires rapid, complex action by multiple agencies that may rarely interact during nondisaster periods. Failures in communication and coordination between agencies have been pitfalls in the advancement of disaster preparedness. Recommendations of the Federal Emergency Management Agency address these needs and demonstrate commitment to successful disaster management, but they are challenging for communities to ensure. In this article we describe the application of Federal Emergency Management Agency guidelines to the 2008 and 2009 Chicago Marathon and discuss the details of our implementation strategy with a focus on optimizing communication. We believe that it is possible to enhance community disaster preparedness through practical application during mass sporting events.

Visual analog scale (VAS) for assessment of acute mountain sickness (AMS) on Aconcagua.

OBJECTIVE: The Lake Louise AMS Self-Report Score (LLSelf) is a commonly used, validated assessment of acute mountain sickness (AMS). We compared LLSelf and visual analog scales (VAS) to quantify AMS on Aconcagua (6962 m). METHODS: Prospective observational cohort study at Plaza de Mulas base camp (4365 m), Aconcagua Provincial Park, Argentina. Volunteers climbing in January 2009 were enrolled at base camp and ascended at their own pace. They completed the LLSelf, an overall VAS [VAS(o)], and 5 individual VAS [VAS(i)] corresponding to the items of the LLSelf when symptoms were maximal. Composite VAS [VAS(c)] was calculated as the sum of the 5 VAS(i). RESULTS: A total of 127 volunteers consented to the study. Response rate was 52.0%. AMS occurred in 77.3% of volunteers, while 48.5% developed severe AMS. Median (interquartile range, IQR) LLSelf was 4 (3-7). Median (IQR) VAS(o) was 36 mm (23-59). VAS(o) was linear and correlated with LLSelf: slope = 6.7 (95% CI: 4.4-9.0), intercept = 3.0 (95% CI: -10.0-16.1), ρ = 0.71, τ = 0.55, R(2) = 0.45, p < 0.001. Median (IQR) VAS(c) was 29 (13-44). VAS(c) was also linear and correlated with LLSelf: slope = 5.9 (95% CI: 4.9-6.9), intercept = -0.6 (95% CI: -6.3-5.1), ρ = 0.83, τ = 0.68, R(2) = 0.73, p < 0.001. The relationship between the 5 VAS(i) and LLSelf(i) was less significant and less linear than that between VAS(o), VAS(c), and LLSelf. CONCLUSIONS: While both VAS(o) and VAS(c) for assessment of AMS appear to be linear with respect to LLSelf, the amount of scatter within the VAS is considerable. The LLSelf remains the gold standard for the diagnosis of AMS.

Postexercise peripheral oxygen saturation after completion of the 6-minute walk test predicts successfully reaching the summit of Aconcagua.

OBJECTIVE: The 6-minute walk test (6MWT) is a single measurement of functional status in patients with cardiovascular disease. It has not been studied at high altitude. We investigate the screening value of 6-minute walk distance (6MWD) and postexercise vital sign (VS) measurements as predictors of successfully reaching the summit or development of acute mountain sickness (AMS) on Aconcagua (6962 m). METHODS: Prospective observational cohort in Aconcagua Provincial Park, Argentina. Adults climbing the normal route who registered with base camp physicians were included. There were no exclusion criteria. VSs were measured before (resting) and after (postexercise) completion of 6MWT while volunteers acclimatized at Plaza de Mulas base camp (4365 m). Volunteers proceeded towards the summit at their own pace and upon descent returned a questionnaire with maximum altitude reached and Lake Louise AMS Self-report Score (LLSelf). RESULTS: One hundred twenty-four volunteers completed the 6MWT. Sixty-four volunteers (51.6%) completed questionnaires; 56% summited. Median LLSelf was 4 (IQR: 3.0-6.5). There was no association between any resting or postexercise VS measurements and AMS. However, mean postexercise SpO(2) was 80.8% in summiters and 76.4% in nonsummiters, a difference of -4.4% (95% CI: -6.7 to -2.0, p = 0.0005). Postexercise SpO(2) < 75% had 97.2% sensitivity and negative likelihood ratio of 0.086 in predicting the outcome of successfully reaching the summit: only one climber with SpO(2) < 75% successfully reached the summit. CONCLUSIONS: This study provides the first published data on 6MWD recorded in the field at high altitude. Postexercise SpO(2) < 75% may be a useful screening test for predicting the outcome of successfully reaching the summit of Aconcagua.

Emergent evaluation of injuries to the shoulder, clavicle, and humerus.

This article provides a review of the evaluation and treatment of common injuries to the shoulder, humerus, and clavicle in the emergency department (ED) setting. In addition to a focused review of the shoulder's physical examination, topics include common emergent injuries such as glenohumeral dislocations, proximal humerus fractures, and acromioclavicular separations as well as less common, but important injuries including pectoralis and biceps tendon injuries and sternoclavicular dislocations. Accurate recognition and management of these injuries is essential in the optimal care of patients in the ED.