<p><strong>Background: </strong>Stress resilience is a dynamic process shaped by the interaction between demands and adaptive resources. Existing measures assess stress and resilience as separate constructs, limiting their use in digital health and workplace interventions. An integrated measure capturing both domains is needed.</p><p><strong>Objective: </strong>We developed and validated the WONE Index, a multidimensional stress resilience tool designed to measure both current stress load and adaptive resources among full-time working adults.</p><p><strong>Methods: </strong>We developed the 32-item WONE Index through literature review, expert consultation, and iterative refinement to assess stress load and resilience resources across behavioral, cognitive, and social domains. Phase 1 (N=1005; United States- or United Kingdom-based full-time employees) evaluated the initial item pool using exploratory and confirmatory factor analyses to establish the preliminary factor structure and assess reliability and validity. Phase 2 (N=306; United States-based adults) expanded underperforming domains, refined items, and tested incremental validity, test-retest reliability, and measurement invariance. Data were collected online through CloudResearch (Connect) and Prolific (Prolific Academic Ltd) using secure survey platforms.</p><p><strong>Results: </strong>Phase 1 supported a 2-domain structure: a Stress Load factor (Work Stress, Personal Stress, and Burnout) and a Resilience Resources factor (Emotion Regulation and Coping, Social Connectedness, and Sleep). Model fit indices were excellent (comparative fit index, CFI=0.95; Tucker-Lewis index, TLI=0.94; and root mean square error of approximation, RMSEA=0.05). Phase 2 replicated and extended this structure, expanding Resilience Resources into 7 domains (adding Purpose and Prosociality, Physical Activity, Dietary Intake, and Perseverative Thinking). Confirmatory factor analyses supported a 2-domain structure, comprising a higher-order Stress Load factor with 3 subdomains (Work Stress, Personal Stress, and Burnout) and a higher-order Resilience Resources factor with 7 subdomains (Emotion Regulation and Coping, Social Connectedness, Purpose and Prosociality, Sleep, Physical Activity, Dietary Intake, and Perseverative Thinking). The Stress Load model demonstrated excellent fit (χ²33=64.18; P=.01; CFI=0.99; TLI=0.98; RMSEA=0.06; and standardized root mean square residual=0.05), and the Resilience Resources model also fit well (χ²443=745.20, P<.001; CFI=0.94; TLI=0.94; RMSEA=0.05; and standardized root mean square residual=0.06). All subscales showed strong internal consistency (composite reliability: mean 0.84, SD 0.10; range 0.74-0.93) and excellent test-retest reliability over 3 weeks (intraclass correlation coefficients 0.77-0.90, 95% CI 0.87-0.93). The index showed strong convergent validity (r=0.73 with Connor-Davidson Resilience Scale and r=-0.66 with Perceived Stress Scale-4) and explained additional varia
{"title":"The WONE Index as a Multidimensional Assessment of Stress Resilience: A Development and Validation Study.","authors":"Lydia Genevieve Roos, Destiny Gilliland, Kelsey Julian, Reeva Misra","doi":"10.2196/81714","DOIUrl":"10.2196/81714","url":null,"abstract":"<p><strong>Background: </strong>Stress resilience is a dynamic process shaped by the interaction between demands and adaptive resources. Existing measures assess stress and resilience as separate constructs, limiting their use in digital health and workplace interventions. An integrated measure capturing both domains is needed.</p><p><strong>Objective: </strong>We developed and validated the WONE Index, a multidimensional stress resilience tool designed to measure both current stress load and adaptive resources among full-time working adults.</p><p><strong>Methods: </strong>We developed the 32-item WONE Index through literature review, expert consultation, and iterative refinement to assess stress load and resilience resources across behavioral, cognitive, and social domains. Phase 1 (N=1005; United States- or United Kingdom-based full-time employees) evaluated the initial item pool using exploratory and confirmatory factor analyses to establish the preliminary factor structure and assess reliability and validity. Phase 2 (N=306; United States-based adults) expanded underperforming domains, refined items, and tested incremental validity, test-retest reliability, and measurement invariance. Data were collected online through CloudResearch (Connect) and Prolific (Prolific Academic Ltd) using secure survey platforms.</p><p><strong>Results: </strong>Phase 1 supported a 2-domain structure: a Stress Load factor (Work Stress, Personal Stress, and Burnout) and a Resilience Resources factor (Emotion Regulation and Coping, Social Connectedness, and Sleep). Model fit indices were excellent (comparative fit index, CFI=0.95; Tucker-Lewis index, TLI=0.94; and root mean square error of approximation, RMSEA=0.05). Phase 2 replicated and extended this structure, expanding Resilience Resources into 7 domains (adding Purpose and Prosociality, Physical Activity, Dietary Intake, and Perseverative Thinking). Confirmatory factor analyses supported a 2-domain structure, comprising a higher-order Stress Load factor with 3 subdomains (Work Stress, Personal Stress, and Burnout) and a higher-order Resilience Resources factor with 7 subdomains (Emotion Regulation and Coping, Social Connectedness, Purpose and Prosociality, Sleep, Physical Activity, Dietary Intake, and Perseverative Thinking). The Stress Load model demonstrated excellent fit (χ²33=64.18; P=.01; CFI=0.99; TLI=0.98; RMSEA=0.06; and standardized root mean square residual=0.05), and the Resilience Resources model also fit well (χ²443=745.20, P<.001; CFI=0.94; TLI=0.94; RMSEA=0.05; and standardized root mean square residual=0.06). All subscales showed strong internal consistency (composite reliability: mean 0.84, SD 0.10; range 0.74-0.93) and excellent test-retest reliability over 3 weeks (intraclass correlation coefficients 0.77-0.90, 95% CI 0.87-0.93). The index showed strong convergent validity (r=0.73 with Connor-Davidson Resilience Scale and r=-0.66 with Perceived Stress Scale-4) and explained additional varia","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"28 ","pages":"e81714"},"PeriodicalIF":6.0,"publicationDate":"2026-01-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12768397/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145906138","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Background: Adherence to antiseizure medications (ASMs) is a cornerstone of effective epilepsy management. However, current consensus guidelines for assessing medication adherence via therapeutic drug monitoring (TDM) may neglect individual patient characteristics, thereby compromising the accuracy of adherence assessments.
Objective: This study proposed an innovative Bayesian-based pharmacokinetic (PK) framework integrated with TDM data to address the above limitations, with a focus on 14 widely prescribed ASMs, including brivaracetam, carbamazepine, clobazam, eslicarbazepine acetate, lacosamide, lamotrigine, levetiracetam, oxcarbazepine, perampanel, phenobarbital, topiramate, valproic acid, vigabatrin, and zonisamide.
Methods: Comprehensive clinical trial simulations were conducted to investigate the PK of ASMs in patients with epilepsy under conditions of full adherence and various nonadherent dosing behaviors, including omission of the last dose and consecutive missed doses. Bayesian posterior probabilities of these dosing behaviors were derived by integrating validated population PK models, individual patient demographics (eg, age, weight, creatinine clearance), dosing history, prior adherence probabilities and TDM measurements. Additionally, the influence of covariates on assessment outcomes was systematically evaluated.
Results: The Bayesian-based PK approach demonstrated robust discriminative ability. Under idealized simulation conditions with minimized variabilities, the approach achieved accurate retrodiction of the last 1 or 2 doses across all 14 ASMs and partial retrodiction of extended nonadherence trajectories for 6 ASMs. Concentration thresholds for adherence classification varied significantly across drugs and are influenced by patient-specific factors, comedications, formulation, sampling time, and prior probability. To translate these insights into practice, an adaptable web-based dashboard was developed using the shiny package in R software to enable precise and real-time assessments of medication adherence.
Conclusions: This study establishes a Bayesian-based PK approach to enhance the assessment of ASMs adherence. This approach facilitates a paradigm shift from population-based management to patient-specific adherence profiling, offering a practical methodology for the precise evaluation of medication-taking behaviors.
{"title":"Bayesian-Based Pharmacokinetic Framework Integrated with Therapeutic Drug Monitoring for Assessing Adherence to Antiseizure Medications: A Clinical Trial Simulation Study.","authors":"Xiao-Qin Liu, Zi-Ran Li, Wei-Wei Lin, Juan Wang, Fu-Qing Gu, Jun-Jie Ding, Zheng Jiao","doi":"10.2196/77917","DOIUrl":"10.2196/77917","url":null,"abstract":"<p><strong>Background: </strong>Adherence to antiseizure medications (ASMs) is a cornerstone of effective epilepsy management. However, current consensus guidelines for assessing medication adherence via therapeutic drug monitoring (TDM) may neglect individual patient characteristics, thereby compromising the accuracy of adherence assessments.</p><p><strong>Objective: </strong>This study proposed an innovative Bayesian-based pharmacokinetic (PK) framework integrated with TDM data to address the above limitations, with a focus on 14 widely prescribed ASMs, including brivaracetam, carbamazepine, clobazam, eslicarbazepine acetate, lacosamide, lamotrigine, levetiracetam, oxcarbazepine, perampanel, phenobarbital, topiramate, valproic acid, vigabatrin, and zonisamide.</p><p><strong>Methods: </strong>Comprehensive clinical trial simulations were conducted to investigate the PK of ASMs in patients with epilepsy under conditions of full adherence and various nonadherent dosing behaviors, including omission of the last dose and consecutive missed doses. Bayesian posterior probabilities of these dosing behaviors were derived by integrating validated population PK models, individual patient demographics (eg, age, weight, creatinine clearance), dosing history, prior adherence probabilities and TDM measurements. Additionally, the influence of covariates on assessment outcomes was systematically evaluated.</p><p><strong>Results: </strong>The Bayesian-based PK approach demonstrated robust discriminative ability. Under idealized simulation conditions with minimized variabilities, the approach achieved accurate retrodiction of the last 1 or 2 doses across all 14 ASMs and partial retrodiction of extended nonadherence trajectories for 6 ASMs. Concentration thresholds for adherence classification varied significantly across drugs and are influenced by patient-specific factors, comedications, formulation, sampling time, and prior probability. To translate these insights into practice, an adaptable web-based dashboard was developed using the shiny package in R software to enable precise and real-time assessments of medication adherence.</p><p><strong>Conclusions: </strong>This study establishes a Bayesian-based PK approach to enhance the assessment of ASMs adherence. This approach facilitates a paradigm shift from population-based management to patient-specific adherence profiling, offering a practical methodology for the precise evaluation of medication-taking behaviors.</p>","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"28 ","pages":"e77917"},"PeriodicalIF":6.0,"publicationDate":"2026-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12772940/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145911927","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
<p><strong>Background: </strong>The incidence of end-stage renal disease continues to rise annually, with dialysis currently serving as the primary replacement therapy. The effectiveness of dialysis treatment and patients' quality of life are highly dependent on their self-management. Mobile health (mHealth), which provides real-time medical support through portable devices, has become an essential tool for assisting patients undergoing dialysis in optimizing their self-management.</p><p><strong>Objective: </strong>This study aimed to systematically explore the core elements of self-management in patients undergoing dialysis and clarify the primary applications of mHealth, including types of mHealth, relevant theories and models, mHealth-based interventions, and evaluation indicators.</p><p><strong>Methods: </strong>This study was guided by Arksey and O'Malley's methodology, PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews), and PRISMA-S (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Literature Search Extension). Databases, such as PubMed, Embase, CINAHL, PsycINFO, and Web of Science, were systematically searched from January 2010 until October 2025. The participants included in this study were patients undergoing dialysis, and the study design must incorporate quantitative research. Published protocols, reviews, editorials, conference papers, books, and non-English studies were excluded. The Mixed Methods Appraisal Tool was used to evaluate the quality of the included studies. Quantitative studies were extracted, mapped, and summarized. The results were collated and synthesized using a structured spreadsheet.</p><p><strong>Results: </strong>Out of 1483 relevant studies, this scoping review ultimately selected 34 studies involving 2068 patients undergoing dialysis. Self-management among patients undergoing dialysis in this study included 6 major areas, including self-monitoring, diet and fluid management, medication management, disease-related knowledge, exercise management, and psychological management. Most studies used a single app (n=22) for management of patients undergoing dialysis, followed by 2 or more online interventions (n=6) and a remote patient monitoring system (n=3). The mHealth-based interventions in this study focused on self-monitoring, dietary and fluid management, and medication management. The transtheoretical model and stages of change (n=5), self-efficacy theory (n=4), and social cognitive theory (n=4) were the most commonly used theories. Among the evaluation indicators, interdialytic weight gain (n=12), serum potassium (n=14), serum phosphorus (n=20), and serum albumin (n=14) were the most commonly used objective indicators. Subjective indicators were assessed using scales, primarily covering adherence (n=17), self-efficacy (n=14), quality of life (n=12), knowledge (n=9), and diet and nutrition (n=9).</p><p><strong>Conclusions: </strong>Altho
背景：终末期肾脏疾病的发病率每年持续上升，透析目前是主要的替代治疗方法。透析治疗的效果和患者的生活质量高度依赖于患者的自我管理。移动医疗（mHealth）通过便携式设备提供实时医疗支持，已成为帮助接受透析的患者优化自我管理的重要工具。目的：本研究旨在系统探讨透析患者自我管理的核心要素，明确移动健康的主要应用，包括移动健康的类型、相关理论和模型、基于移动健康的干预措施和评估指标。方法：本研究采用Arksey和O'Malley的方法，PRISMA-ScR（系统评价和meta分析扩展的首选报告项目）和PRISMA-S（系统评价和meta分析文献搜索扩展的首选报告项目）。从2010年1月到2025年10月，系统地检索了PubMed、Embase、CINAHL、PsycINFO和Web of Science等数据库。本研究纳入的受试者为透析患者，研究设计必须纳入定量研究。已发表的协议、评论、社论、会议论文、书籍和非英语研究被排除在外。采用混合方法评价工具评价纳入研究的质量。定量研究被提取、绘制和总结。使用结构化电子表格对结果进行整理和综合。结果：在1483项相关研究中，本综述最终选择了34项研究，涉及2068例透析患者。本研究透析患者的自我管理包括自我监测、饮食与液体管理、药物管理、疾病相关知识、运动管理和心理管理6个主要方面。大多数研究使用单一应用程序（n=22）来管理接受透析的患者，其次是2个或更多的在线干预（n=6）和远程患者监测系统（n=3）。本研究中基于移动健康的干预措施侧重于自我监测、饮食和液体管理以及药物管理。跨理论模型和变化阶段理论（n=5）、自我效能理论（n=4）和社会认知理论（n=4）是最常用的理论。评价指标中，最常用的客观指标为透析间期体重增加（n=12）、血清钾（n=14）、血清磷（n=20）、血清白蛋白（n=14）。使用量表评估主观指标，主要包括依从性（n=17）、自我效能（n=14）、生活质量（n=12）、知识（n=9）以及饮食和营养（n=9）。结论：尽管移动医疗有望改善透析患者的自我管理和预后，但仍有很大的发展空间。未来该领域的研究应侧重于加强自适应软件开发，深度融合人工智能技术，解决特殊人群的需求，建立标准化的自我管理评估体系。我们的研究结果不仅为优化透析患者的临床管理策略提供了理论框架，也为后续应用程序的开发提供了有针对性的指导和实践见解。
{"title":"The Application of Mobile Health in Self-Management Among Patients Undergoing Dialysis: Scoping Review.","authors":"Qin Xu, Yulin Xu, Xiaoqin Liu, Xiaolin Ma","doi":"10.2196/76880","DOIUrl":"10.2196/76880","url":null,"abstract":"<p><strong>Background: </strong>The incidence of end-stage renal disease continues to rise annually, with dialysis currently serving as the primary replacement therapy. The effectiveness of dialysis treatment and patients' quality of life are highly dependent on their self-management. Mobile health (mHealth), which provides real-time medical support through portable devices, has become an essential tool for assisting patients undergoing dialysis in optimizing their self-management.</p><p><strong>Objective: </strong>This study aimed to systematically explore the core elements of self-management in patients undergoing dialysis and clarify the primary applications of mHealth, including types of mHealth, relevant theories and models, mHealth-based interventions, and evaluation indicators.</p><p><strong>Methods: </strong>This study was guided by Arksey and O'Malley's methodology, PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews), and PRISMA-S (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Literature Search Extension). Databases, such as PubMed, Embase, CINAHL, PsycINFO, and Web of Science, were systematically searched from January 2010 until October 2025. The participants included in this study were patients undergoing dialysis, and the study design must incorporate quantitative research. Published protocols, reviews, editorials, conference papers, books, and non-English studies were excluded. The Mixed Methods Appraisal Tool was used to evaluate the quality of the included studies. Quantitative studies were extracted, mapped, and summarized. The results were collated and synthesized using a structured spreadsheet.</p><p><strong>Results: </strong>Out of 1483 relevant studies, this scoping review ultimately selected 34 studies involving 2068 patients undergoing dialysis. Self-management among patients undergoing dialysis in this study included 6 major areas, including self-monitoring, diet and fluid management, medication management, disease-related knowledge, exercise management, and psychological management. Most studies used a single app (n=22) for management of patients undergoing dialysis, followed by 2 or more online interventions (n=6) and a remote patient monitoring system (n=3). The mHealth-based interventions in this study focused on self-monitoring, dietary and fluid management, and medication management. The transtheoretical model and stages of change (n=5), self-efficacy theory (n=4), and social cognitive theory (n=4) were the most commonly used theories. Among the evaluation indicators, interdialytic weight gain (n=12), serum potassium (n=14), serum phosphorus (n=20), and serum albumin (n=14) were the most commonly used objective indicators. Subjective indicators were assessed using scales, primarily covering adherence (n=17), self-efficacy (n=14), quality of life (n=12), knowledge (n=9), and diet and nutrition (n=9).</p><p><strong>Conclusions: </strong>Altho","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"28 ","pages":"e76880"},"PeriodicalIF":6.0,"publicationDate":"2026-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12791203/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145892531","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Myungsung Kim, Min Jeon, Yerin Lee, Sangil Lee, Hwang Kim, Dooyoung Jung
<p><strong>Background: </strong>Exposure therapy effectively treats anxiety disorders but faces implementation barriers, including cost, time constraints, and reluctance from therapists and clients. Virtual reality exposure therapy (VRET) offers a controlled digital alternative addressing these issues. However, adoption remains limited, with previous studies focusing mainly on hospital settings without considering individual or workplace factors.</p><p><strong>Objective: </strong>This study examined factors affecting counselors' VRET acceptance across diverse settings. We used the Unified Theory of Acceptance and Use of Technology (UTAUT) extended with job stress and resistance to change. Open-ended questions provided a deeper understanding of counselors' perspectives on VRET.</p><p><strong>Methods: </strong>A cross-sectional mixed methods study was conducted with 258 certified counselors across various settings, including universities, public institutions, and private clinics. Participants watched a 4-minute VRET introduction video and completed a survey measuring UTAUT variables (performance expectancy, effort expectancy, facilitating conditions, and social influence), resistance to change, and job stress. Stepwise forward selection multiple linear regression with moderation analyses was conducted to identify key predictors and test interaction effects. Open-ended responses (N=257, 290 meaning units) on VRET applicability and improvement suggestions were analyzed using team-based thematic analysis with iterative consensus coding.</p><p><strong>Results: </strong>Performance expectancy (β=.404, 95% CI 0.297-0.512, P<.001) and social influence (β=.387, 95% CI 0.280-0.494, P<.001) significantly predicted VRET adoption intentions (R2=0.494). Moderation analysis revealed that routine seeking weakened performance expectancy impact (β=-.160, 95% CI -0.277 to -0.043, P<.01), low job control strengthened it (β=.162, 95% CI 0.280-0.494, P<.005), and high job demands reduced social influence effects (β=-.150, 95% CI -0.263 to -0.036, P=.01). The narrow confidence intervals indicate precise estimation of these moderation effects. Younger counselors were more sensitive to contextual moderators, while older counselors prioritized performance expectancy. Thematic analysis identified 3 themes: counselor evaluation criteria for VRET, emphasizing content diversity and scientific validation; considerations for promoting and introducing VRET to counselors, addressing implementation challenges; and areas requiring continuous improvement for VRET field implementation, emphasizing professional competence and system reliability.</p><p><strong>Conclusions: </strong>This study advances VRET acceptance research by examining certified counselors across diverse nonhospital settings-unlike prior hospital-focused physician studies-and extending UTAUT with profession-specific moderators. Performance expectancy and social influence emerged as primary predictors, with routine se
背景：暴露疗法可有效治疗焦虑症，但存在实施障碍，包括费用、时间限制以及治疗师和来访者的不情愿。虚拟现实暴露疗法（VRET）为解决这些问题提供了一种可控的数字替代方案。然而，采用仍然有限，以前的研究主要集中在医院环境，没有考虑个人或工作场所的因素。目的：本研究探讨了不同环境下心理咨询师VRET接受度的影响因素。我们使用了技术接受和使用的统一理论（UTAUT），扩展了工作压力和对变化的抵制。开放式问题提供了辅导员对VRET的观点的更深层次的理解。方法：一项横断面混合方法研究对来自不同环境的258名认证咨询师进行了研究，包括大学、公共机构和私人诊所。参与者观看了一段4分钟的VRET介绍视频，并完成了一份测量UTAUT变量（表现预期、努力预期、促进条件和社会影响）、对变化的抵抗力和工作压力的调查。采用逐步前向选择多元线性回归和适度分析来确定关键预测因子和检验交互效应。采用基于团队的主题分析和迭代共识编码，对开放式反馈（N=257、290个意义单位）的VRET适用性和改进建议进行分析。结果：绩效预期（β= 0.404, 95% CI 0.297-0.512, p）结论：本研究通过在不同的非医院环境中检查认证咨询师（不像以前以医院为中心的医生研究），并将UTAUT扩展为专业特定调节因子，从而推进了VRET接受度研究。表现预期和社会影响是主要的预测因素，常规求职和工作环境在不同年龄组中显著调节了这些影响。专题分析显示，辅导员将VRET评估为需要科学验证、内容多样化和结构化培训的辅助工具，而不仅仅是技术可用性。研究结果为以下实际策略提供了依据：传播有效性证据，利用专业网络，解决高需求环境下的工作环境障碍，以及制定适合年龄的方法。洞察指导内容开发者、政策制定者和研究人员在医院之外实施VRET。
{"title":"Exploring Predictors of Counselors' Acceptance of Virtual Reality Exposure Therapy With Resistance and Job Contexts as Moderators: Cross-Sectional Mixed Methods Study.","authors":"Myungsung Kim, Min Jeon, Yerin Lee, Sangil Lee, Hwang Kim, Dooyoung Jung","doi":"10.2196/81803","DOIUrl":"10.2196/81803","url":null,"abstract":"<p><strong>Background: </strong>Exposure therapy effectively treats anxiety disorders but faces implementation barriers, including cost, time constraints, and reluctance from therapists and clients. Virtual reality exposure therapy (VRET) offers a controlled digital alternative addressing these issues. However, adoption remains limited, with previous studies focusing mainly on hospital settings without considering individual or workplace factors.</p><p><strong>Objective: </strong>This study examined factors affecting counselors' VRET acceptance across diverse settings. We used the Unified Theory of Acceptance and Use of Technology (UTAUT) extended with job stress and resistance to change. Open-ended questions provided a deeper understanding of counselors' perspectives on VRET.</p><p><strong>Methods: </strong>A cross-sectional mixed methods study was conducted with 258 certified counselors across various settings, including universities, public institutions, and private clinics. Participants watched a 4-minute VRET introduction video and completed a survey measuring UTAUT variables (performance expectancy, effort expectancy, facilitating conditions, and social influence), resistance to change, and job stress. Stepwise forward selection multiple linear regression with moderation analyses was conducted to identify key predictors and test interaction effects. Open-ended responses (N=257, 290 meaning units) on VRET applicability and improvement suggestions were analyzed using team-based thematic analysis with iterative consensus coding.</p><p><strong>Results: </strong>Performance expectancy (β=.404, 95% CI 0.297-0.512, P<.001) and social influence (β=.387, 95% CI 0.280-0.494, P<.001) significantly predicted VRET adoption intentions (R2=0.494). Moderation analysis revealed that routine seeking weakened performance expectancy impact (β=-.160, 95% CI -0.277 to -0.043, P<.01), low job control strengthened it (β=.162, 95% CI 0.280-0.494, P<.005), and high job demands reduced social influence effects (β=-.150, 95% CI -0.263 to -0.036, P=.01). The narrow confidence intervals indicate precise estimation of these moderation effects. Younger counselors were more sensitive to contextual moderators, while older counselors prioritized performance expectancy. Thematic analysis identified 3 themes: counselor evaluation criteria for VRET, emphasizing content diversity and scientific validation; considerations for promoting and introducing VRET to counselors, addressing implementation challenges; and areas requiring continuous improvement for VRET field implementation, emphasizing professional competence and system reliability.</p><p><strong>Conclusions: </strong>This study advances VRET acceptance research by examining certified counselors across diverse nonhospital settings-unlike prior hospital-focused physician studies-and extending UTAUT with profession-specific moderators. Performance expectancy and social influence emerged as primary predictors, with routine se","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e81803"},"PeriodicalIF":6.0,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12755899/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145878542","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
<p><strong>Background: </strong>The Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) is a federal nutrition assistance program for low-income, food-insecure mothers and young children in the United States. Despite its intended goals, many eligible individuals forgo WIC benefits, in part due to administrative burden-defined as the complex, often frustrating processes encountered when navigating public benefit programs. In response, a range of digital interventions and policy waivers were introduced during the COVID-19 pandemic, but their effectiveness in reducing barriers remains unclear.</p><p><strong>Objective: </strong>Drawing from administrative burden theory and human-computer interaction research, this study examined user reviews of WIC smartphone apps (WIC Apps) used by local agencies. Specifically, it investigated (1) how obstacles to WIC access manifested in daily app use, (2) how user experiences shifted after the onset of the COVID-19 pandemic, and (3) how these changes were associated with app ratings.</p><p><strong>Methods: </strong>An original dataset of user reviews (Nreview=28,212) was compiled for 26 WIC Apps between 2013 and 2024. Structural topic modeling identified 8 key themes, and sentiment was examined with Robustly Optimized Bidirectional Encoder Representations From Transformers Pretraining Approach. Analyses compared topic prevalence and sentiment distributions before and after COVID-19. Mixed-effects models examined the relationship between topics, sentiment, and app ratings.</p><p><strong>Results: </strong>Technical concerns related to account authentication and login, document upload, and app updates were among the most prevalent themes. These issues were typically expressed with negative sentiment and appeared more frequently in pre-COVID-19 reviews than in post-COVID-19 reviews. Although reliability problems (eg, outages and maintenance) persisted, post-COVID-19 reviews increasingly emphasized features that facilitated program tracking, shopping and benefit redemption, and general ease of use, which were generally described with positive sentiment. Mixed-effects analyses indicated that these post-COVID-19 topics were significantly associated with higher app ratings (program tracking: B=0.21, SE=0.06; P=.001; shopping and redemption: B=0.18, SE=0.07; P=.01; and ease of use: B=0.10, SE=0.05; P=.04), whereas pre-COVID-19 concerns were not associated with ratings (Ps>.05). When sentiment was added to the mixed-effect model, it became the dominant factor: negative sentiment was associated with lower ratings (B=-1.71, SE=0.03; P<.001), and positive sentiment was associated with higher ratings (B=1.78, SE=0.03; P<.001). After accounting for sentiment, no individual topic was significantly associated with ratings (Ps>.05), suggesting that sentiment contributed to much of the variance previously linked to topics.</p><p><strong>Conclusions: </strong>User-centered digital interventions, such as WIC
{"title":"Effectiveness of Digital Interventions for Low-Income, Food-Insecure Populations: Natural Language Processing Study of WIC Smartphone App User Reviews, 2013-2024.","authors":"Jihye Lee","doi":"10.2196/78984","DOIUrl":"10.2196/78984","url":null,"abstract":"<p><strong>Background: </strong>The Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) is a federal nutrition assistance program for low-income, food-insecure mothers and young children in the United States. Despite its intended goals, many eligible individuals forgo WIC benefits, in part due to administrative burden-defined as the complex, often frustrating processes encountered when navigating public benefit programs. In response, a range of digital interventions and policy waivers were introduced during the COVID-19 pandemic, but their effectiveness in reducing barriers remains unclear.</p><p><strong>Objective: </strong>Drawing from administrative burden theory and human-computer interaction research, this study examined user reviews of WIC smartphone apps (WIC Apps) used by local agencies. Specifically, it investigated (1) how obstacles to WIC access manifested in daily app use, (2) how user experiences shifted after the onset of the COVID-19 pandemic, and (3) how these changes were associated with app ratings.</p><p><strong>Methods: </strong>An original dataset of user reviews (Nreview=28,212) was compiled for 26 WIC Apps between 2013 and 2024. Structural topic modeling identified 8 key themes, and sentiment was examined with Robustly Optimized Bidirectional Encoder Representations From Transformers Pretraining Approach. Analyses compared topic prevalence and sentiment distributions before and after COVID-19. Mixed-effects models examined the relationship between topics, sentiment, and app ratings.</p><p><strong>Results: </strong>Technical concerns related to account authentication and login, document upload, and app updates were among the most prevalent themes. These issues were typically expressed with negative sentiment and appeared more frequently in pre-COVID-19 reviews than in post-COVID-19 reviews. Although reliability problems (eg, outages and maintenance) persisted, post-COVID-19 reviews increasingly emphasized features that facilitated program tracking, shopping and benefit redemption, and general ease of use, which were generally described with positive sentiment. Mixed-effects analyses indicated that these post-COVID-19 topics were significantly associated with higher app ratings (program tracking: B=0.21, SE=0.06; P=.001; shopping and redemption: B=0.18, SE=0.07; P=.01; and ease of use: B=0.10, SE=0.05; P=.04), whereas pre-COVID-19 concerns were not associated with ratings (Ps>.05). When sentiment was added to the mixed-effect model, it became the dominant factor: negative sentiment was associated with lower ratings (B=-1.71, SE=0.03; P<.001), and positive sentiment was associated with higher ratings (B=1.78, SE=0.03; P<.001). After accounting for sentiment, no individual topic was significantly associated with ratings (Ps>.05), suggesting that sentiment contributed to much of the variance previously linked to topics.</p><p><strong>Conclusions: </strong>User-centered digital interventions, such as WIC ","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e78984"},"PeriodicalIF":6.0,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12755294/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145878547","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Moses Okumu, Carmen Hellen Logie, Isaac Koomson, Thabani Nyoni, Joshua Muzei, Bonita B Sharma, Flora Cohen, William Byansi, Michelle G Thompson, Joseph Cedrick Wabwire, Catherine Naluwende Nafula, Robert Hakiza, Peter Kyambadde, Liliane Cambraia Windsor
<p><strong>Background: </strong>Digital health interventions can enhance sexual health equity among marginalized and underserved populations, including displaced youth. However, there is limited understanding of displaced youth's digital health literacy (DHL) and its association with knowledge of and access to sexual health products and services.</p><p><strong>Objective: </strong>This study aims to identify patterns of DHL among displaced youth and assess how these patterns are associated with awareness of and access to sexual health products and services, while considering gender differences.</p><p><strong>Methods: </strong>We conducted a cross-sectional tablet-assisted survey in Kampala, Uganda. We used peer-driven sampling to recruit displaced youth aged 16-24 years living in 5 informal urban settlements. We identified DHL patterns using latent profile analysis. Gender-disaggregated multivariate probit models were constructed to estimate the relationship between DHL and awareness of and access to sexual health products and services (eg, sexual and reproductive health [SRH] information, external condoms, condom use training, sexually transmitted infection testing, and HIV testing).</p><p><strong>Results: </strong>Among the participants (N=445), our latent profile analysis identified 4-DHL classes named: low (class 1, 51/444, 11.5%), moderate (class 2, 99/444, 22.2%), high (class 3, 138/444, 31%), and very high (class 4, 157/444, 35.3%). Our adjusted multivariate probit model indicated that, compared to class 1, class 4 participants were more likely to know where to access condom use training (marginal effect [ME]=0.23; P<.001), external condoms (ME=0.19; P<.001), and HIV testing (ME=0.23; P<.001). We also noted gender-based differences. Men with very high DHL, compared with those with low DHL, were more likely to know where to access SRH information (ME=0.46; P<.001) and condom use training (ME=0.40; P<.050), while women with very high DHL were more likely than those with low DHL to report knowing how to access condom use training (ME=0.12, SE=0.06; P<.050), external condoms (ME=0.34; P<.001), and HIV testing (ME=0.22, SE=0.10; P<.050). Regarding access to sexual health products and services in the last 3 months, class 4 respondents reported higher access to condom use training (ME=0.13, SE=0.04; P<.001), external condoms (ME=0.14; P<.050), and HIV testing (ME=0.24; P<.050) than class 1 respondents. Gender differences showed that among men, those with very high DHL were more likely to access condom use training (ME=0.28; P<.010) than those with low DHL. In contrast, among women, those with very high DHL were less likely to access SRH information (ME=-0.20; P<.001).</p><p><strong>Conclusions: </strong>Our findings reveal a generally high level of DHL but suboptimal awareness of and access to SRH services among urban displaced youth in Kampala. Improving SRH among urban displaced populations will require gender-responsive and culturally grounded
{"title":"Digital Health Literacy and Its Role in Awareness of and Access to Sexual Health Products and Services Among Displaced Youth in Uganda's Informal Urban Settlements: Community-Based Cross-Sectional Study.","authors":"Moses Okumu, Carmen Hellen Logie, Isaac Koomson, Thabani Nyoni, Joshua Muzei, Bonita B Sharma, Flora Cohen, William Byansi, Michelle G Thompson, Joseph Cedrick Wabwire, Catherine Naluwende Nafula, Robert Hakiza, Peter Kyambadde, Liliane Cambraia Windsor","doi":"10.2196/78343","DOIUrl":"10.2196/78343","url":null,"abstract":"<p><strong>Background: </strong>Digital health interventions can enhance sexual health equity among marginalized and underserved populations, including displaced youth. However, there is limited understanding of displaced youth's digital health literacy (DHL) and its association with knowledge of and access to sexual health products and services.</p><p><strong>Objective: </strong>This study aims to identify patterns of DHL among displaced youth and assess how these patterns are associated with awareness of and access to sexual health products and services, while considering gender differences.</p><p><strong>Methods: </strong>We conducted a cross-sectional tablet-assisted survey in Kampala, Uganda. We used peer-driven sampling to recruit displaced youth aged 16-24 years living in 5 informal urban settlements. We identified DHL patterns using latent profile analysis. Gender-disaggregated multivariate probit models were constructed to estimate the relationship between DHL and awareness of and access to sexual health products and services (eg, sexual and reproductive health [SRH] information, external condoms, condom use training, sexually transmitted infection testing, and HIV testing).</p><p><strong>Results: </strong>Among the participants (N=445), our latent profile analysis identified 4-DHL classes named: low (class 1, 51/444, 11.5%), moderate (class 2, 99/444, 22.2%), high (class 3, 138/444, 31%), and very high (class 4, 157/444, 35.3%). Our adjusted multivariate probit model indicated that, compared to class 1, class 4 participants were more likely to know where to access condom use training (marginal effect [ME]=0.23; P<.001), external condoms (ME=0.19; P<.001), and HIV testing (ME=0.23; P<.001). We also noted gender-based differences. Men with very high DHL, compared with those with low DHL, were more likely to know where to access SRH information (ME=0.46; P<.001) and condom use training (ME=0.40; P<.050), while women with very high DHL were more likely than those with low DHL to report knowing how to access condom use training (ME=0.12, SE=0.06; P<.050), external condoms (ME=0.34; P<.001), and HIV testing (ME=0.22, SE=0.10; P<.050). Regarding access to sexual health products and services in the last 3 months, class 4 respondents reported higher access to condom use training (ME=0.13, SE=0.04; P<.001), external condoms (ME=0.14; P<.050), and HIV testing (ME=0.24; P<.050) than class 1 respondents. Gender differences showed that among men, those with very high DHL were more likely to access condom use training (ME=0.28; P<.010) than those with low DHL. In contrast, among women, those with very high DHL were less likely to access SRH information (ME=-0.20; P<.001).</p><p><strong>Conclusions: </strong>Our findings reveal a generally high level of DHL but suboptimal awareness of and access to SRH services among urban displaced youth in Kampala. Improving SRH among urban displaced populations will require gender-responsive and culturally grounded ","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e78343"},"PeriodicalIF":6.0,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12805323/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145878590","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
<p><strong>Background: </strong>Machine learning (ML) has been investigated for its predictive value in knee osteoarthritis (KOA) progression. However, systematic evidence on the effectiveness of ML is still lacking, posing a challenge to precision prevention.</p><p><strong>Objective: </strong>This systematic review aimed to systematically assess the application status and accuracy of ML in predicting KOA progression and to compare the predictive performance of ML, traditional methods, and deep learning under different datasets, model types, modeling variables, and definitions of KOA progression.</p><p><strong>Methods: </strong>Following the PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) statement, a systematic search was conducted in Embase, Web of Science, PubMed, and Cochrane Library up to October 10, 2025. Two investigators were independently responsible for study screening, data extraction, and risk-of-bias assessment in included studies using the Prediction Model Risk of Bias Assessment Tool. Meta-analyses were conducted on the concordance index (C-index) and diagnostic 4-fold table using a random effects model, with prediction intervals (PIs) reported. In addition, subgroup analyses were performed by model type, modeling variable, and definition of KOA progression.</p><p><strong>Results: </strong>A total of 32 studies were included. The overall risk of bias was considered low in 8 studies, high in 13 studies, and unclear in 11 studies. For predicting all progression, the pooled C-index was 0.773 (95% CI 0.727-0.821; 95% PI 0.567-1.000) for the clinical feature-based model, 0.798 (95% CI 0.755-0.843; 95% PI 0.646-0.984) for the magnetic resonance imaging (MRI)-based model, 0.712 (95% CI 0.657-0.772; 95% PI 0.526-0.965) for the X-ray-based model, 0.806 (95% CI 0.765-0.849; 95% PI 0.639-1.000) for the MRI+clinical feature-based model, 0.772 (95% CI 0.731-0.815; 95% PI 0.610-0.976) for the X-ray+clinical feature-based model, and 0.731 (95% CI 0.669-0.798; 95% PI 0.518-1.000) for the clinical feature+X-ray+MRI-based model. The clinical feature-based model was established mainly using logistic regression and exhibited accuracy comparable to other ML models. Among image-based models, traditional ML or deep learning possessed higher accuracy.</p><p><strong>Conclusions: </strong>This systematic review used CIs to estimate mean effects and PIs to estimate the potential range of effects in future scenarios. It systematically compared the performance of ML in predicting KOA progression under different model types, modeling variables, and definitions of KOA progression. ML models demonstrate certain discriminatory power in predicting KOA progression, but current evidence should be interpreted with caution due to various sources of significant heterogeneity, such as variations in the definition of KOA progression and validation strategies. Future research should standardize the definition of KOA progression, enhance methodol
背景：机器学习（ML）在膝关节骨关节炎（KOA）进展中的预测价值已被研究。然而，关于机器学习有效性的系统证据仍然缺乏，这对精确预防提出了挑战。目的：本系统综述旨在系统评估机器学习在KOA进展预测中的应用现状和准确性，并比较机器学习、传统方法和深度学习在不同数据集、模型类型、建模变量和KOA进展定义下的预测性能。方法：根据PRISMA （Preferred Reporting Items for Systematic reviews and meta - analysis）声明，系统检索Embase、Web of Science、PubMed和Cochrane Library，检索时间截止到2025年10月10日。两名研究者使用预测模型偏倚风险评估工具独立负责纳入研究的研究筛选、数据提取和偏倚风险评估。采用随机效应模型对一致性指数（C-index）和诊断4重表进行meta分析，并报告预测区间（pi）。此外，根据模型类型、建模变量和KOA进展定义进行亚组分析。结果：共纳入32项研究。8项研究认为总体偏倚风险低，13项研究认为偏倚风险高，11项研究认为偏倚风险不明确。对于预测所有进展，临床特征模型的合并c -指数为0.773 (95% CI 0.727-0.821; 95% PI 0.567-1.000)，磁共振成像（MRI）模型的合并c -指数为0.798 (95% CI 0.755-0.843; 95% PI 0.646-0.984)， x线模型的合并c -指数为0.712 (95% CI 0.657-0.772; 95% PI 0.526-0.965)， MRI+临床特征模型的合并c -指数为0.772 (95% CI 0.765-0.849; 95% PI 0.639-1.000)， MRI+临床特征模型的合并c -指数为0.772 (95% CI 0.731-0.815；基于x线+临床特征的模型为95% PI 0.610-0.976)，基于临床特征+ x线+ mri的模型为0.731 （95% CI 0.669-0.798; 95% PI 0.518-1.000）。基于临床特征的模型主要使用逻辑回归建立，其准确性与其他ML模型相当。在基于图像的模型中，传统的ML或深度学习具有更高的准确性。结论：本系统综述使用ci来估计平均效应，使用pi来估计未来情景的潜在影响范围。系统比较了ML在不同模型类型、建模变量和KOA进展定义下预测KOA进展的性能。ML模型在预测KOA进展方面显示出一定的歧视性，但由于各种来源的显著异质性，如KOA进展的定义和验证策略的差异，目前的证据应谨慎解释。未来的研究应规范KOA进展的定义，提高方法的严谨性，并进行严格的外部验证，以提高模型的可靠性，促进临床翻译。
{"title":"Predictive Value of Machine Learning in Knee Osteoarthritis Progression: Systematic Review and Meta-Analysis.","authors":"Yanwen Liu, Guangzhi Xiao, Youqun Zhang, Xinyi Wang, Junfeng Jia, Aiguo Xie, Zhaohui Zheng, Kui Zhang","doi":"10.2196/80430","DOIUrl":"10.2196/80430","url":null,"abstract":"<p><strong>Background: </strong>Machine learning (ML) has been investigated for its predictive value in knee osteoarthritis (KOA) progression. However, systematic evidence on the effectiveness of ML is still lacking, posing a challenge to precision prevention.</p><p><strong>Objective: </strong>This systematic review aimed to systematically assess the application status and accuracy of ML in predicting KOA progression and to compare the predictive performance of ML, traditional methods, and deep learning under different datasets, model types, modeling variables, and definitions of KOA progression.</p><p><strong>Methods: </strong>Following the PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) statement, a systematic search was conducted in Embase, Web of Science, PubMed, and Cochrane Library up to October 10, 2025. Two investigators were independently responsible for study screening, data extraction, and risk-of-bias assessment in included studies using the Prediction Model Risk of Bias Assessment Tool. Meta-analyses were conducted on the concordance index (C-index) and diagnostic 4-fold table using a random effects model, with prediction intervals (PIs) reported. In addition, subgroup analyses were performed by model type, modeling variable, and definition of KOA progression.</p><p><strong>Results: </strong>A total of 32 studies were included. The overall risk of bias was considered low in 8 studies, high in 13 studies, and unclear in 11 studies. For predicting all progression, the pooled C-index was 0.773 (95% CI 0.727-0.821; 95% PI 0.567-1.000) for the clinical feature-based model, 0.798 (95% CI 0.755-0.843; 95% PI 0.646-0.984) for the magnetic resonance imaging (MRI)-based model, 0.712 (95% CI 0.657-0.772; 95% PI 0.526-0.965) for the X-ray-based model, 0.806 (95% CI 0.765-0.849; 95% PI 0.639-1.000) for the MRI+clinical feature-based model, 0.772 (95% CI 0.731-0.815; 95% PI 0.610-0.976) for the X-ray+clinical feature-based model, and 0.731 (95% CI 0.669-0.798; 95% PI 0.518-1.000) for the clinical feature+X-ray+MRI-based model. The clinical feature-based model was established mainly using logistic regression and exhibited accuracy comparable to other ML models. Among image-based models, traditional ML or deep learning possessed higher accuracy.</p><p><strong>Conclusions: </strong>This systematic review used CIs to estimate mean effects and PIs to estimate the potential range of effects in future scenarios. It systematically compared the performance of ML in predicting KOA progression under different model types, modeling variables, and definitions of KOA progression. ML models demonstrate certain discriminatory power in predicting KOA progression, but current evidence should be interpreted with caution due to various sources of significant heterogeneity, such as variations in the definition of KOA progression and validation strategies. Future research should standardize the definition of KOA progression, enhance methodol","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e80430"},"PeriodicalIF":6.0,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12753132/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145862935","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Tabinda Sarwar, Antonio José Jimeno Yepes, Lawrence Cavedon
<p><strong>Background: </strong>Data collected in controlled settings typically results in high-quality datasets. However, in real-world applications, the quality of data collection is often compromised. It is well established that the quality of a dataset significantly impacts the performance of machine learning models. In this context, detailed information about individuals is often recorded in progress notes. Given the critical nature of health applications, it is essential to evaluate the impact of textual data quality, as any incorrect prediction can have serious, potentially life-threatening consequences.</p><p><strong>Objective: </strong>This study aims to quantify the quality of textual datasets and systematically evaluate the impact of varying levels of errors on feature representation and machine learning models. The primary goal is to determine whether feature representations and machine learning models are tolerant to errors and to assess whether investing additional time and computational resources to improve data quality is justified.</p><p><strong>Methods: </strong>We developed a rudimentary error rate metric to evaluate textual dataset quality at the token level. The Mixtral large language model (LLM) was used to quantify and correct errors in low-quality datasets. The study analyzed two health care datasets: the high-quality MIMIC-III public hospital dataset (for mortality prediction) and a lower-quality private dataset from Australian aged care homes (AACHs; for depression and fall risk prediction). Errors were systematically introduced into MIMIC-III at varying rates, while the AACH dataset quality was improved using the LLM. Feature representations and machine learning models were assessed using the area under the receiver operating curve.</p><p><strong>Results: </strong>For the sampled 35,774 and 6336 patients from the MIMIC and AACH datasets, respectively, we used Mixtral to introduce errors in MIMIC and correct errors in AACH. Mixtral correctly detected errors in 63% of progress notes, with 17% containing a single token misclassified due to medical terminology. LLMs demonstrated potential for improving progress note quality by addressing various errors. Under varying error rates (5%-20%, in 5% increments), feature representation performance was tolerant to lower error rates (<10%) but declined significantly at higher rates. This aligned with the AACH dataset's 8% error rate, where no major performance drop was observed. Across both datasets, term frequency-inverted document frequency outperformed embedding features, and machine learning models varied in effectiveness, highlighting that optimal feature representation and model choice depend on the specific task.</p><p><strong>Conclusions: </strong>This study revealed that models performed relatively well on datasets with lower error rates (<10%), but their performance declined significantly as error rates increased (≥10%). Therefore, it is crucial to evaluate the quality of
{"title":"Assessing the Impact of the Quality of Textual Data on Feature Representation and Machine Learning Models: Quantitative Study Using Large Language Models.","authors":"Tabinda Sarwar, Antonio José Jimeno Yepes, Lawrence Cavedon","doi":"10.2196/73325","DOIUrl":"10.2196/73325","url":null,"abstract":"<p><strong>Background: </strong>Data collected in controlled settings typically results in high-quality datasets. However, in real-world applications, the quality of data collection is often compromised. It is well established that the quality of a dataset significantly impacts the performance of machine learning models. In this context, detailed information about individuals is often recorded in progress notes. Given the critical nature of health applications, it is essential to evaluate the impact of textual data quality, as any incorrect prediction can have serious, potentially life-threatening consequences.</p><p><strong>Objective: </strong>This study aims to quantify the quality of textual datasets and systematically evaluate the impact of varying levels of errors on feature representation and machine learning models. The primary goal is to determine whether feature representations and machine learning models are tolerant to errors and to assess whether investing additional time and computational resources to improve data quality is justified.</p><p><strong>Methods: </strong>We developed a rudimentary error rate metric to evaluate textual dataset quality at the token level. The Mixtral large language model (LLM) was used to quantify and correct errors in low-quality datasets. The study analyzed two health care datasets: the high-quality MIMIC-III public hospital dataset (for mortality prediction) and a lower-quality private dataset from Australian aged care homes (AACHs; for depression and fall risk prediction). Errors were systematically introduced into MIMIC-III at varying rates, while the AACH dataset quality was improved using the LLM. Feature representations and machine learning models were assessed using the area under the receiver operating curve.</p><p><strong>Results: </strong>For the sampled 35,774 and 6336 patients from the MIMIC and AACH datasets, respectively, we used Mixtral to introduce errors in MIMIC and correct errors in AACH. Mixtral correctly detected errors in 63% of progress notes, with 17% containing a single token misclassified due to medical terminology. LLMs demonstrated potential for improving progress note quality by addressing various errors. Under varying error rates (5%-20%, in 5% increments), feature representation performance was tolerant to lower error rates (<10%) but declined significantly at higher rates. This aligned with the AACH dataset's 8% error rate, where no major performance drop was observed. Across both datasets, term frequency-inverted document frequency outperformed embedding features, and machine learning models varied in effectiveness, highlighting that optimal feature representation and model choice depend on the specific task.</p><p><strong>Conclusions: </strong>This study revealed that models performed relatively well on datasets with lower error rates (<10%), but their performance declined significantly as error rates increased (≥10%). Therefore, it is crucial to evaluate the quality of ","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e73325"},"PeriodicalIF":6.0,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12811037/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145862942","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ayesha Thanthrige, Bruce Lu, Zaid Sako, Nilmini Wickramasinghe
<p><strong>Background: </strong>Health care technology adoption is key to improving patient care, enhancing operational efficiency, and ensuring better health outcomes. Examining the determinants that influence the acceptance and sustainable use of health care technologies is crucial for system developers, health care providers, and policymakers. The Unified Theory of Acceptance and Use of Technology (UTAUT) and task-technology fit (TTF) theoretical models offer a comprehensive framework to assess these determinants systematically, with UTAUT focusing on usage intentions (UI) and TTF emphasizing task-technology alignment for system usefulness, usability, and satisfaction.</p><p><strong>Objective: </strong>This systematic review and meta-analysis aimed to identify and analyze the key factors influencing the adoption of health care technologies based on an integrated UTAUT and TTF framework. By synthesizing existing literature, the study seeks to provide valuable insights for stakeholders to implement innovative and effective solutions in the health care domain.</p><p><strong>Methods: </strong>A search was conducted across a range of databases, including MEDLINE and Embase, IEEE Xplore, ScienceDirect, Scopus, CINAHL, Google Scholar, and Web of Science. Inclusion criteria covered studies applying either the UTAUT model, the TTF model, or both to health care technology adoption, published in English between 2012 and 2025. Exclusion criteria included nonquantitative studies, studies not focused on a health care setting, and those lacking sufficient data for meta-analysis. The reviewers collaborated to decide on the final papers for inclusion in the review through Covidence, the Cochrane Collaboration's platform for systematic reviews. Data collection involved extracting quantitative data (eg, sample sizes, reliabilities, and standardized path coefficients) analyzed using meta-analytic techniques with a random-effects model in R software (R Development Core Team) to combine findings and calculate effect sizes.</p><p><strong>Results: </strong>A total of 50 studies (35 UTAUT with 20,723 participants and 15 TTF with 4041 participants) met the inclusion criteria, representing various health care technologies, such as electronic health records, telemedicine platforms, and mobile health apps. The meta-analysis revealed that performance expectancy emerged as the most significant predictor of UI (β=.304; P<.001), while UI was the primary predictor of usage behavior (β=.199; P<.001). Other UTAUT predictors included effort expectancy (β=.177; P<.001), social influence (β=.167; P<.001), and facilitating conditions (β=.105; P<.001). For TTF, technology characteristics had the strongest effect on TTF (β=.445; P<.001), followed by TTF on UI (β=.271; P<.001) and task characteristics on TTF (β=.263; P<.001). Variability across settings and regions suggests contextual influences, with high heterogeneity (I²=81.90%-94.87%).</p><p><strong>Conclusions: </strong>This study
背景：卫生保健技术的采用是改善患者护理、提高操作效率和确保更好的健康结果的关键。检查影响接受和可持续使用卫生保健技术的决定因素对系统开发者、卫生保健提供者和政策制定者至关重要。技术接受和使用统一理论（UTAUT）和任务-技术契合（TTF）理论模型提供了一个全面的框架来系统地评估这些决定因素，UTAUT侧重于使用意图（UI），而TTF强调任务-技术对系统有用性、可用性和满意度的一致性。目的：本系统综述和荟萃分析旨在识别和分析影响医疗保健技术采用的关键因素，基于综合UTAUT和TTF框架。通过综合现有文献，本研究旨在为利益相关者在医疗保健领域实施创新和有效的解决方案提供有价值的见解。方法：检索MEDLINE和Embase、IEEE explore、ScienceDirect、Scopus、CINAHL、b谷歌Scholar和Web of Science等数据库。纳入标准包括在2012年至2025年期间以英文发表的应用UTAUT模型、TTF模型或两者均适用于医疗保健技术采用的研究。排除标准包括非定量研究、不关注卫生保健环境的研究和缺乏足够数据进行meta分析的研究。审稿人合作决定通过Cochrane协作组织的系统评价平台covid - ence纳入审稿的最终论文。数据收集涉及提取定量数据（例如，样本量、可靠性和标准化路径系数），使用R软件（R Development Core Team）中的随机效应模型的元分析技术进行分析，以组合结果并计算效应大小。结果：共有50项研究（35项UTAUT， 20,723名参与者，15项TTF， 4041名参与者）符合纳入标准，代表了各种医疗技术，如电子健康记录、远程医疗平台和移动健康应用程序。结论：通过整合UTAUT和TTF，本研究为促进医疗保健技术采用提供了有价值的见解，突出了绩效期望、努力期望、社会影响、促进条件、任务特征、技术特征和TTF是关键驱动因素。调查结果评估了系统的有用性、可用性和满意度，可以指导干预措施，以改善采用率和卫生保健服务。
{"title":"Determinants of Health Care Technology Adoption Using an Integrated Unified Theory of Acceptance and Use of Technology and Task Technology Fit Model: Systematic Review and Meta-Analysis.","authors":"Ayesha Thanthrige, Bruce Lu, Zaid Sako, Nilmini Wickramasinghe","doi":"10.2196/64524","DOIUrl":"10.2196/64524","url":null,"abstract":"<p><strong>Background: </strong>Health care technology adoption is key to improving patient care, enhancing operational efficiency, and ensuring better health outcomes. Examining the determinants that influence the acceptance and sustainable use of health care technologies is crucial for system developers, health care providers, and policymakers. The Unified Theory of Acceptance and Use of Technology (UTAUT) and task-technology fit (TTF) theoretical models offer a comprehensive framework to assess these determinants systematically, with UTAUT focusing on usage intentions (UI) and TTF emphasizing task-technology alignment for system usefulness, usability, and satisfaction.</p><p><strong>Objective: </strong>This systematic review and meta-analysis aimed to identify and analyze the key factors influencing the adoption of health care technologies based on an integrated UTAUT and TTF framework. By synthesizing existing literature, the study seeks to provide valuable insights for stakeholders to implement innovative and effective solutions in the health care domain.</p><p><strong>Methods: </strong>A search was conducted across a range of databases, including MEDLINE and Embase, IEEE Xplore, ScienceDirect, Scopus, CINAHL, Google Scholar, and Web of Science. Inclusion criteria covered studies applying either the UTAUT model, the TTF model, or both to health care technology adoption, published in English between 2012 and 2025. Exclusion criteria included nonquantitative studies, studies not focused on a health care setting, and those lacking sufficient data for meta-analysis. The reviewers collaborated to decide on the final papers for inclusion in the review through Covidence, the Cochrane Collaboration's platform for systematic reviews. Data collection involved extracting quantitative data (eg, sample sizes, reliabilities, and standardized path coefficients) analyzed using meta-analytic techniques with a random-effects model in R software (R Development Core Team) to combine findings and calculate effect sizes.</p><p><strong>Results: </strong>A total of 50 studies (35 UTAUT with 20,723 participants and 15 TTF with 4041 participants) met the inclusion criteria, representing various health care technologies, such as electronic health records, telemedicine platforms, and mobile health apps. The meta-analysis revealed that performance expectancy emerged as the most significant predictor of UI (β=.304; P<.001), while UI was the primary predictor of usage behavior (β=.199; P<.001). Other UTAUT predictors included effort expectancy (β=.177; P<.001), social influence (β=.167; P<.001), and facilitating conditions (β=.105; P<.001). For TTF, technology characteristics had the strongest effect on TTF (β=.445; P<.001), followed by TTF on UI (β=.271; P<.001) and task characteristics on TTF (β=.263; P<.001). Variability across settings and regions suggests contextual influences, with high heterogeneity (I²=81.90%-94.87%).</p><p><strong>Conclusions: </strong>This study","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e64524"},"PeriodicalIF":6.0,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12753102/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145862937","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Andreas Nielsen Hald, Peter Fallesen, Frank Eriksson, Gert Martin Hald
<p><strong>Background: </strong>Digital health interventions are increasingly promoted as scalable and cost-effective approaches to support mental health and resilience. Short-term benefits are well documented, but evidence on long-term outcomes (beyond 12 mo) remains scarce, particularly when assessed with objective measures in large cohorts. Most studies to date have focused on small samples, relied on self-reported outcomes, and used follow-up periods of less than a year. This leaves uncertainty about whether early changes are sustained over time and whether they can be observed in objective indicators of health. This gap is particularly relevant for stressful life transitions, where the risk of long-term adverse health outcomes is high. Divorce, a common and stressful transition linked to poorer mental and physical health, thus provides an ideal case for investigating the long-term potential of digital health interventions.</p><p><strong>Objective: </strong>This study examined the association between SES One, a digital health intervention for Danish divorcees, and mental health medication use, primary care usage, and hospitalizations over a 5-year follow-up period using Danish national health registers.</p><p><strong>Methods: </strong>Participants (n=1856) from a randomized controlled trial of SES One in Denmark were followed for 5 years after divorce. Outcomes included mental health medication prescriptions (eg, antipsychotics, anxiolytics, hypnotics, sedatives, and antidepressants), primary care usage (eg, billable interactions with general practitioners, specialist practitioners, and psychologists), and hospitalizations. Odds ratios and incidence rate ratios were calculated to compare outcomes between SES One participants and the control group.</p><p><strong>Results: </strong>Over 5 years, SES One participants did not have significantly lower odds of filling a prescription (odds ratio [OR] 0.836; P=.09) but filled 28% fewer prescriptions overall (incidence rate ratio 0.720; P=.045), indicating a reduce-not-remove effect. No overall differences were observed in primary care usage or hospitalizations. However, participants had 38% (OR 0.624, P=.003) and 27% (OR 0.730, P=.001) lower odds of visiting primary care in years 2 and 3, respectively, and 32% (OR 0.677, P=.046) lower odds of hospitalization in year 4, suggesting possible late-onset effects.</p><p><strong>Conclusions: </strong>The findings advance the field by showing that a targeted digital health intervention can generate measurable long-term health benefits in a large cohort when evaluated with objective registry data. The results suggest that such interventions may reduce reliance on medication and health care services over time, not by eliminating needs entirely but by reducing them. These patterns can be interpreted as reflecting both legacy and late-onset pathways. Long-term evaluations with objective data are essential to fully capture the durability and timing of digital heal
{"title":"Digital Health Intervention for and Long-Term Health Outcomes of a Divorce Cohort With Linked Danish Data: 5-Year Posttrial Follow-Up of a Randomized Controlled Trial.","authors":"Andreas Nielsen Hald, Peter Fallesen, Frank Eriksson, Gert Martin Hald","doi":"10.2196/69387","DOIUrl":"10.2196/69387","url":null,"abstract":"<p><strong>Background: </strong>Digital health interventions are increasingly promoted as scalable and cost-effective approaches to support mental health and resilience. Short-term benefits are well documented, but evidence on long-term outcomes (beyond 12 mo) remains scarce, particularly when assessed with objective measures in large cohorts. Most studies to date have focused on small samples, relied on self-reported outcomes, and used follow-up periods of less than a year. This leaves uncertainty about whether early changes are sustained over time and whether they can be observed in objective indicators of health. This gap is particularly relevant for stressful life transitions, where the risk of long-term adverse health outcomes is high. Divorce, a common and stressful transition linked to poorer mental and physical health, thus provides an ideal case for investigating the long-term potential of digital health interventions.</p><p><strong>Objective: </strong>This study examined the association between SES One, a digital health intervention for Danish divorcees, and mental health medication use, primary care usage, and hospitalizations over a 5-year follow-up period using Danish national health registers.</p><p><strong>Methods: </strong>Participants (n=1856) from a randomized controlled trial of SES One in Denmark were followed for 5 years after divorce. Outcomes included mental health medication prescriptions (eg, antipsychotics, anxiolytics, hypnotics, sedatives, and antidepressants), primary care usage (eg, billable interactions with general practitioners, specialist practitioners, and psychologists), and hospitalizations. Odds ratios and incidence rate ratios were calculated to compare outcomes between SES One participants and the control group.</p><p><strong>Results: </strong>Over 5 years, SES One participants did not have significantly lower odds of filling a prescription (odds ratio [OR] 0.836; P=.09) but filled 28% fewer prescriptions overall (incidence rate ratio 0.720; P=.045), indicating a reduce-not-remove effect. No overall differences were observed in primary care usage or hospitalizations. However, participants had 38% (OR 0.624, P=.003) and 27% (OR 0.730, P=.001) lower odds of visiting primary care in years 2 and 3, respectively, and 32% (OR 0.677, P=.046) lower odds of hospitalization in year 4, suggesting possible late-onset effects.</p><p><strong>Conclusions: </strong>The findings advance the field by showing that a targeted digital health intervention can generate measurable long-term health benefits in a large cohort when evaluated with objective registry data. The results suggest that such interventions may reduce reliance on medication and health care services over time, not by eliminating needs entirely but by reducing them. These patterns can be interpreted as reflecting both legacy and late-onset pathways. Long-term evaluations with objective data are essential to fully capture the durability and timing of digital heal","PeriodicalId":16337,"journal":{"name":"Journal of Medical Internet Research","volume":"27 ","pages":"e69387"},"PeriodicalIF":6.0,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12753028/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145862914","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：