Shivam Shukla, Emily Chen, Mahnaz Roshanaei
et al.
There has been a growing research interest in Digital Therapeutic Alliance (DTA) as the field of AI-powered conversational agents are being deployed in mental health care, particularly those delivering CBT (Cognitive Behaviour Therapy). Our proposition argues that the current design paradigm which seeks to optimize the bond between a patient in need of support and an AI agent contains a subtle but consequential trap: it risks producing an "appearance of connection" that unintentionally disrupts the fundamental human need for relatedness, which potentially displaces the authentic human relationships upon which long-term psychological recovery depends. We propose a reorientation from designing artificial intelligence tools that simulate relationships to designing AI that scaffolds them. To operationalize our argument, we propose an interdisciplinary model that translates the Responsible AI Six Sphere Framework through the lens of Self-Determination Theory (SDT), with a specific focus on the basic psychological need for relatedness. The resulting model offers the technical and other clinical communities a set of relationship-centered design guidelines and relevant provocations for building AI systems that function not just as companions, but as a catalyst for strengthening a patient's entire relational ecology; their connections with therapists, caregivers, family, and peers. In doing so, we discuss a model towards a more sustainable ecosystem of relationship-centered AI in mental health care.
Marco Cascella, Alessandro Simonini, Sergio Coluccia
et al.
Abstract Background Burnout (BO) is a serious issue affecting professionals across various sectors, leading to adverse psychological and occupational consequences, even in anesthesiologists. Machine learning, particularly neural networks, can offer effective data-driven approaches to identifying BO risk more accurately. This study aims to develop and evaluate different artificial dense neural network (DNN)-based models to predict BO based on occupational, psychological, and behavioral factors. Methods A dataset (300 Italian anesthesiologists) comprising workplace stressors, psychological well-being indicators, and demographic variables was used to train DNN models. Model performance was measured using standard evaluation metrics, including accuracy, precision, recall, and F1 score. Statistical tests were adopted to assess differences in prediction across the DNNs. Results The best neural architecture achieved a predictive accuracy of 0.68, with key contributors to BO including workload, emotional exhaustion, job dissatisfaction, and lack of work-life balance. Despite substantial differences among the six implemented algorithms, no significant variation in prediction performance was observed. Conclusion Psychological distress scores are significantly higher in the high-risk BO group, suggesting greater anxiety, depression, and overall distress in this category. While challenges remain, continued advancements in artificial intelligence and data science promise more effective and personalized mental health care solutions. Trial registration Not applicable.
Anesthesiology, Medical emergencies. Critical care. Intensive care. First aid
Chronic illnesses are a global concern with essential hypertension and diabetes mellitus among the most common conditions. Remote patient monitoring has shown promising results on clinical and health outcomes. However, access to care and digital health solutions is limited among rural, lower-income, and older adult populations. This paper repots on a pre-post study of a comprehensive care coordination program including connected, wearable blood pressure and glucometer devices, tablets, and medical assistant-provided health coaching in a community health center in rural California. The participants (n=221) had a mean age of 54.6 years, were majority female, two-thirds spoke Spanish, 19.9% had hypertension, 49.8% diabetic, and 30.3% both conditions. Participants with hypertension achieved a mean reduction in systolic blood pressure of 20.24 (95% CI: 13.61, 26.87) at six months while those with diabetes achieved a mean reduction of 3.85 points (95% CI: 3.73, 4.88). These outcomes compare favorably to the small but growing body of evidence supporting digital care coordination and remote monitoring. These results also support the feasibility of well-designed digital health solutions yielding improved health outcomes among underserved communities.
Subhashis Das, Debashis Naskar, Sara Rodriguez Gonzalez
et al.
The rapid advancement of digital technologies and recent global pandemic scenarios have led to a growing focus on how these technologies can enhance healthcare service delivery and workflow to address crises. Action plans that consolidate existing digital transformation programs are being reviewed to establish core infrastructure and foundations for sustainable healthcare solutions. Reforming health and social care to personalize home care, for example, can help avoid treatment in overcrowded acute hospital settings and improve the experiences and outcomes for both healthcare professionals and service users. In this information-intensive domain, addressing the interoperability challenge through standards-based roadmaps is crucial for enabling effective connections between health and social care services. This approach facilitates safe and trustworthy data workflows between different healthcare system providers. In this paper, we present a methodology for extracting, transforming, and loading data through a semi-automated process using a Common Semantic Standardized Data Model (CSSDM) to create personalized healthcare knowledge graph (KG). The CSSDM is grounded in the formal ontology of ISO 13940 ContSys and incorporates FHIR-based specifications to support structural attributes for generating KGs. We propose that the CSSDM facilitates data harmonization and linking, offering an alternative approach to interoperability. This approach promotes a novel form of collaboration between companies developing health information systems and cloud-enabled health services. Consequently, it provides multiple stakeholders with access to high-quality data and information sharing.
Beatriz Severes, Ana O. Henriques, Rory Clark
et al.
Academic well-being is deeply influenced by peer-support networks, yet they remain informal, inequitable, and unsustainable, often relying on personal connections and social capital rather than structured, inclusive systems. Additionally, institutional well-being responses frequently focus on student populations, neglecting the emotional labour of faculty and staff, reinforcing an exclusionary academic culture. Drawing on HCI methodologies, participatory design, and care ethics, this workshop will provide a space for rethinking how academic communities can support inclusive networks. Through pre-workshop engagement, co-design activities, and reflection, participants will examine systemic gaps in networks and explore ways to embed care, equity, and sustainability into academic peer-support frameworks -- from informal, exclusionary models to structured, inclusive care-based ecosystems. At the end of the workshop, participants will co-develop design strategies for integrating care and resilience in academic ecosystems, resources for designing equitable support systems, and a peer network invested and committed to fostering a supportive academic community.
The multivariate, asynchronous nature of real-world clinical data, such as that generated in Intensive Care Units (ICUs), challenges traditional AI-based decision-support systems. These often assume data regularity and feature independence and frequently rely on limited data scopes and manual feature engineering. The potential of generative AI technologies has not yet been fully exploited to analyze clinical data. We introduce ICU-BERT, a transformer-based model pre-trained on the MIMIC-IV database using a multi-task scheme to learn robust representations of complex ICU data with minimal preprocessing. ICU-BERT employs a multi-token input strategy, incorporating dense embeddings from a biomedical Large Language Model to learn a generalizable representation of complex and multivariate ICU data. With an initial evaluation of five tasks and four additional ICU datasets, ICU-BERT results indicate that ICU-BERT either compares to or surpasses current performance benchmarks by leveraging fine-tuning. By integrating structured and unstructured data, ICU-BERT advances the use of foundational models in medical informatics, offering an adaptable solution for clinical decision support across diverse applications.
Background Advances in emergency and critical care have improved outcomes, but gaps in communication and decision-making persist, especially in the emergency department (ED), prompting the development of a checklist to aid in serious illness conversations (SIC) in China. Methods This was a single-centre prospective interventional study on the quality improvement of SIC for life-sustaining treatment (LST). The study recruited patients consecutively for both its observational baseline and interventional stages until its conclusion. Eligible participants were adults over 18 years old admitted to the Emergency Intensive Care Unit (EICU) of a tertiary teaching hospital, possessing full decisional capacity or having a legal proxy. Exclusions were made for pregnant women, patients deceased upon arrival, those who refused participation, and individuals with incomplete data for analysis. First, a two-round Delphi process was organized to identify major elements and generate a standard process through a checklist. Subsequently, the efficacy of SIC in adult patients admitted to the EICU was compared using the Decisional Conflict Scale (DCS) score before (baseline group) and after (intervention group) implementing the checklist. Results The study participants presented with the most common comorbidities, such as diabetes, myocardial infarction, cerebrovascular disease, moderate-to-severe renal disease, congestive heart failure, and chronic pulmonary disease. The median Charlson Index did not differ between the baseline and intervention cohorts. The median length of hospital stay was 11.0 days, and 82.9% of patients survived until hospital discharge. The total DCS score was lower in the intervention group than in the baseline group. Three subscales, including the informed, values clarity, and support subscales, demonstrated significant differences between the intervention and baseline groups. Fewer intervention group patients agreed with and changed their minds about cardiopulmonary resuscitation (CPR) compared to the baseline group. Conclusion The use of a SIC checklist in the EICU reduced the DCS score by increasing medical information disclosure, patient value awareness, and decision-making support.
Adam J. Singer, Neena S. Abraham, Latha Ganti
et al.
Abstract This manuscript is a consensus document of an expert panel on the Evaluation and Treatment of Gastrointestinal Bleeding in Patients Taking Anticoagulants Presenting to the Emergency Department, sponsored by the American College of Emergency Physicians.
Medical emergencies. Critical care. Intensive care. First aid
Andrew Piner, Spencer S. Lovegrove, Laura J. Bontempo
et al.
A 77-year-old male who presented to the emergency department with generalized weakness and worsening chronic dysphagia was diagnosed with pneumonia. Shortly after receiving vascular access for his treatment, he had a rapid change in his mental status and neurological examination.
Medical emergencies. Critical care. Intensive care. First aid
Objective: Question answering (QA) systems have the potential to improve the quality of clinical care by providing health professionals with the latest and most relevant evidence. However, QA systems have not been widely adopted. This systematic review aims to characterize current medical QA systems, assess their suitability for healthcare, and identify areas of improvement. Materials and methods: We searched PubMed, IEEE Xplore, ACM Digital Library, ACL Anthology and forward and backward citations on 7th February 2023. We included peer-reviewed journal and conference papers describing the design and evaluation of biomedical QA systems. Two reviewers screened titles, abstracts, and full-text articles. We conducted a narrative synthesis and risk of bias assessment for each study. We assessed the utility of biomedical QA systems. Results: We included 79 studies and identified themes, including question realism, answer reliability, answer utility, clinical specialism, systems, usability, and evaluation methods. Clinicians' questions used to train and evaluate QA systems were restricted to certain sources, types and complexity levels. No system communicated confidence levels in the answers or sources. Many studies suffered from high risks of bias and applicability concerns. Only 8 studies completely satisfied any criterion for clinical utility, and only 7 reported user evaluations. Most systems were built with limited input from clinicians. Discussion: While machine learning methods have led to increased accuracy, most studies imperfectly reflected real-world healthcare information needs. Key research priorities include developing more realistic healthcare QA datasets and considering the reliability of answer sources, rather than merely focusing on accuracy.
Melike Sirlanci, George Hripcsak, Cecilia C. Low Wang
et al.
Intensive care unit (ICU) patients exhibit erratic blood glucose (BG) fluctuations, including hypoglycemic and hyperglycemic episodes, and require exogenous insulin delivery to keep their BG in healthy ranges. Glycemic control via glycemic management (GM) is associated with reduced mortality and morbidity in the ICU, but GM increases the cognitive load on clinicians. The availability of robust, accurate, and actionable clinical decision support (CDS) tools reduces this burden and assists in the decision-making process to improve health outcomes. Clinicians currently follow GM protocol flow charts for patient intravenous insulin delivery rate computations. We present a mechanistic model-based control algorithm that predicts the optimal intravenous insulin rate to keep BG within a target range; the goal is to develop this approach for eventual use within CDS systems. In this control framework, we employed a stochastic model representing BG dynamics in the ICU setting and used the linear quadratic Gaussian control methodology to develop a controller. We designed two experiments, one using virtual (simulated) patients and one using a real-world retrospective dataset. Using these, we evaluate the safety and efficacy of this model-based glycemic control methodology. The presented controller avoids hypoglycemia and hyperglycemia in virtual patients, maintaining BG levels in the target range more consistently than two existing GM protocols. Moreover, this methodology could theoretically prevent a large proportion of hypoglycemic and hyperglycemic events recorded in a real-world retrospective dataset.
With the intensification of global aging, health management of the elderly has become a focus of social attention. This study designs and implements a smart elderly care service model to address issues such as data diversity, health status complexity, long-term dependence and data loss, sudden changes in behavior, and data privacy in the prediction of health behaviors of the elderly. The model achieves accurate prediction and dynamic management of health behaviors of the elderly through modules such as multimodal data fusion, data loss processing, nonlinear prediction, emergency detection, and privacy protection. In the experimental design, based on multi-source data sets and market research results, the model demonstrates excellent performance in health behavior prediction, emergency detection, and personalized services. The experimental results show that the model can effectively improve the accuracy and robustness of health behavior prediction and meet the actual application needs in the field of smart elderly care. In the future, with the integration of more data and further optimization of technology, the model will provide more powerful technical support for smart elderly care services.
Socially assistive robots are increasingly being used to support the social, cognitive, and physical well-being of those who provide care (healthcare professionals) and those in need of care (older adults). However, the effectiveness of persuasive socially assistive robot behaviors and their impact on the sustained motivation of older adults is still not well understood. This extended abstract describes our prior human-robot interaction study on investigating the effectiveness of persuasive social robot behaviors with care providers, followed by our current research assessing the impact of these persuasive robot behaviors on the well-being of older adults in long-term care. The findings provide insights into engagement and sustained motivation of older adults when providing assistance.
Marie Kristine Jessen, Anna Drescher Petersen, Hans Kirkegaard
Introduction: Sepsis is a life-threatening and common cause of Emergency department (ED) referrals. Out-of-hour staffing is limited in ED, which may potentially affect fluid administration. This study aimed to investigate fluid volume variation in out-of-hour vs. routine-hour admissions.
Methods: The present study is a post-hoc analysis of a multicentre, prospective, observational study investigating fluid administration in ED patients with suspected infection, from Jan 20th - March 2nd, 2020. Patient groups were “routine-hours” (RH): weekdays 07:00-18:59 or “out-of-hours” (OOH): weekdays 19:00-06:59 or Friday 19:00-Monday 06:59. Primary outcome was 24-hour total fluid volumes (oral + intravenous (IV)). Secondary outcomes were total fluids 0-6 hours, oral fluids 0-6 and 0-24 hours, and IV fluids 0-6 and 0-24 hours. Linear regression adjusted for site and illness severity was used.
Results: 734 patients had suspected infection; 449 were admitted during RH and 287 during OOH. Mean (95% CI) total 24-hour fluid volumes were equal in simple infection and sepsis regardless of admission time: Simple infection RH: 3640 (3410 - 3871) ml and OOH: 3681 (3451 - 3913) ml. Sepsis RH: 3671 (3443;3898) ml and OOH: 3896 (3542;4250) ml. Oral fluids 0-6h were reduced in simple infection and sepsis among OOH vs. RH. Sepsis patients received more 0-6-hour IV fluid when admitted OOH vs. RH. There were no associations between admission time and 0-24-hour oral or IV volumes in simple infection or sepsis.
Conclusion: Admission time did not have an association with 24-hour total fluid volumes. Sepsis patients admitted during OOH received more 0-6-hour IV fluids than RH patients, and simple infection and sepsis patients received less oral fluid in 0-6 hours if admitted during OOH vs. RH.
Medical emergencies. Critical care. Intensive care. First aid
Cerebral air embolism (CAE) is a rare life-threatening condition. It may clinically mimic acute ischemic stroke by decreasing cerebral perfusion pressure and brain tissue oxygenation and may cause impaired consciousness and epileptic seizures. In its etiology; iatrogenic causes such as central venous catheterization, endoscopy, sclerotherapy, major surgeries, and invasive lung interventions are mostly seen. The most useful imaging method for diagnosis is cranial computed tomography (CT). In this presentation, we present the case of a 63-year-old male patient admitted to the emergency department (ED) with complaints of mental status changes and seizures and weakness in the left upper and lower extremities on physical examination. The patient had a history of thoracentesis performed three days ago and was discharged. Brain CT of the patient showed signs of newly developing CAE, and diffusion magnetic resonance imaging showed findings consistent with right middle cerebral artery infarction in his second admission. We thought that the symptoms of the patient, who had no history of additional intervention, were due to CAE, which developed as a complication of thoracentesis. Thoracentesis is one of the invasive procedures that can be applied for diagnostic/therapeutic purposes in the ED and does not have complications such as pneumothorax, hemothorax, soft tissue infection, or intra-abdominal organ injury. Physicians should be aware that CAE may also occur in patients who develop neurological deficits after thoracentesis. The clinician’s high suspicion, prompt diagnosis, and treatment can be lifesaving.
Medicine, Medical emergencies. Critical care. Intensive care. First aid