Conférence sur l’intelligence artificielle : comprendre ses mécanismes, enjeux éthiques et souveraineté numérique en France et en Europe.

Background: With 28%-35% of individuals aged 65 years and older experiencing incidents of falling, falls are the second leading cause of unintentional injury–related deaths globally. Limited availability of clinical staff often impedes the timely detection and prevention of potential falls. Advances in artificial intelligence (AI) could complement existing fall risk assessment and help better allocate nursing care resources. Yet, many studies are based on small datasets from a single institution, which can restrict the generalizability of the model, and do not investigate important aspects in AI model development, such as fairness across demographic groups. Objective: This study aimed to provide a comprehensive empirical evaluation of the potential of AI in nursing care, focusing on the case of fall risk prediction. To account for demographic and contextual differences in fall incidences, we analyze data from a university and a geriatric hospital in Germany. To the best of our knowledge, these are the largest fall risk prediction datasets to date with heterogeneous data distributions. We focus on 3 key objectives. First, does AI help in improving fall risk prediction? Second, how can AI models be trained safely across different hospitals? Finally, are these models fair? Methods: This study used 2 datasets for fall risk prediction: one from a university hospital with 931,726 participants, 10,442 of whom experienced falls, and another from a geriatric hospital with 12,773 participants, 1728 of whom have fallen. State-of-the-art AI models were trained with 3 approaches, including 2 decentralized learning paradigms. First, separate models were trained on data from each hospital; second, models were retrained on the respective other dataset; and federated learning (FL) was applied to both datasets. The performance of these models was compared with the rule-based systems as implemented in clinical practice for fall risk prediction. Additional analyses were conducted to test for model fairness. Results: Our findings demonstrate that AI models consistently outperform rule-based systems across all experimental setups, with the area under the receiver operating characteristic curve of 0.735 (90% CI 0.727-0.744) for the geriatric hospital, and 0.926 (90% CI 0.924-0.928) for the university hospital. FL did not improve the fall risk prediction in this setting. Our fairness analysis ruled out disparities in model performance between different sex groups, but we found fairness infringements across age groups. Conclusions: This study demonstrates that AI models consistently outperform traditional rule-based systems across heterogeneous datasets in predicting fall risk. However, it also reveals the challenges related to demographic shifts and label distribution imbalances, which limited the FL models’ ability to generalize. While the fairness analysis indicated fair results across sex subgroups, age-related disparities emerged. Addressing data imbalances and ensuring broader representation across demographic groups will be crucial for developing more fair and generalizable models.

Background: #Digitalisation is seen as a way to reduce the negative environmental impacts of healthcare production, but research is still limited. Objective: Our main focus is on the assessment of sustainability aspects of #Digital services in home care and health care. This #Study demonstrates the approach to identify the climate impacts and social impacts – both positive and negative – on a selection of #Digital home care and healthcare services, such as medicine robot services for older home care clients, through two Finnish case studies. Methods: Impacts are identified in interviews and statistical data collected from public service providers and technology suppliers using both quantitative and qualitative assessments. Results: While a well-planned and well-implemented #Digital service is likely to be a climate-friendly option, every #Digitalisation action carries at least some negative impacts. The design, architecture and practical implementation of these services greatly affect their climate and social impacts. Conclusions: This #Study employs a novel combination of impact assessment methods, highlighting the importance of qualitative understanding alongside quantitative approaches for interpreting results, especially when numerical data are limited. Advocating for multi-method impact assessments is crucial to properly capturing the service context and promoting holistic sustainability thinking.

Background: Telemonitoring can be implemented using either centralized or distributed organizational models. However, few published studies explore which conditions make one model preferable over the other, or how to choose between these two. Objective: This study investigated the decision-making factors across several domains (e.g. technological, personal, organizational) when selecting the telemonitoring model. Methods: We conducted a multiple case study across four purposively sampled hospitals to gain a range of perspectives on organizational models for telemonitoring. Selection criteria included: 1) type of organizational model, 2) type of collaborating partners, 3) task division of handling notifications and 4) it had to be implemented at scale, rather than being in an exploratory phase. Data was collected in a document study, 13 semi-structured interviews, and focus group. The topic list was based on the domains of the NASSS (non-adoption, abandonment, scale-up, spread, and sustainability) framework. Interviewees (n=13) were five project leaders, two tele-nurses, four #healthcare professionals, and two clinical informaticians. Data analysis was performed iteratively and included reflective thematic analysis. A member-checking focus group was organized to verify and reflect on the findings. Results: Various preferential factors based on the seven domains of the NASSS framework were explored for both centralized and distributed telemonitoring models: 1) Condition: The choice of objective, usually based on organizational strategy, determines whether telemonitoring will be centralized or distributed. 2) Technology: The preference for a model is determined by the anticipated number of notifications the application generates for a specific #patient group. 3) Value Proposition: The perceived cost-effectiveness and overall value to the #patient shape the value proposition for each model. 4) Adopters: The new role of tele-nurse emerged in centralized monitoring centers (CMCs), necessitating new competencies, task redistribution, and shifts in responsibility. The importance of trust among staff became evident in the context of task redistribution. 5) Organization: CMCs are typically organized regionally, in partnerships or network arrangements, which can be time-consuming yet offer significant potential for impact. 6) Wider System: The existing Dutch reimbursement system does not incentivize CMCs because the payment structure is still based on a per-treatment model. 7) Adaption Over Time: With advancements in technology, including artificial intelligence (#AI), organizing telemonitoring through CMCs is likely to gain popularity. Conclusions: Our study highlights that when decision-makers are choosing which telemonitoring model—centralized and/or distributed—to implement in their organization, deciding on the suitability of the model depends on multiple contextual factors. Our findings illustrate that decisions made for #patient group selection, technology design and value proposition significantly influence each other. It is therefore crucial for decision-makers to understand these interactions and corresponding dynamics to better align their strategies with the operational realities of their organization.

Background: During the #COVID19 #coronavirus pandemic, the continuous spread of misinformation on the internet posed an ongoing threat to public trust and understanding of epidemic #Prevention policies. Although the pandemic is now under control, information regarding the risks of long-term #COVID19 #coronavirus effects and reinfection still needs to be integrated into #COVID19 #coronavirus policies. Objective: This #Study aims to develop a robust and generalizable deep learning framework for detecting misinformation related to the prolonged impacts of #COVID19 #coronavirus by integrating pretrained language models (PLMs) with an innovative fuzzy rank-based ensemble approach. Methods: A comprehensive dataset comprising 566 genuine and 2361 fake samples was curated from reliable open sources and processed using advanced techniques. The dataset was randomly split using the scikit-learn package to facilitate both training and evaluation. Deep learning models were trained for 20 epochs on a Tesla T4 for hierarchical attention networks (HANs) and an RTX A5000 (for the other models). To enhance performance, we implemented an ensemble learning strategy that incorporated a reparameterized Gompertz function, which assigned fuzzy ranks based on each model’s prediction confidence for each test case. This method effectively fused outputs from state-of-the-art PLMs such as robustly optimized bidirectional encoder representations from transformers pretraining approach (RoBERTa), decoding-enhanced bidirectional encoder representations from transformers with disentangled attention (DeBERTa), and XLNet. Results: After training on the dataset, various classification methods were evaluated on the test set, including the fuzzy rank-based method and state-of-the-art large language models. Experimental results reveal that language models, particularly XLNet, outperform traditional approaches that combine term frequency–inverse document frequency features with support vector machine or utilize deep models like HAN. The evaluation metrics—including accuracy, precision, recall, F1-score, and area under the curve (AUC)—indicated a clear performance advantage for models that had a larger number of parameters. However, this #Study also highlights that model architecture, training procedures, and optimization techniques are critical determinants of classification effectiveness. XLNet’s permutation language modeling approach enhances bidirectional context understanding, allowing it to surpass even larger models in the bidirectional encoder representations from transformers (BERT) series despite having relatively fewer parameters. Notably, the fuzzy rank-based ensemble method, which combines multiple language models, achieved impressive results on the test set, with an accuracy of 93.52%, a precision of 94.65%, an F1-score of 96.03%, and an AUC of 97.15%. Conclusions: The fusion of ensemble learning with PLMs and the Gompertz function, employing fuzzy rank-based methodology, introduces a novel prediction approach with prospects for enhancing accuracy and reliability. Additionally, the experimental results imply that training solely on textual content can yield high prediction accuracy, thereby providing valuable insights into the optimization of fake news detection systems. These findings not only aid in detecting misinformation but also have broader implications for the application of advanced deep learning techniques in #PublicHealth #Policy and communication.

Date Submitted: Apr 17, 2025. Open Peer Review Period: May 2, 2025 - Jun 27, 2025.

Background: The rising demand for advanced home care services, driven by an aging population and the preference for aging in place, presents both challenges and opportunities. While advanced home care can improve cost-effectiveness and patient outcomes, gaps remain in understanding how eHealth technologies can optimize these services. eHealth tools have the potential to offer personalized, coordinated care that increases patient engagement. However, research exploring health care professionals’ (HCPs) perspectives on the use of eHealth tools in advanced home care and their impact on the HCP-patient relationship is limited. Objective: This study aims to explore HCPs’ perspectives on using eHealth tools in advanced home care and these tools’ impact on HCP-patient relationships. Methods: In total, 20 HCPs from 9 clinics specializing in advanced home care were interviewed using semistructured interviews. The discussions focused on their experiences with 2 eHealth tools: a mobile documentation tool and a mobile preconsultation form. The data were analyzed using content analysis to identify recurring themes. Results: The data analysis identified one main theme: optimizing health care with eHealth; that is, enhancing care delivery and overcoming challenges for future health care. Two subthemes emerged: (1) enhancing care delivery, collaboration, and overcoming adoption barriers and (2) streamlining implementation and advancing eHealth tools for future health care delivery. Five categories were also identified: (1) positive experiences and benefits, (2) interactions between HCPs and patients, (3) challenges and difficulties with eHealth tools, (4) integration into the daily workflow, and (5) future directions. Most HCPs expressed positive experiences with the mobile documentation tool, highlighting improved efficiency, documentation quality, and patient safety. While all found the mobile preconsultation form beneficial, patient-related factors limited its utility. Regarding HCP-patient relationships, interactions with patients remained unchanged with the implementation of both tools. HCPs successfully maintained their interpersonal skills and patient-centered approach while integrating eHealth tools into their practice. The tools allowed more focused, in-depth discussions, enhancing patient engagement without affecting relationships. Difficulties with the tools originated from tool-related issues, organizational challenges, or patient-related complexities, occasionally affecting the time available for direct patient interaction. Conclusions: The study underscores the importance of eHealth tools in enhancing advanced home care while maintaining the HCP-patient relationship. While eHealth tools modify care delivery techniques, they do not impact the core dynamics of the relationships between HCPs and patients. While most of the HCPs in the study had a positive attitude toward using the eHealth tools, understanding the challenges they encounter is crucial for improving user acceptance and success in implementation. Future development should focus on features that not only improve efficiency but also actively enhance HCP-patient relationships, such as facilitating more meaningful interactions and supporting personalized care in the advanced home care setting.

In this blog post, we reflect on the progress we’ve made at ORCID, and how we've remained true to the vision set forth in our earliest days.

Date Submitted: Mar 17, 2025. Open Peer Review Period: Mar 19, 2025 - May 14, 2025.

Cardiac surgery involves surgical treatments for heart and great vessel disorders, including procedures like coronary artery bypass grafting, valve re…

Journal of Medical Internet Research AI, 2023. Recognizes the imperative to strengthen AI and ML literacy in underserved communities and build a ...

The University of Sheffield leads the UK in engineering research income and investment, according to the latest figures from the Higher Education Statistics Agency (HESA).

The U.S. healthcare system is facing an alarming crisis: a severe shortage of hospital staff, particularly nurses. According to the American Association of Colleges of Nursing (AACN), the demand for r...

The Health Innovation Network is the Academic Health Science Network (AHSN) for south London, one of 15 AHSNs across England.

Date Submitted: Feb 5, 2025. Open Peer Review Period: Feb 17, 2025 - Apr 14, 2025.