Journal of Medical Internet Research最新文献

{"title":"Designing Clinical Decision Support Systems (CDSS)-A User-Centered Lens of the Design Characteristics, Challenges, and Implications: Systematic Review.","authors":"Andrew A Bayor, Jane Li, Ian A Yang, Marlien Varnfield","doi":"10.2196/63733","DOIUrl":"https://doi.org/10.2196/63733","url":null,"abstract":"<p><strong>Background: </strong>Clinical decision support systems (CDSS) have the potential to play a crucial role in enhancing health care quality by providing evidence-based information to clinicians at the point of care. Despite their increasing popularity, there is a lack of comprehensive research exploring their design characterization and trends. This limits our understanding and ability to optimize their functionality, usability, and adoption in health care settings.</p><p><strong>Objective: </strong>This systematic review examined the design characteristics of CDSS from a user-centered perspective, focusing on user-centered design (UCD), user experience (UX), and usability, to identify related design challenges and provide insights into the implications for future design of CDSS.</p><p><strong>Methods: </strong>This review followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) recommendations and used a grounded theory analytical approach to guide the conduct, data analysis, and synthesis. A search of 4 major electronic databases (PubMed, Web of Science, Scopus, and IEEE Xplore) was conducted for papers published between 2013 and 2023, using predefined design-focused keywords (design, UX, implementation, evaluation, usability, and architecture). Papers were included if they focused on a designed CDSS for a health condition and discussed the design and UX aspects (eg, design approach, architecture, or integration). Papers were excluded if they solely covered technical implementation or architecture (eg, machine learning methods) or were editorials, reviews, books, conference abstracts, or study protocols.</p><p><strong>Results: </strong>Out of 1905 initially identified papers, 40 passed screening and eligibility checks for a full review and analysis. Analysis of the studies revealed that UCD is the most widely adopted approach for designing CDSS, with all design processes incorporating functional or usability evaluation mechanisms. The CDSS reported were mainly clinician-facing and mostly stand-alone systems, with their design lacking consideration for integration with existing clinical information systems and workflows. Through a UCD lens, four key categories of challenges relevant to CDSS design were identified: (1) usability and UX, (2) validity and reliability, (3) data quality and assurance, and (4) design and integration complexities. Notably, a subset of studies incorporating Explainable artificial intelligence highlighted its emerging role in addressing key challenges related to validity and reliability by fostering explainability, transparency, and trust in CDSS recommendations, while also supporting collaborative validation with users.</p><p><strong>Conclusions: </strong>While CDSS show promise in enhancing health care delivery, identified challenges have implications for their future design, efficacy, and utilization. Adopting pragmatic UCD design approaches that actively involve users is essent","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e63733"},"PeriodicalIF":5.8,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144336690","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Using Social Media to Disseminate Behavior Change Interventions: Scoping Review of Systematic Reviews.","authors":"Porooshat Dadgostar, Qiuyuan Qin, Suiyue Cui, Laura Ellen Ashcraft, Reza Yousefi-Nooraie","doi":"10.2196/57370","DOIUrl":"https://doi.org/10.2196/57370","url":null,"abstract":"<p><strong>Background: </strong>Compared with implementation, the conceptual frameworks, strategies, and outcomes of efforts to disseminate behavioral interventions are less developed. We conducted a scoping review of the systematic reviews of social media strategies to disseminate behavior change interventions. We focused on the common themes in the methodology and evaluation frameworks of social media-based dissemination strategies.</p><p><strong>Objective: </strong>This scoping review aims to identify common themes in the design, delivery, and impact assessment of social media-based dissemination strategies for behavior change interventions.</p><p><strong>Methods: </strong>We searched the Epistemonikos database (until 2024) to retrieve systematic reviews on social media dissemination. A total of 2 independent reviewers screened the abstracts and full texts. We extracted and classified the data on the characteristics of the included reviews and outcome assessments. We followed the reflexive thematic analysis steps to identify the main themes of the ingredients of the social media dissemination strategies.</p><p><strong>Results: </strong>We screened 613 records based on the title and abstract, followed by the assessment of 100 full texts of potentially eligible reviews. The 43 included reviews assessed a median of 20 empirical studies (IQ range 21). The study designs, intervention strategies, and evaluation measures of social media dissemination interventions were diverse. We classified the main themes of the ingredients of social media dissemination strategies into 4 main categories: 1-way spread (aiming for spread and diffusion, with little or no effort to develop 2-way communications or engage target users in conversation and feedback; n=37), invoking conversations (facilitating and enhancing the 1-way spread using conversational and community features of social media to promote dialogue among users or between the users and experts; n=21), peer motivation (facilitate sharing individual behavior on social media to receive confirmation, feedback, and support, to further personalize the dissemination; n=11), and miscellaneous (eg, dissemination through online multiplayer games; n=3). The main outcomes of dissemination efforts were reach and engagement (n=12), user perception of their knowledge, intention to change the behavior, feasibility and acceptability of the intervention (n=24), and impact on health and health-related behaviors (n=43). The majority of theoretical frameworks that were identified by the reviews were individual and social behavior change models (including the theory of planned behavior and Social Cognitive Theories). A smaller number of reviews also identified social and contextual models (eg, Social Network Theory), dissemination and implementation frameworks (eg, Diffusion of Innovation), and social marketing and action models (eg, community mobilization and Reader-to-Leader framework).</p><p><strong>Conclusions: </strong>R","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e57370"},"PeriodicalIF":5.8,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144336693","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Using Natural Language Processing to Explore Patient Perspectives on AI Avatars in Support Materials for Patients With Breast Cancer: Survey Study.","authors":"Eleanor Cheese, Raouef Ahmed Bichoo, Kartikae Grover, Dorin Dumitru, Alexandros Zenonos, Joanne Groark, Douglas Gibson, Rebecca Pope","doi":"10.2196/70971","DOIUrl":"https://doi.org/10.2196/70971","url":null,"abstract":"<p><strong>Background: </strong>Having well-informed patients is crucial to enhancing patient satisfaction, quality of life, and health outcomes, which in turn optimizes health care use. Traditional methods of delivering information, such as booklets and leaflets, are often ineffective and can overwhelm patients. Educational videos represent a promising alternative; however, their production typically requires significant time and financial resources. Video production using generative artificial intelligence (AI) technology may provide a solution to this problem.</p><p><strong>Objective: </strong>This study aimed to use natural language processing (NLP) to understand free-text patient feedback on 1 of 7 AI-generated patient educational videos created in collaboration with Roche UK and the Hull University Teaching Hospitals NHS Trust breast cancer team, titled \"Breast Cancer Follow Up Programme.\"</p><p><strong>Methods: </strong>A survey was sent to 400 patients who had completed the breast cancer treatment pathway, and 98 (24.5%) free-text responses were received for the question \"Any comments or suggestions to improve its [the video's] contents?\" We applied and evaluated different NLP machine learning techniques to draw insights from these unstructured data, namely sentiment analysis, topic modeling, summarization, and term frequency-inverse document frequency word clouds.</p><p><strong>Results: </strong>Sentiment analysis showed that 81% (79/98) of the responses were positive or neutral, while negative comments were predominantly related to the AI avatar. Topic modeling using BERTopic with k-means clustering was found to be the most effective model and identified 4 key topics: the breast cancer treatment pathway, video content, the digital avatar or narrator, and short responses with little or no content. The term frequency-inverse document frequency word clouds indicated positive sentiment about the treatment pathway (eg, \"reassured\" and \"faultless\") and video content (eg, \"informative\" and \"clear\"), whereas the AI avatar was often described negatively (eg, \"impersonal\"). Summarization using the text-to-text transfer transformer model effectively created summaries of the responses by topic.</p><p><strong>Conclusions: </strong>This study demonstrates the success of NLP techniques in efficiently generating insights into patient feedback related to generative AI educational content. Combining NLP methods resulted in clear visuals and insights, enhancing the understanding of patient feedback. Analysis of free-text responses provided clinicians at Hull University Teaching Hospitals NHS Trust with deeper insights than those obtained from quantitative Likert scale responses alone. Importantly, the results validate the use of generative AI in creating patient educational videos, highlighting its potential to address the challenges of costly video production and the limitations of traditional, often overwhelming educational leaflets. Despite the positiv","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e70971"},"PeriodicalIF":5.8,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144336692","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Impact of Virtual Reality-Based Biofeedback on Sleep Quality Among Individuals With Depressive Symptoms, Anxiety Symptoms, or Both: 4-Week Randomized Controlled Study.","authors":"Sisu Seong, Hyewon Kim, Yaehee Cho, Min-Ji Kim, Ka Ram Park, Jooeun Choi, Seonah Lee, Dong Jun Kim, Seog Ju Kim, Hong Jin Jeon","doi":"10.2196/65772","DOIUrl":"https://doi.org/10.2196/65772","url":null,"abstract":"<p><strong>Background: </strong>Use of virtual reality (VR)-based biofeedback (BF) represents an emerging nonpharmacological intervention for enhancing sleep quality in individuals exhibiting depressive symptoms, anxiety symptoms, or both. However, empirical evidence regarding its efficacy in addressing sleep disturbances remains limited and inconclusive.</p><p><strong>Objective: </strong>This 3-arm randomized controlled trial aimed (1) to compare the efficacy of VR-based BF with conventional BF in improving sleep quality, as measured by the Pittsburgh Sleep Quality Index (PSQI), among individuals with depressive symptoms, anxiety symptoms, or both (DAS); (2) to examine the effects of VR-based BF in a demographically similar healthy control (HC) group; and (3) to evaluate between-group differences in sleep quality improvements at the 4-week follow-up.</p><p><strong>Methods: </strong>Participants scoring ≥10 on the Patient Health Questionnaire-9 or ≥9 on the Panic Disorder Severity Scale were allocated to a group with DAS while others were assigned to a HC group. The DAS group was subsequently randomized into VR-based BF or conventional BF interventions with a therapist. All participants attended sessions at weeks 0, 2, and 4, completing assessments including the Montgomery-Asberg Depression Rating Scale, State-Trait Anxiety Inventory, and Visual Analog Scale in interviews. The PSQI was administered at baseline and postintervention to evaluate alterations in sleep quality over a 4-week period.</p><p><strong>Results: </strong>A total of 118 participants were randomized into a VR-based BF group (DAS/VR, n=40) or a conventional BF group (DAS/BF, n=38), and a control group (HC/VR, n=40) received VR-based BF. Sleep disturbance scores of both DAS/VR and DAS/BF groups had significant improvements (mean reductions of -0.58, SD 0.75 and -0.66, SD 0.75, respectively) compared to those preintervention, showing no significant difference after adjusting for age and sex (P=.49). The DAS/VR group had a greater improvement in sleep disturbance (mean -0.08, SD 0.53; P=0.0014) than the HC/VR group. Global PSQI scores in both DAS/VR and DAS/BF groups improved compared to those preintervention, showing decreases by -2.50 (SD 2.89) and -3.39 (SD 2.80), respectively. The difference between the 2 groups was not statistically significant (P=.14). The Global PSQI score in the DAS/VR group showed significant improvement (-0.95, SD 2.09; P=.01) compared to that in the HC/VR group.</p><p><strong>Conclusions: </strong>This study provides evidence that both VR-based BF and conventional BF with a therapist are efficacious psychological interventions for enhancing sleep quality in individuals with depressive symptoms, anxiety symptoms, or both, with no significant differences observed between these 2 approaches. Both interventions showed significant improvements compared to baseline measurements. These findings suggest potential applications of these interventions in clinical se","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e65772"},"PeriodicalIF":5.8,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144333298","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Social Media and Youth Mental Health: Scoping Review of Platform and Policy Recommendations.","authors":"Jasleen Chhabra, Vita Pilkington, Ruben Benakovic, Michael James Wilson, Louise La Sala, Zac Seidler","doi":"10.2196/72061","DOIUrl":"https://doi.org/10.2196/72061","url":null,"abstract":"<p><strong>Background: </strong>High rates of social media use and mental ill-health among young people have drawn significant public, policy, and research concern. Rapid technological advancements and changes in platform design have outpaced our understanding of the health effects of social media and hampered timely evidence-based regulatory responses. While a proliferation of recommendations to social media companies and governments has been published, a comprehensive summary of recommendations for protecting young people's mental health and digital safety does not yet exist.</p><p><strong>Objective: </strong>This scoping review synthesized published recommendations for social media companies and governments in relation to young people's (aged 12-25 years) mental health. A qualitative approach was used to undertake inductive content analysis, where recommendations were grouped under conceptually similar themes.</p><p><strong>Methods: </strong>We searched academic (PubMed, Scopus, and PsycINFO) and nonacademic (Overton and Google) databases for relevant documents. Eligible documents provided recommendations to regulators and social media companies that pertained to social media, young people, and mental health. This review excluded recommendations for young people, caregivers, educators, or clinicians surrounding strategies for managing individual social media use; instead, the recommendations emphasized the regulation or design of social media products and practices of social media companies. Peer-reviewed and gray literature from selected Western contexts (Australia, Canada, the United Kingdom, and the United States) were relevant for inclusion. Documents were published between January 2020 and September 2024.</p><p><strong>Results: </strong>Of the identified 4980 unique reports, 120 (2.41%) progressed to full-text screening, and 70 (1.41%) met the inclusion criteria. Five interrelated themes were identified: (1) legislating and overseeing accountability, (2) transparency, (3) collaboration, (4) safety by design, and (5) restricting young people's access to social media.</p><p><strong>Conclusions: </strong>This review emphasizes the need for multipronged approaches to address the rapidly increasing presence and reach of social media platforms in the lives of young people. These recommendations provide practical and tangible paths forward for governments and industry, backed by expert organizations in youth mental health and technology regulation at a time when expert-informed guidance is sorely needed. Rigorous evaluation of the proposed recommendations is needed while continuing to build on the emerging peer-reviewed evidence base that should form the foundation of policy and regulatory changes.</p>","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e72061"},"PeriodicalIF":5.8,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144336691","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Large Language Model-Assisted Surgical Consent Forms in Non-English Language: Content Analysis and Readability Evaluation.","authors":"Namkee Oh, Jongman Kim, Sunghae Park, Sunghyo An, Eunjin Lee, Hayeon Do, Jiyoung Baik, Suk Min Gwon, Jinsoo Rhu, Gyu-Seong Choi, Seonmin Park, Jai Young Cho, Hae Won Lee, Boram Lee, Eun Sung Jeong, Jeong-Moo Lee, YoungRok Choi, Jieun Kwon, Kyeong Deok Kim, Seok-Hwan Kim, Gwang-Sik Chun","doi":"10.2196/73222","DOIUrl":"https://doi.org/10.2196/73222","url":null,"abstract":"<p><strong>Background: </strong>Surgical consent forms convey critical information; yet, their complex language can limit patient comprehension. Large language models (LLMs) can simplify complex information and improve readability, but evidence of the impact of LLM-generated modifications on content preservation in non-English consent forms is lacking.</p><p><strong>Objective: </strong>This study evaluates the impact of LLM-assisted editing on the readability and content quality of surgical consent forms in Korean-particularly consent documents for standardized liver resection-across multiple institutions.</p><p><strong>Methods: </strong>Standardized liver resection consent forms were collected from 7 South Korean medical institutions, and these forms were simplified using ChatGPT-4o. Thereafter, readability was assessed using KReaD and Natmal indices, while text structure was evaluated based on character count, word count, sentence count, words per sentence, and difficult word ratio. Content quality was analyzed across 4 domains-risk, benefit, alternative, and overall impression-using evaluations from 7 liver resection specialists. Statistical comparisons were conducted using paired 2-sided t tests, and a linear mixed-effects model was applied to account for institutional and evaluator variability.</p><p><strong>Results: </strong>Artificial intelligence-assisted editing significantly improved readability, reducing the KReaD score from 1777 (SD 28.47) to 1335.6 (SD 59.95) (P<.001) and the Natmal score from 1452.3 (SD 88.67) to 1245.3 (SD 96.96) (P=.007). Sentence length and difficult word ratio decreased significantly, contributing to increased accessibility (P<.05). However, content quality analysis showed a decline in the risk description scores (before: 2.29, SD 0.47 vs after: 1.92, SD 0.32; P=.06) and overall impression scores (before: 2.21, SD 0.49 vs after: 1.71, SD 0.64; P=.13). The linear mixed-effects model confirmed significant reductions in risk descriptions (β₁=-0.371; P=.01) and overall impression (β₁=-0.500; P=.03), suggesting potential omissions in critical safety information. Despite this, qualitative analysis indicated that evaluators did not find explicit omissions but perceived the text as overly simplified and less professional.</p><p><strong>Conclusions: </strong>Although LLM-assisted surgical consent forms significantly enhance readability, they may compromise certain aspects of content completeness, particularly in risk disclosure. These findings highlight the need for a balanced approach that maintains accessibility while ensuring medical and legal accuracy. Future research should include patient-centered evaluations to assess comprehension and informed decision-making as well as broader multilingual validation to determine LLM applicability across diverse health care settings.</p>","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e73222"},"PeriodicalIF":5.8,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144333300","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Large Language Model Architectures in Health Care: Scoping Review of Research Perspectives.","authors":"Florian Leiser, Richard Guse, Ali Sunyaev","doi":"10.2196/70315","DOIUrl":"10.2196/70315","url":null,"abstract":"<p><strong>Background: </strong>Large language models (LLMs) can support health care professionals in their daily work, for example, when writing and filing reports or communicating diagnoses. With the rise of LLMs, current research investigates how LLMs could be applied in medical practice and their benefits for physicians in clinical workflows. However, most studies neglect the importance of selecting suitable LLM architectures.</p><p><strong>Objective: </strong>In this literature review, we aim to provide insights on the different LLM model architecture families (ie, Bidirectional Encoder Representations from Transformers [BERT]-based or generative pretrained transformer [GPT]-based models) used in previous research. We report on the suitability and benefits of different LLM model architecture families for various research foci.</p><p><strong>Methods: </strong>To this end, we conduct a scoping review to identify which LLMs are used in health care. Our search included manuscripts from PubMed, arXiv, and medRxiv. We used open and selective coding to assess the 114 identified manuscripts regarding 11 dimensions related to usage and technical facets and the research focus of the manuscripts.</p><p><strong>Results: </strong>We identified 4 research foci that emerged previously in manuscripts, with LLM performance being the main focus. We found that GPT-based models are used for communicative purposes such as examination preparation or patient interaction. In contrast, BERT-based models are used for medical tasks such as knowledge discovery and model improvements.</p><p><strong>Conclusions: </strong>Our study suggests that GPT-based models are better suited for communicative purposes such as report generation or patient interaction. BERT-based models seem to be better suited for innovative applications such as classification or knowledge discovery. This could be due to the architectural differences where GPT processes language unidirectionally and BERT bidirectionally, allowing more in-depth understanding of the text. In addition, BERT-based models seem to allow more straightforward extensions of their models for domain-specific tasks that generally lead to better results. In summary, health care professionals should consider the benefits and differences of the LLM architecture families when selecting a suitable model for their intended purpose.</p>","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e70315"},"PeriodicalIF":5.8,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144326024","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development of a Machine Learning-Based Predictive Model for Postoperative Delirium in Older Adult Intensive Care Unit Patients: Retrospective Study.","authors":"Houfeng Li, Qinglai Zang, Qi Li, Yanchen Lin, Jintao Duan, Jing Huang, Huixiu Hu, Ying Zhang, Dengyun Xia, Miao Zhou","doi":"10.2196/67258","DOIUrl":"https://doi.org/10.2196/67258","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;Delirium is a prevalent phenomenon among patients admitted to the geriatric intensive care unit (ICU) and can adversely impact prognosis and augment the risk of complications.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Objective: &lt;/strong&gt;We aimed to construct and validate a predictive model for postoperative delirium state in older ICU patients, providing timely and effective early identification of high-risk individuals and assisting clinicians in decision-making.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;The data from patients admitted to the ICU for over 24 hours were extracted from the Medical Information Marketplace for Intensive Care IV (MIMIC-IV) database and the eICU Collaborative Research Database (eICU-CRD). The MIMIC-IV data were split (7:3) into training and internal validation sets, while the eICU-CRD data served as an external validation set. Delirium predictions were conducted for the subsequent prediction windows (12 h, 24 h, 48 h, and whole stay time) using data from the first 24 hours post admission. The corresponding feature variables were subjected to Boruta feature selection, and the prediction models were constructed using logistic regression, support vector classifier, random forest classifier, and extreme gradient boosting (XGB). Subsequently, model performance was evaluated using areas under the receiver operating characteristic curves (AUCs), Brier scores, and decision curve analysis, and external validation.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;The MIMIC-IV and eICU-CRD datasets comprised 6129 and 709 patients, respectively, who were included in the analysis. Fifty-four features were selected to construct the predictive model. Regarding internal validation, the XGB model demonstrated the most effective prediction of delirium across different prediction windows. The AUCs for the 4 prediction windows (12 h, 24 h, 48 h, and whole stay time) were 0.848 (95% CI 0.826-0.869), 0.852 (95% CI 0.831-0.872), 0.851 (95% CI 0.831-0.871), and 0.844 (95% CI 0.823-0.863), respectively, and those of the external validation set were 0.777 (95% CI 0.726-0.825), 0.761 (95% CI 0.710-0.808), 0.753 (95% CI 0.704-0.798), and 0.737 (95% CI 0.695-0.777), respectively. Furthermore, the XGB model demonstrated the most accurate calibration across all prediction windows, with values of 0.129, 0.136, 0.144, and 0.148, respectively. Additionally, decision curve analysis revealed that the XGB model outperformed the other models in terms of net gain for the majority of threshold probability values. The 6 most significant predictive features identified were the first day's delirium assessment results, type of first care unit, minimum Glasgow Coma Scale (GCS) score, Acute Physiology Score III, acetaminophen, and nonsteroidal anti-inflammatory drugs.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusions: &lt;/strong&gt;The high-performance XGB model for predicting postoperative delirium state in older adult ICU patients has been successfully developed and validated. The model predicts the delirium st","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e67258"},"PeriodicalIF":5.8,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144333296","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"It is Time to Realize the Promise of the Digital Mental Health Transformation: Application for Population Mental Health.","authors":"Jonathan Adler, Deryk Van Brunt","doi":"10.2196/63791","DOIUrl":"https://doi.org/10.2196/63791","url":null,"abstract":"<p><strong>Unlabelled: </strong>The past 25 years have seen the explosion of digital health care-from 1s and 0s initially serving most researchers for accomplishing their work, to the creation of smartphones, mHealth, and more recently artificial intelligence. The revolution for digital mental health is no longer in its infancy, as new tools are created to address mental health, sometimes even undergoing evaluation for adoption and efficacy. In fact, a recent study reporting on National Health Interview Survey data (annually conducted by the National Center for Health Statistics) indicated that, in 2024, 40% of adults reporting serious psychological distress used a digital health tool, which has increased from 21% in 2017 and 10% in 2013. Given the widespread access to digital tools and the potential of digital mental health, it is time for a new paradigm of care to address the mental health crisis in the United States. Reactive care, consisting largely of medication and counseling provided to those already experiencing severe or debilitating symptoms of mental anguish, is not adequate to address the needs of 22.8% of the US population (>55 million people) experiencing symptoms of a mental illness, and the larger number of people with preclinical mental health concerns. A population mental health approach is needed that includes early identification, intervention, and prevention, in addition to reactive care.</p>","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e63791"},"PeriodicalIF":5.8,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144333299","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The Rapid Online Cognitive Assessment for the Detection of Neurocognitive Disorder: Open-Label Study.","authors":"Calvin Howard, Amy Johnson, Joseph Peedicail, Marcus C Ng","doi":"10.2196/66735","DOIUrl":"https://doi.org/10.2196/66735","url":null,"abstract":"<p><strong>Background: </strong>The rising prevalence of dementia necessitates a scalable solution to cognitive screening. Paper-based cognitive screening examinations are well-validated but minimally scalable. If a digital cognitive screening examination could replicate paper-based screening, it may improve scalability while potentially maintaining the performance of these well-validated paper-based tests. Here, we evaluate the Rapid Online Cognitive Assessment (RoCA), a remote and self-administered digital cognitive screening examination.</p><p><strong>Objective: </strong>The objective of this study was to validate the ability of RoCA to reliably evaluate patient input, identify patients with cognitive impairment relative to the established tests, and evaluate its potential as a screening tool.</p><p><strong>Methods: </strong>RoCA uses a convolutional neural network to evaluate a patient's ability to perform common cognitive screening tasks: wireframe diagram copying and clock drawing tests. To evaluate RoCA, we compared its evaluations with those of established paper-based tests. This open-label study consists of 46 patients (age range 33-82 years) who were enrolled from neurology clinics. Patients completed the RoCA screening examination and either Addenbrooke's Cognitive Examination-3 (ACE-3, n=35) or Montreal Cognitive Assessment (MoCA, n=11). We evaluated 3 primary metrics of RoCA's performance: (1) ability to correctly evaluate patient inputs, (2) ability to identify patients with cognitive impairment compared to ACE-3 and MoCA, and (3) performance as a screening tool.</p><p><strong>Results: </strong>RoCA classifies patients similarly to gold standard paper-based tests, with a receiver operating characteristic area under the curve of 0.81 (95% CI 0.67-0.91; P<.001). RoCA achieved sensitivity of 0.94 (95% CI 0.80-1.0; P<.001). This was robust to multiple control analyses. Approximately 83% (16/19) of the patient respondents reported RoCA as highly intuitive, with 95% (18/19) perceiving it as adding value to their care.</p><p><strong>Conclusions: </strong>RoCA may act as a simple and highly scalable digital cognitive screening examination. However, due to the limitations of this study, further work is required to evaluate the ability of RoCA to be generalizable across patient populations, assess its performance in an entirely remote manner, and analyze the effect of digital literacy.</p>","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e66735"},"PeriodicalIF":5.8,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144333302","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}