Department of Emergency Medicine (UCI)

Importance

Objective

Design, setting, and participants

Main outcomes and measures

Results

Conclusions and relevance

Patient-controlled data-sharing systems are increasingly promoted as a way to empower patients with greater autonomy over their health data. Yet it remains unclear how different stakeholders, especially patients and health system leaders, perceive the benefits and challenges of enabling granular control over the sharing of de-identified medical data for research. To address this gap, we developed a high-fidelity prototype of a patient-controlled, web-based consent platform and conducted a two-phase mixed-methods study: semi-structured interviews with 16 health system leaders and a survey with 523 patient participants. While both groups appreciated the potential of such a platform to enhance transparency and autonomy, their views diverged in meaningful ways. Leaders viewed transparency and granular control through the lens of informed consent and institutional ethics, whereas patients interpreted these factors as safeguards against potential risks and uncertainties. Our findings underscore critical tensions such as individual control and research integrity. We offer design implications for building trustworthy, context-aware systems that support flexible granularity, provide ongoing benefit‑centered transparency, and adapt to diverse literacy and privacy needs.

AbstractBackground

In 2023, road traffic collisions caused 2.44 million injuries and 40,901 deaths in the U.S., with young adults disproportionately affected. This is the implementation report of the Virtual Ambassador Program at the University of California, Irvine (UCI), supported by grant from the California Office of Traffic Safety. The program engages students as virtual ambassadors to spread sober driving messages through their personal social media networks and prepares them to assume leadership of the initiative.

To foster student partnership and engagement in traffic safety efforts at UCI and ultimately reducing impaired driving.

The campaign began in 2022 on Facebook and Twitter, with student ambassadors disseminating sober driving messages through their personal networks. It later expanded to TikTok and Instagram, and the content scope broadened to include marijuana alongside alcohol and drugs. In 2024, the ambassadors posted 1,057 posts, reaching 40,401 unique users with 27,253 video views. In 2025, leadership began transitioning from UCI faculty to the student organization Drive Sober Orange County. This marked the beginning of student-led leadership, supported by mentorship, training, and funding. Full student ownership is expected after the transition, enabling sustained peer-driven engagement in traffic safety.

This report highlights the untapped potential of student partnership in road safety. With structured support, university students served as effective contributors to road safety. Using a cascade model of peer dissemination, they reached broad audiences with messages likely to resonate with their high-risk age group.

• University students are willing and capable contributors to road safety. Peer influence is underutilized in impaired driving prevention.

• With structured mentorship and support, student-led strategies can extend the reach of traffic safety messaging and foster effective partnerships.

Youth empowerment, Impaired driving, Virtual Ambassador.

INTRODUCTION: Hemorrhage continues to be the leading cause of preventable death in trauma, and tourniquet application has been associated with survival. The purpose of our study was to evaluate the efficacy of telesimulation (TeleSIM) versus in-person training (SIM) for teaching tourniquet application for life-threatening hemorrhage control. METHODS: We performed a prospective randomized study of participants enrolled in a Stop The Bleed course at a university medical school. The TeleSIM group completed the course with an instructor streaming live from an off-site location. The SIM group completed the course in the standard fashion with a live instructor present. The primary endpoint was the successful application of a combat application tourniquet to a bleeding extremity in a simulation scenario. We also evaluated the time for successful tourniquet application according to training modality. Participants' thoughts, feelings, and attitudes pertaining to their experience in the course were obtained via a postcourse survey. RESULTS: Ninety-four of 97 (96.9%) eligible subjects participated in the study. There was no difference in the proportion of participants in each group who successfully applied a combat application tourniquet during their simulation scenario: TeleSIM group, 100% (95% CI, 92.5-100.0); SIM group, 100% (95% CI, 92.7-100.0). We also observed no significant difference in the mean time it took the participants to apply a tourniquet regardless of their training modality. Both groups reported their learning modality as an effective way to learn hemorrhage control. CONCLUSION: A telesimulation-based instructional delivery design is an effective way to teach tourniquet application for hemorrhage control.

Cognitive rumination, a transdiagnostic symptom across mental health disorders, has traditionally been assessed through self-report measures. However, these measures are limited by their temporal nature and subjective bias. The rise in wearable technologies offers the potential for continuous, real-time monitoring of physiological indicators associated with rumination. This scoping review investigates the current state of research on using wearable technology to detect cognitive rumination. Specifically, we examine the sensors and wearable devices used, physiological biomarkers measured, standard measures of rumination used, and the comparative validity of specific biomarkers in identifying cognitive rumination. The review was performed according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines on IEEE, Scopus, PubMed, and PsycInfo databases. Studies that used wearable devices to measure rumination-related physiological responses and biomarkers were included (n = 9); seven studies assessed one biomarker, and two studies assessed two biomarkers. Electrodermal Activity (EDA) sensors capturing skin conductance activity emerged as both the most prevalent sensor (n = 5) and the most comparatively valid biomarker for detecting cognitive rumination via wearable devices. Other commonly investigated biomarkers included electrical brain activity measured through Electroencephalogram (EEG) sensors (n = 2), Heart Rate Variability (HRV) measured using Electrocardiogram (ECG) sensors and heart rate fitness monitors (n = 2), muscle response measured through Electromyography (EMG) sensors (n = 1) and movement measured through an accelerometer (n = 1). The Empatica E4 and Empatica Embrace 2 wrist-worn devices were the most frequently used wearable (n = 3). The Rumination Response Scale (RRS), was the most widely used standard scale for assessing rumination. Experimental induction protocols, often adapted from Nolen-Hoeksema and Morrow's 1993 rumination induction paradigm, were also widely used. In conclusion, the findings suggest that wearable technology offers promise in capturing real-time physiological responses associated with rumination. However, the field is still developing, and further research is needed to validate these findings and explore the impact of individual traits and contextual factors on the accuracy of rumination detection.

Background/Objective

Most pediatric stroke survivors suffer long-term impairments. To minimize injury, it is essential to quickly restore perfusion to viable brain tissue. Minimizing the time to stroke diagnosis requires recognition of a possible stroke by prehospital and emergency healthcare personnel, and rapid neuroimaging. While CT suffices for diagnosing hemorrhagic stroke, MRI is necessary to diagnose acute ischemic stroke (IS), contributing to significant diagnostic delays and potentially missed opportunities for intervention.

We conducted a retrospective study of children 1–14 years old with acute neurological symptoms presenting by Emergency Medical Services (EMS) to the study institution from 1/2019–6/2023. We described patient characteristics and neuroimaging studies, then evaluated predictors of MRI acquisition and actionable findings, including stroke. To assess the generalizability of these data we analyzed a secondary retrospective cohort of all children admitted during this period with out-of-hospital strokes regardless of presentation modality [EMS, emergency department (ED) walk-in, and transfer].

Among 3,888 pediatric patients with acute neurological symptoms presenting via EMS, 695 (17.9%) had neuroimaging: CT only in 570 patients (14.7%); CT and MRI in 125 (3.2%). Median (IQR) times from EMS activation to neuroimaging were 2.29 (1.56, 3.21) hours for CT and 26.8 (16.3, 43.8) hours for MRI. An EMS primary impression of “stroke” was rare (n = 13) but strongly predictive of imaging acquisition: all had CT and 11 had MRI. Thirty-one of the 125 patients with MRI had actionable MRIs, including nine acute strokes. During the study period another 14 stroke patients presented as ED walk-ins. Median time from ED arrival to CT was 0.92 (0.47, 1.08) hours for EMS patients with hemorrhagic stroke and 5.69 (1.50, 9.76) hours for walk-ins; for MRI, median time was 4.15 (3.00, 5.31) hours for EMS patients with ischemic stroke and 10.2 (1.99, 36.3) hours for walk-ins.

Among children with acute neurological symptoms selected for neuroimaging, CT was the most common modality while MRIs were performed with a substantial time delay. While EMS providers rarely suspected stroke, their diagnosis impacted imaging decisions in the ED, suggesting a need to raise awareness among prehospital providers. To measure quality improvement in pediatric stroke, new pediatric-specific metrics like “door to diagnosis” time, should be further explored.

BackgroundResearch demonstrates that Point-of-care ultrasound (POCUS) improves clinical outcomes for patients. Improving clinician satisfaction with POCUS should promote utilization into everyday practice, leading to improved clinical outcomes. Despite this benefit, there are still barriers to use including POCUS workflow. This project was undertaken to improve the functionality of the existing POCUS workflow and move toward an “encounter-based” system by automating worklist generation. It aimed to streamline the POCUS workflow, primarily determine if there was improved clinician satisfaction with the new workflow, and secondarily determine the change in revenue generation from decreased errors in data entry.MethodsA new workflow was created which automatically populated every registered Emergency Department (ED) patient into the worklist upon patient registration. Clinician feedback on their use of the new workflow was sought via survey after implementation. The number of medical record number (MRN) entry errors prior to and following implementation was manually reviewed and calculated.ResultsThere was a strong preference for the new workflow, with 36 of 38 (94.7%) clinicians finding it to be more convenient and 37 of 38 (97.4%) finding it to be preferable to use compared to the old workflow. Implementation also resulted in a 36% reduction in database studies containing an MRN data entry error.ConclusionsAn “encounter-based” workflow is strongly preferred over manual data entry for POCUS workflow among clinicians. Additionally, there was no cost to the intervention and the total data entry errors were significantly reduced, allowing for improved quality review and increased revenue.

BACKGROUND: Acute pain management is critical in postoperative care, especially in vulnerable patient populations that may be unable to self-report pain levels effectively. Current methods of pain assessment often rely on subjective patient reports or behavioral pain observation tools, which can lead to inconsistencies in pain management. Multimodal pain assessment, integrating physiological and behavioral data, presents an opportunity to create more objective and accurate pain measurement systems. However, most previous work has focused on healthy subjects in controlled environments, with limited attention to real-world postoperative pain scenarios. This gap necessitates the development of robust, multimodal approaches capable of addressing the unique challenges associated with assessing pain in clinical settings, where factors like motion artifacts, imbalanced label distribution, and sparse data further complicate pain monitoring. OBJECTIVE: This study aimed to develop and evaluate a multimodal machine learning-based framework for the objective assessment of pain in postoperative patients in real clinical settings using biosignals such as electrocardiogram, electromyogram, electrodermal activity, and respiration rate (RR) signals. METHODS: The iHurt study was conducted on 25 postoperative patients at the University of California, Irvine Medical Center. The study captured multimodal biosignals during light physical activities, with concurrent self-reported pain levels using the Numerical Rating Scale. Data preprocessing involved noise filtering, feature extraction, and combining handcrafted and automatic features through convolutional and long-short-term memory autoencoders. Machine learning classifiers, including support vector machine, random forest, adaptive boosting, and k-nearest neighbors, were trained using weak supervision and minority oversampling to handle sparse and imbalanced pain labels. Pain levels were categorized into baseline and 3 levels of pain intensity (1-3). RESULTS: The multimodal pain recognition models achieved an average balanced accuracy of over 80% across the different pain levels. RR models consistently outperformed other single modalities, particularly for lower pain intensities, while facial muscle activity (electromyogram) was most effective for distinguishing higher pain intensities. Although single-modality models, especially RR, generally provided higher performance compared to multimodal approaches, our multimodal framework still delivered results that surpassed most previous works in terms of overall accuracy. CONCLUSIONS: This study presents a novel, multimodal machine learning framework for objective pain recognition in postoperative patients. The results highlight the potential of integrating multiple biosignal modalities for more accurate pain assessment, with particular value in real-world clinical settings.

BACKGROUND:  Data exploration in modern electronic health records (EHRs) is often aided by user-friendly graphical interfaces providing "self-service" tools for end users to extract data for quality improvement, patient safety, and research without prerequisite training in database querying. Other resources within the same institution, such as Honest Brokers, may extract data sourced from the same EHR but obtain different results leading to questions of data completeness and correctness. OBJECTIVES:  Our objectives were to (1) examine the differences in aggregate output generated by a "self-service" graphical interface data extraction tool and our institution's clinical data warehouse (CDW), sourced from the same database, and (2) examine the causative factors that may have contributed to these differences. METHODS:  Aggregate demographic data of patients who received influenza vaccines at three static clinics and three drive-through clinics in similar locations between August 2020 and December 2020 was extracted separately from our institution's EHR data exploration tool and our CDW by our organization's Honest Brokers System. We reviewed the aggregate outputs, sliced by demographics and vaccination sites, to determine potential differences between the two outputs. We examined the underlying data model, identifying the source of each database. RESULTS:  We observed discrepancies in patient volumes between the two sources, with variations in demographic information, such as age, race, ethnicity, and primary language. These variations could potentially influence research outcomes and interpretations. CONCLUSION:  This case study underscores the need for a thorough examination of data quality and the implementation of comprehensive user education to ensure accurate data extraction and interpretation. Enhancing data standardization and validation processes is crucial for supporting reliable research and informed decision-making, particularly if demographic data may be used to support targeted efforts for a specific population in research or quality improvement initiatives.

Background: Case studies have shown ChatGPT can run clinical simulations at the medical student level. However, no data have assessed ChatGPT's reliability in meeting desired simulation criteria such as medical accuracy, simulation formatting, and robust feedback mechanisms. Objective: This study aims to quantify ChatGPT's ability to consistently follow formatting instructions and create simulations for preclinical medical student learners according to principles of medical simulation and multimedia educational technology. Methods: Using ChatGPT-4 and a prevalidated starting prompt, the authors ran 360 separate simulations of an acute asthma exacerbation. A total of 180 simulations were given correct answers and 180 simulations were given incorrect answers. ChatGPT was evaluated for its ability to adhere to basic simulation parameters (stepwise progression, free response, interactivity), advanced simulation parameters (autonomous conclusion, delayed feedback, comprehensive feedback), and medical accuracy (vignette, treatment updates, feedback). Significance was determined with χ² analyses using 95% CIs for odds ratios. Results: In total, 100% (n=360) of simulations met basic simulation parameters and were medically accurate. For advanced parameters, 55% (200/360) of all simulations delayed feedback, while the Correct arm (157/180, 87%) delayed feedback was significantly more than the Incorrect arm (43/180, 24%; P<.001). A total of 79% (285/360) of simulations concluded autonomously, and there was no difference between the Correct and Incorrect arms in autonomous conclusion (146/180, 81% and 139/180, 77%; P=.36). Overall, 78% (282/360) of simulations gave comprehensive feedback, and there was no difference between the Correct and Incorrect arms in comprehensive feedback (137/180, 76% and 145/180, 81%; P=.31). ChatGPT-4 was not significantly more likely to conclude simulations autonomously (P=.34) and provide comprehensive feedback (P=.27) when feedback was delayed compared to when feedback was not delayed. Conclusions: These simulations have the potential to be a reliable educational tool for simple simulations and can be evaluated by a novel 9-part metric. Per this metric, ChatGPT simulations performed perfectly on medical accuracy and basic simulation parameters. It performed well on comprehensive feedback and autonomous conclusion. Delayed feedback depended on the accuracy of user inputs. A simulation meeting one advanced parameter was not more likely to meet all advanced parameters. Further work must be done to ensure consistent performance across a broader range of simulation scenarios.