Delaware journal of public health最新文献

{"title":"When Collaborative Robots Meet the Bedside: Nurses Informing Emerging Artificial Intelligence Technology.","authors":"Susan D Smith, Danielle Weber","doi":"10.32481/djph.2026.03.04","DOIUrl":"10.32481/djph.2026.03.04","url":null,"abstract":"<p><p>Across the United States, clinicians working in the acute care hospital settings continue to face persistent workforce shortages. As artificial intelligence (AI) becomes increasingly integrated into healthcare delivery, it is critical to explore technology-enabled solutions that reduce non-clinical workload without compromising care. At ChristianaCare, a nurse-led team of executive and clinical leaders, researchers, and informaticians, launched a three-year, grant funded collaborative robot (cobot) pilot program starting in 2022 to evaluate whether cobots could offload repetitive, time-consuming non-clinical tasks from staff. At the conclusion of the grant period in 2025, the findings demonstrated operational value in select high-volume workflows, such as non-urgent medication and equipment deliveries. However, the pilot also revealed that current cobot capabilities were unable to meaningfully augment nurses' complex and often time-sensitive patient care workflows. These findings underscore the need for engineers and scientists to partner closely with nurses and frontline hospital staff throughout the design and implementation processes to ensure AI powered cobots deliver high impact, workforce-supporting solutions. Additionally, healthcare leaders should not underestimate their role in facilitating these collaborations and removing organizational barriers that could influence operational success.</p>","PeriodicalId":72774,"journal":{"name":"Delaware journal of public health","volume":"12 1","pages":"10-11"},"PeriodicalIF":0.0,"publicationDate":"2026-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13048758/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147629411","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Bias Patterns in the Application of LLMs for Clinical Decision Support: A Comprehensive Study.","authors":"Raphael Poulain, Farzana Islam Adiba, Hamed Fayyaz, Rahmatollah Beheshti","doi":"10.32481/djph.2026.03.10","DOIUrl":"10.32481/djph.2026.03.10","url":null,"abstract":"<p><strong>Objectives: </strong>To investigate the extent to which Large Language Models (LLMs) exhibit social bias based on protected patient attributes and to determine how design choices, such as architecture and prompting strategies, influence these observed biases in clinical decision support.</p><p><strong>Methods: </strong>We evaluated eight popular LLMs, including general-purpose and clinically trained models, across three standardized question-answering datasets using clinical vignettes. We employed red-teaming strategies to analyze the impact of demographics on LLM outputs and compared various prompting techniques, including Zero-shot and Chain of Thought.</p><p><strong>Results: </strong>Our experiments reveal various disparities across protected groups. Notably, larger models were not necessarily less biased, and medical fine-tuning did not consistently outperform general-purpose models. Furthermore, specific prompt phrasing significantly influenced bias patterns, whereas reflection-type approaches like Chain of Thought effectively reduced biased outcomes.</p><p><strong>Conclusions: </strong>LLMs demonstrate significant social biases in clinical scenarios that are influenced by model architecture and prompt engineering. These findings highlight the critical need for rigorous evaluation and enhancement of LLMs before their integration into clinical decision support systems. Consistent with prior studies, we call for additional scrutiny to ensure equity in AI-driven healthcare applications. All code and data are available at https://github.com/healthylaife/FairCDSLLM.</p>","PeriodicalId":72774,"journal":{"name":"Delaware journal of public health","volume":"12 1","pages":"54-67"},"PeriodicalIF":0.0,"publicationDate":"2026-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13048765/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147629318","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"From the Guest Editors: Artificial Intelligence and Big Data in the Health Sciences.","authors":"Weisong Shi, Yixiang Deng","doi":"10.32481/djph.2026.03.02","DOIUrl":"10.32481/djph.2026.03.02","url":null,"abstract":"","PeriodicalId":72774,"journal":{"name":"Delaware journal of public health","volume":"12 1","pages":"4"},"PeriodicalIF":0.0,"publicationDate":"2026-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13048752/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147629370","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Human-AI Cooperation in Healthcare and Rehabilitation.","authors":"Austin Brockmeier, Panagiotis Artemiadis, Hacene Boukari, Chris Callison-Burch, Eric Eaton, Joel Harley, Thomas Powers, Jose C Principe, Darcy Reisman, Cathy Wu","doi":"10.32481/djph.2026.03.06","DOIUrl":"10.32481/djph.2026.03.06","url":null,"abstract":"<p><p>Rehabilitation after injury or to manage chronic health conditions requires continuous reassessment and intervention across time scales ranging from seconds to months. Advances in sensors and data collection, coupled with new technology to administer interventions, create numerous possibilities-including at-home care. The increased capabilities enable automated analysis and control using artificial intelligence (AI). In this essay, we analyze the need, the potential and the requirements for an intense and enduring physical human-AI cooperation framework, i.e., a symbiosis, where both AI and humans contribute to realize improved solutions. The focus is the development of knowledge and expertise to realize a new generation of AI-enabled therapy for the next decades. With an aging population, prevalence of stroke and chronic diseases, there is a demand for more efficient and effective rehabilitation powered by human-AI cooperation, especially in cases that enable remote participation in areas with limited access. This essay analyzes how the potential for advances in human-AI cooperation can impact rehabilitation in Delaware.</p>","PeriodicalId":72774,"journal":{"name":"Delaware journal of public health","volume":"12 1","pages":"20-27"},"PeriodicalIF":0.0,"publicationDate":"2026-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13082460/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147983988","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Autonomous Wheelchairs Deployment in Healthcare Facilities: Requirements and Challenges.","authors":"Mingyu Guo, Weisong Shi","doi":"10.32481/djph.2026.03.05","DOIUrl":"10.32481/djph.2026.03.05","url":null,"abstract":"<p><p>Deployed autonomous wheelchairs are reliable when autonomy is constrained to bounded domains and predefined destinations, but this design choice leaves key clinical requirements unmet. In health facilities, the primary barriers are not only navigation performance, but also robust interaction with building infrastructure (doors, elevators, access control), socially and operationally appropriate behavior in crowded corridors, and safety-centered fallback when perception or planning degrades. This vision paper characterizes these recurring gaps and distills a requirements agenda for next-generation autonomous wheelchairs: (1) building and workflow compatibility as first-class subsystems; (2) explicit recovery and caregiver-in-the-loop modes; (3) semantic \"last-meter\" goal grounding for user-referenced destinations; and (4) on-device, timing-predictable architectures aligned with privacy and scalable deployment. To make these requirements actionable, at the CAR lab at UD, we present SWee, a working prototype that emphasizes clinical deployability through an on-device, modular edge-box architecture (VOCAR) integrating onboard sensing and language-grounded goal specification.</p>","PeriodicalId":72774,"journal":{"name":"Delaware journal of public health","volume":"12 1","pages":"12-18"},"PeriodicalIF":0.0,"publicationDate":"2026-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13048751/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147629382","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Pioneering the Nation's First Nursing Research Fellowship in Robotics and Innovation.","authors":"Susan D Smith, Kathryn Shady, Briana Abernathy, Morgan Tallo","doi":"10.32481/djph.2026.03.03","DOIUrl":"10.32481/djph.2026.03.03","url":null,"abstract":"<p><strong>Objective: </strong>Clinical nurse attrition from the bedside calls for innovative professional development strategies that diversify skills and support wellbeing and retention. To address this issue, the largest health system in Delaware implemented the nation's first Nursing Research Fellowship in Robotics and Innovation using external grant funding.</p><p><strong>Programmatic methods: </strong>Following a competitive application pool, four bachelors-prepared clinical nurses were selected from two hospital campuses across four diverse practice areas. This eight-month, paid fellowship grounded in adult learning theory combined weekly didactic instruction with mentored, hands-on research in a structured, collaborative, and independent format. The nurse fellows serve as co-investigators on an IRB approved robotics study. Longitudinal pre-, mid-, and post-fellowship surveys assessed knowledge acquisition, program experience, and well-being.</p><p><strong>Programmatic results: </strong>Nurse fellows demonstrated gains in research competencies and specialty areas that included protocol development, informatics, artificial intelligence, robotics, and techquity. All fellows reported increased job satisfaction, improved psychological wellbeing, enhanced professional confidence, and intent to remain at the bedside. Scholarly outcomes included multiple accepted national and regional conference abstracts, published commentary articles, and co-authorship of an original research manuscript.</p><p><strong>Conclusions: </strong>This novel fellowship effectively integrated research education, innovation, and paid protected time to strengthen clinical nurses' research capability, professional fulfillment, and retention to the bedside. This program offers a replicable model for advancing nursing workforce wellbeing through immersive, mentored research experiences.</p>","PeriodicalId":72774,"journal":{"name":"Delaware journal of public health","volume":"12 1","pages":"6-9"},"PeriodicalIF":0.0,"publicationDate":"2026-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13048760/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147629315","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Perspective: Delaware's Vision for Responsible Innovation in Health Care.","authors":"Christen Linke Young","doi":"10.32481/djph.2026.03.18","DOIUrl":"10.32481/djph.2026.03.18","url":null,"abstract":"","PeriodicalId":72774,"journal":{"name":"Delaware journal of public health","volume":"12 1","pages":"128-129"},"PeriodicalIF":0.0,"publicationDate":"2026-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13048764/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147629375","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"How Strong Data Infrastructure Could Transform Delaware's Firearm Violence Prevention Ecosystem.","authors":"Lauren Footman, Danielle Fisher, Alexandra Wynn","doi":"10.32481/djph.2026.03.12","DOIUrl":"10.32481/djph.2026.03.12","url":null,"abstract":"<p><p>Firearm violence is a persistent and preventable public health crisis in Delaware, resulting in avoidable loss of life, long-term injury, and community harm. Over the past decade, firearm violence has become prevalent across the state and disproportionately in communities with historic disinvestment and limited access to protective factors. Trends in firearm violence have shifted over time in Delaware, with periods of improvement in some jurisdictions occurring alongside rising violence in others. This context makes clear the need to treat firearm violence as a population-level health issue that requires timely and accessible data, coordinated systems, and intervention and prevention strategies beyond traditional criminal justice responses. This commentary examines Delaware's current capacity to address community violence intervention (CVI) through a public health lens, with a focus on the role of data infrastructure. The commentary argues that Delaware's fragmented data systems jeopardize the state's ability to effectively prevent and respond to firearm violence. By outlining the consequences of these gaps and drawing lessons from surrounding states, this provides best practices for strengthening Delaware's data infrastructure to support reductions in firearm violence statewide.</p>","PeriodicalId":72774,"journal":{"name":"Delaware journal of public health","volume":"12 1","pages":"92-95"},"PeriodicalIF":0.0,"publicationDate":"2026-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13048761/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147629338","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Improving Context-Aware Personalized Nudging: Using Wearable Sensors to Reduce Sedentary Behavior.","authors":"Tanvir Rahman, Ajith Vemuri, Cora J Firkin, Barry Bodt, Elizabeth Orsega-Smith, Gregory M Dominick, Keith Decker","doi":"10.32481/djph.2026.03.14","DOIUrl":"10.32481/djph.2026.03.14","url":null,"abstract":"<p><strong>Objectives: </strong>To improve nudge outcome classification accuracy in a context-aware personalized nudging framework using wearable sensor data targeted to reduce sedentary behavior using Just- in-Time Adaptive Interventions (JITAIs).</p><p><strong>Methods: </strong>Data were collected using a custom smartwatch application in a free-living observational study conducted at the University of Delaware (Newark, Delaware, USA) between Spring 2021 and Fall 2022. A total of 18 participants were enrolled. The system continuously recorded motion, physiological, and contextual data and delivered adaptive behavioral prompts. A decision- tree model was trained using sitting and walking bouts enriched with contextual features such as time, location, physiological state, and prior intervention outcomes. Behavioral responses were automatically evaluated using sensor-derived outcomes.</p><p><strong>Results: </strong>The proposed model improved classification accuracy for nudge outcomes from 0.42 to 0.78 across 787 sitting bouts. A walking-nudge model achieved an accuracy of 0.70 on 207 walking bouts. Nudged walking bouts were longer in duration, covered greater distances, and exhibited higher average speeds than non-nudged bouts.</p><p><strong>Conclusions: </strong>Context-aware adaptive nudging can improve both the timing and behavioral effectiveness of wearable-based interventions. Incorporating contextual and historical features enables personalized and behaviorally meaningful intervention delivery.</p><p><strong>Policy implications: </strong>Wearable-based adaptive interventions offer a scalable and cost-effective strategy to reduce sedentary behavior and support population-level health promotion.</p>","PeriodicalId":72774,"journal":{"name":"Delaware journal of public health","volume":"12 1","pages":"106-112"},"PeriodicalIF":0.0,"publicationDate":"2026-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13048762/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147629307","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Beyond Cognitive Load: AI-Based Estimation of Cognitive Effort Using Brain Signals During Digital Tasks.","authors":"Shayla Sharmin, Mohammad Fahim Abrar, Gael Lucero-Palacios, Aditya Raikwar, Roghayeh Leila Barmaki","doi":"10.32481/djph.2026.03.08","DOIUrl":"10.32481/djph.2026.03.08","url":null,"abstract":"<p><strong>Background: </strong>Cognitive effort, defined as the relationship between cognitive load and task performance, offers insight into how individuals efficiently allocate mental resources during cognitively demanding activities. This metric is crucial in high-stakes public health and clinical training, where unmanaged cognitive overload has been linked to medical errors and workforce burnout. This study aims to examine whether cognitive effort varies systematically across task segments and whether it can be estimated at the individual level using brain signal data and machine learning.</p><p><strong>Method: </strong>Functional near-infrared spectroscopy (fNIRS) data were collected from 16 participants during a structured digital cognitive task comprising four sequential segments separated by short and long rest intervals. Cognitive effort was defined through relative neural efficiency and relative neural involvement, which combined measures of prefrontal hemodynamic activity with task performance. The analysis followed a two-stage approach. First, a segment-level group analysis assessed whether cognitive effort differed significantly across predefined task segments, thereby confirming that the task structure produced meaningful changes in cognitive demand. Second, participant-independent machine learning models predicted task performance from brain signal features. These predicted performance scores were then combined with neural measures to estimate cognitive effort at the individual level.</p><p><strong>Results: </strong>First, statistical analysis showed significant differences in cognitive effort across the four task segments. This confirms that even small changes in the assessment structure impact on the collective cognitive efficiency of the trainees. Then, we used machine learning on fNIRS data to predict individual performance scores. We found that the effort calculated from these predicted scores was nearly identical to that calculated from the actual scores. This result suggests that our effort metric is strongly based on brain signals.</p><p><strong>Conclusion: </strong>The findings demonstrate the feasibility of estimating cognitive effort from brain signals using artificial intelligence at both group and individual levels.</p><p><strong>Public health implications: </strong>Estimation of cognitive effort from digital task data may support scalable monitoring of cognitive workload and mental fatigue in technology-mediated environments, which complements subjective assessment methods used in public and digital health research.</p>","PeriodicalId":72774,"journal":{"name":"Delaware journal of public health","volume":"12 1","pages":"32-43"},"PeriodicalIF":0.0,"publicationDate":"2026-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13082466/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147984036","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}