Therapist-guided internet-delivered cognitive behavioral therapy (guided iCBT) is increasingly implemented in routine mental health care to expand access to evidence-based treatments. Although the clinical effectiveness and patient acceptability of guided iCBT for common mental health disorders are well established, less is known about how introducing such digitally mediated interventions reshapes therapists' everyday work practices and professional roles. Existing research has primarily focused on treatment outcomes, service uptake, and organizational implementation challenges, leaving therapists' lived professional experiences and role enactment in routine practice underexplored. This study aimed to explore how the introduction of guided iCBT in routine clinical practice influences mental health care professionals' perceptions and enactment of their professional roles, and to examine how these roles are shaped by interactions between technology and health care professionals. We conducted an exploratory qualitative study using semistructured interviews and observations with 31 health care professionals who delivered guided iCBT in specialized mental health care settings between January and December 2022. Participants were recruited through purposive and snowball sampling, and data were analyzed using reflexive thematic analysis. The study was theoretically informed by actor-network theory and sociological perspectives on professional roles, enabling a sociotechnical interpretation of how professional work is shaped through interactions between human and nonhuman actors. Two overarching themes described how guided iCBT reconfigures therapeutic work. First, participants reported increased professional agency through enhanced flexibility and autonomy, including greater control over work schedules and opportunities to tailor standardized treatment elements to individual patients. Second, the therapist's role expanded beyond traditional clinical tasks, encompassing increased administrative responsibilities ("the administrative therapist") and expectations to actively promote and legitimize guided iCBT to patients ("the professional salesperson"). While guided iCBT enabled more flexible and individualized care, it also introduced new constraints through standardized protocols, digital platforms, and organizational expectations, resulting in professional discretion and authority being actively negotiated rather than simply reduced or maintained. This study offers a practice-oriented, sociotechnical account of how guided iCBT is implemented in routine care. The findings show how therapists' roles are shaped by everyday clinical workflows, in which responsibility, decision-making, and clinical judgment are distributed among therapists, patients, digital platforms, treatment protocols, and organizational arrangements. The concept of distributed therapeutic authority offers a useful lens for understanding how therapeutic work is reorganized in digitally mediated care without implying a loss of professional expertise. Professional judgment remains central but is exercised in interaction with digital tools and standardized treatment structures. These findings underscore the importance of aligning digital systems, clinical routines, and professional roles when implementing guided iCBT to support both access to care and the quality of therapeutic practice.
Historic and ongoing problematic practices regarding the collection, storage, and use of Indigenous health data have led to the need to ensure principles of Indigenous Data Sovereignty (IDS) are followed in research practices and technology development. This project, a partnership between UC San Diego and the Native BioData Consortium (NativeBio), sought to explore the practical application of blockchain technology and its potential to facilitate Indigenous-led research collaboration. This project first undertook purposeful relationship building with NativeBio to form a Community Advisory Board (CAB) for identifying community and technology needs for a blockchain research collaboration platform with an initial focus on genomic data. Over a 2-year project period, a series of public meetings and presentations at Indigenous-led conferences introduced the concept of exploring compatibility between blockchain and IDS principles, followed by iterative prototyping and co-design of a blockchain platform with NativeBio, using Ethereum as the underlying protocol. Direct engagement with NativeBio and the CAB informed the initial design and development of a "b-IDS" proof-of-concept (POC) blockchain platform. The POC consists of three main components: (1) the web front-end layer, (2) the Ethereum network that executes the smart contract and blockchain storage aspects of the framework, and (3) the back-end database that stores off-chain interactions and data for future use with external genomic data repositories. After refinement of the POC, a community-based participatory research (CBPR) use case aligned with IDS principles was identified as a practical workflow and incorporated into the design of the POC for implementation. The findings from this project demonstrated the potential use of operationalizing IDS through blockchain technology with proactive and sustained engagement with Indigenous partners. Blockchain technology may have certain advantages over other data governance approaches and systems, facilitating timely oversight, shared decision-making and consent structures, and direct involvement of Indigenous communities in technology design, respecting the core principles of IDS and CBPR. Future development of the blockchain-IDS POC will need to incorporate other research practices and ethics frameworks to expand its use to other public health and biomedical research use cases.
Neuromuscular diseases (NMD) affect nerves and muscles, resulting in weakness and often profound disability. Family caregivers of individuals with NMD experience significant psychological burden, stress, and reduced well-being. Digital peer support interventions may help to ameliorate these negative impacts. This study evaluated the effect of a 12-week digital peer support intervention compared to usual care on caregiver mastery, competence, stress, burden, anxiety, and depression among family caregivers of individuals with NMD. We conducted a parallel-group multicenter randomized controlled superiority trial in Ontario, Canada. Family caregivers of children or adults with NMD were recruited between August 2022 and September 2023 through 7 sites, social media, and national organizations. Participants were randomized 1:1 to a 12-week digital peer support intervention or usual care. The 12-week intervention comprised access to a trained peer mentor, private app-based communication via aTouchAway (Aetonix, Canada), and weekly moderated digital group discussion forums. The primary outcome was caregiver mastery measured using the Pearlin Mastery Scale, adjusted for baseline score. Secondary outcomes included caregiver stress, competence, burden, anxiety, and depression. We calculated adjusted (for baseline score) mean differences using analysis of covariance and generated multivariable linear regression models exploring associations with the intervention and caregiver age, years of caregiving, care recipient medical diagnosis, care recipient ventilation type, adjusting for baseline outcome scores. Intervention fidelity was evaluated through participant engagement metrics. A total of 100 participants were randomized (n=50 intervention and n=50 control). Participants had a mean age of 46.8 (SD 11.6) years, 70% (n=70) were mothers, with a mean length of caregiving of 11.8 (SD 7.6) years. We found no difference in 12-week Pearlin Mastery Scale scores (adjusted mean difference 0.67, 95% CI -1.7 to 3.1). We also found no difference in any of our secondary outcomes. Mentors and participants sent a mean of 21.3 (SD 33.3) and 17.7 (SD 33.0) messages, respectively. Overall, 62% (n=31) of participants and 92% (n=11) of mentors engaged in at least 1 program element for ≥8 of the 12 weeks. Our 12-week digital peer support program had no effect on caregiver mastery or other caregiving or psychological outcomes among family caregivers of individuals with NMD. This might be partly due to moderate fidelity and variability in participant engagement. Unlike prior caregiver interventions that incorporated structured psychoeducation or self-management training, this intervention evaluated primarily peer support delivered through a digital platform. This study contributes important evidence regarding the feasibility and limitations of scalable digital peer support programs for caregivers of individuals with NMD. These findings highlight the importance of intervention tailoring, participant matching, and sustained engagement. Future research should evaluate longer-duration and more individualized peer support models targeting caregivers earlier in the caregiving trajectory to improve intervention fidelity and ultimately caregiver well-being.
The rapid emergence of artificial intelligence (AI) has outpaced its formal adoption in health care organizations, contributing to the emergence of Shadow AI, defined here as the use of unauthorized AI tools by medical professionals. Under the European Union Medical Device Regulation, AI tools used for clinical purposes must undergo conformity assessment before use; general-purpose tools such as ChatGPT have not done so, rendering their clinical application unauthorized at the regulatory level. While Shadow AI offers potential efficiency gains and higher performance, it poses significant risks to data privacy, clinical safety, and regulatory compliance. Despite its growing prevalence, empirical research on the purposes for which physicians use Shadow AI remains scarce. This study explores the purposes for which physicians describe using Shadow AI in their work. We conducted a cross-sectional survey of physicians employed in Swedish health care organizations (N=357; response rate~64%). Data were collected between December 2023 and January 2024 via a verified online panel. We conducted a qualitative content analysis of free-text responses on the use of unauthorized AI tools. We applied theoretical lenses from the sociology of professions and paradox theory to interpret the empirical findings. Physicians use Shadow AI for several purposes, which we grouped into 4 categories: clinical work and decision-making, administrative work, research and professional development, and technological interest and curiosity. More specifically, Shadow AI is used as a colleague and second opinion for clinical decision support (eg, differential diagnoses and rare cases), administrative tasks such as patient communication and documentation, and research aimed at staying up to date and exploring developments in generative AI. Physicians described using these tools compensated for perceived gaps in institutional systems, reducing workload, and accessing knowledge considered difficult to obtain through conventional channels. The findings reveal a tension between physicians' drive to improve their practice and the regulatory and organizational constraints that render such use unauthorized. Shadow AI used by physicians presents both opportunities and risks for health care professionals and organizations. Shadow AI indicates gaps where formal hospital systems may fail to meet health care professionals' needs and signals a way for physicians to strengthen their experience-based knowledge. It represents a renegotiation of professional boundaries, as physicians bypass institutional constraints to maintain professional efficacy. The findings highlight a paradox in which the same tools that pose regulatory and safety risks also address real gaps in clinical and administrative support, suggesting that governance approaches must account for this tension rather than relying on prohibition alone.
Sleep apnea (SA) is a serious sleep disorder, and its diagnostic gold standard, polysomnography, is costly and time-consuming. Electroencephalogram (EEG) signals, due to their direct correlation with neural activity and ease of extraction, represent a promising tool. Despite increasing research on machine learning (ML) and deep learning for EEG-based SA detection, model performance has not been consistently evaluated. This systematic review evaluated the accuracy of ML in detecting SA from EEG data and provided an evidence base for further clinical application and future research. Following the PRISMA-DTA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses of Diagnostic Test Accuracy) and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 expanded checklists, we systematically searched PubMed, Embase, Web of Science, Cochrane Library (CENTRAL), Scopus, IEEE Xplore, and ClinicalTrials.gov databases from inception to April 2026. Studies evaluating the value of ML algorithms for detecting SA based only on EEG data were included. The Quality Assessment of Diagnostic Accuracy Studies-2 and Prediction Model Risk of Bias Assessment Tool for Artificial Intelligence tools were used to assess the risk of bias in each study. Statistical analysis was performed using the mada and metafor packages in R (version 4.6.0; R Foundation for Statistical Computing) and the Meta-DiSc (version 1.4; Hospital Ramón y Cajal) software. We used GRADE (Grading of Recommendations Assessment, Development and Evaluation) to evaluate the certainty of evidence. A total of 27 retrospective studies were included. Segment-level analyses showed high diagnostic performance, with a pooled sensitivity of 0.90 (95% CI 0.85-0.94; 95% prediction interval 0.43-0.99) and specificity of 0.92 (95% CI 0.87-0.95; 95% prediction interval 0.46-0.99). The pooled area under the summary receiver operating characteristic curve was 0.95 (95% CI 0.92-0.99). Meta-regression identified EEG channel configuration, region, and validation strategy as significant sources of heterogeneity (P=.004, P=.003, and P=.046, respectively). Multichannel EEG, deep learning approaches, and hold-out validation strategies generally demonstrated better diagnostic performance. Only 2 studies evaluated patient-level diagnostic performance, which was summarized qualitatively. To our knowledge, this is the first systematic review and meta-analysis specifically focused on the diagnostic accuracy of EEG-based ML models in the detection of SA. This meta-analysis indicates that ML models based on EEG demonstrate good diagnostic accuracy in detecting SA at the segment level and show promise as tools for SA screening and clinical decision support. However, most current studies are retrospective segment-level analyses, which may overestimate the practical value of this technology in real-world clinical settings. To reliably integrate EEG-based ML models into clinical diagnostic workflows, further prospective studies incorporating full-night monitoring and patient-level validation are needed.
Motivation to quit smoking and decisions to smoke or forgo smoking vary throughout the day. However, little is known about how within-day patterns of psychological states such as self-efficacy and attitudes toward smoking relate to these determinants of smoking cessation attempts. Identifying these dynamic processes can inform the development of more precisely timed and tailored digital interventions. This study aimed to identify distinct within-day trajectories of self-efficacy for cutting down on cigarettes smoked and attitudes toward smoking, and to examine how these trajectories predicted end-of-day motivation to quit and same-day cigarette forgoing (ie, choosing not to smoke cigarettes that one would normally smoke). People who smoked at least 10 cigarettes a day at baseline (N=348, mean age 44.6, SD 12.1 years; n=212, 60.9% female) received smartphone surveys about 4-5 times a day after logging each cigarette, producing 15,614 surveys over 2561 days. Trajectories of self-efficacy and smoking attitudes were modeled at the person-day level using smooth functions, and 6 daily parameters of change (overall level, range of change, volatility, overall trend, acceleration of change, and trajectory shape [trend×acceleration]) were extracted. These parameters were then entered as predictors of (1) end-of-day motivation to quit (linear mixed models) and (2) whether participants forwent cigarettes that day (binomial generalized linear mixed models). Higher overall self-efficacy consistently predicted both greater end-of-day motivation and greater odds of forgoing. Upward trends and acceleration in self-efficacy further predicted greater odds of forgoing, indicating that days when confidence not only increased but did so quicker were most strongly associated with forgoing cigarettes that day. Less favorable attitudes toward smoking predicted greater motivation to quit and increased likelihood of forgoing cigarettes. Broader ranges of daily change in attitudes were linked with stronger motivation to quit and greater odds of forgoing, while more moment-to-moment volatility was associated with reduced odds of forgoing cigarettes that day. Dynamic features of self-efficacy and smoking attitudes, such as overall level, trend, and acceleration, were robust predictors of daily motivation to quit and cigarette forgoing. These findings highlight that the way self-efficacy and attitudes shift across the day is meaningful beyond their overall levels. Just-in-time adaptive interventions may be more effective if they monitor and respond to varying trajectory features rather than focusing on static states, supporting a shift toward dynamically aware intervention strategies in digital health.
Sleep disorders represent a significant public health burden associated with cardiovascular and neurocognitive morbidities. While AI technologies offer potential for personalized sleep medicine, clinical integration remains limited. This translational disparity is often attributed to a lack of human-centered design, specifically insufficient stakeholder engagement in the development and implementation of these technologies. Current research frequently prioritizes algorithmic performance over usability and patient trust. This scoping review systematically maps the extent and nature of human-centered AI (HCAI) research within sleep medicine across different AI modalities, evaluating how diverse stakeholders are involved in the design, validation, and implementation of AI tools, including patients, clinicians, and technologists. Following the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines, we searched 8 databases (PubMed, Web of Science, Embase, Scopus, IEEE Xplore, ACM Digital Library, APA PsycINFO, and CINAHL) for literature published up to June 18, 2026. We identified primary research describing the design, development, or evaluation of AI technologies for sleep health with explicit human-centered components. Included studies (n=34) were categorized based on AI technology type and the method of stakeholder engagement. Data were extracted and synthesized using a thematic analysis approach. Based on the included studies, the analysis reveals an uneven distribution of research focus across technological domains as descriptive patterns rather than definitive trends. Research on generative AI (GenAI) is predominantly restricted to downstream expert auditing of output accuracy (comprising 7/11, 64% of GenAI studies), with a noticeable gap in upstream participatory design involving patients. Conversely, deep learning research primarily focuses on technical explainable AI methods to address algorithmic opacity for clinicians, yet lacks progression to real-world clinical implementation. Mobile health and wearable technologies (17/34, 50%) demonstrate the most balanced HCAI ecosystem, evidencing a complete translational cycle from upstream co-design to downstream clinical implementation. Furthermore, an emerging trend is observed where AI is evolving from an automated diagnostic tool into an interactive therapeutic agent, with recent studies indicating that lay users may perceive responses from large language models as more empathetic than those from physicians. Lacking formal quality appraisal, our findings reflect research activity patterns rather than confirmed clinical effectiveness. Nevertheless, this scoping review innovatively applies the HCAI framework to the sleep AI lifecycle. Unlike existing reviews prioritizing algorithmic performance metrics over usability, clinical workflow integration, and patient trust, this study systematically maps these essential sociotechnical factors. It contributes to the field by revealing distinct methodological disparities and the urgent need for upstream participatory design, particularly for GenAI. In the real world, establishing standardized protocols for human-AI interaction, ensuring algorithmic transparency, and addressing demographic biases are essential to foster the clinical trust required for effective AI adoption.
Chronic low back pain (CLBP) is a major global health challenge. While nonpharmacological therapies are recommended, patient compliance is often hindered by kinesiophobia. Virtual reality (VR) offers an immersive, distraction-based approach, but the comparative effectiveness of different VR modalities remains unclear. The aim of the study is to compare and rank the efficacy of different VR-based training modalities on pain intensity, disability, and kinesiophobia in patients with CLBP. Systematic searches were conducted in PubMed, Web of Science, Scopus, Embase, CINAHL, and the Cochrane Library from inception until June 2025. Randomized controlled trials (RCTs) assessing the effects of VR-based training on individuals with CLBP were selected. Primary outcomes were pain intensity, disability (Oswestry Disability Index), and kinesiophobia (Tampa Scale of Kinesiophobia). The Cochrane Risk of Bias 2 tool was used for quality assessment. Confidence in Network Meta-Analysis (CINeMA) framework was used to evaluate the credibility of cumulative evidence. A Bayesian network meta-analysis with standardized mean difference (SMD) as effect size was performed to synthesize evidence and rank interventions using surface under the cumulative ranking curve values. The GRADE (Grading of Recommendations Assessment, Development and Evaluation) framework was adapted to evaluate the quality of evidence. In total, 25 RCTs with a total of 2610 participants were included in the analysis. For pain intensity, shooting games (SMD -4.40, 95% credible interval [CrI] -6.80 to -2.20) and VR-based equestrian training (SMD -2.00, 95% CrI -3.70 to -0.57) were significantly superior to all types of controls. Surface under the cumulative ranking curve indicated that shooting games had the highest probability (98%) of being the most effective intervention for pain relief. For disability, no intervention demonstrated statistically significant superiority. For kinesiophobia, shooting games (SMD -3.40, 95% CrI -5.60 to -1.10) significantly outperformed traditional exercise controls. The quality of evidence ranged from very low to moderate across outcomes. This first network meta-analysis to compare and rank distinct VR modalities for CLBP offers several key innovations and contributions to the field. By moving beyond aggregate VR categorizations, we provide a granular, comparative ranking of specific, actionable VR interventions. Unlike previous reviews that treated VR as a homogeneous group or only compared it to sham, our network meta-analysis directly and indirectly compares 7 distinct VR modalities, revealing that not all VR is equally effective. Our findings suggest that shooting games have the potential to be the most effective VR therapy for relieving pain intensity and kinesiophobia, though evidence for disability remains limited. Unfortunately, due to heterogeneity and low-quality evidence, there is no evidence demonstrating significant improvement in specific outcomes for patients with CLBP. More RCTs are needed to provide robust clinical evidence.
Spiritual care providers are increasingly challenged to address the introduction of AI within the ethical, theological, and organizational bounds of their employers and religious or worldview communities. However, empirical data on professional perspectives regarding AI implementation in spiritual care remain scarce. This study aimed to conduct the first empirical investigation of AI use cases, risks and benefits, theological and ethical considerations, and relevant professional competencies from a multistakeholder expert perspective. An international, multistakeholder modified Delphi study was conducted in 2 rounds. A purposive sample of 149 subject experts was recruited. Panelists rated 213 items spanning task assistance and substitution, risks and benefits, theological and ethical considerations, limits, and competencies. Consensus was defined using combined measures of variance and directionality. Exploratory subgroup analyses assessed whether ratings differed across professional and demographic groups. Round 1 was completed by 102 of 149 invited panelists (response rate 68.5%); round 2 was completed by 83 panelists (response rate 81.4%). In round 2, strong agreement emerged that AI can currently assist with or enhance administrative and routine tasks (77/81, 95.1%), informational tasks (74/79, 93.7%), documentation (67/80, 83.8%), and spiritual care research (65/77, 84.4%). Agreement was lower for relational, patient-facing tasks such as creating supportive spaces (30/69, 43.5%), direct patient engagement (32/76, 42.1%), and conducting ritual tasks (32/76, 42.1%). Panelists favored AI assistance over substitution and rated the future potential of AI above its current capabilities. The highest-ranked benefit that reached consensus was improved screening and triage; the highest-ranked risk was loss of human connection. Overall, 45.8% (38/83) of panelists judged the benefits of AI in spiritual care to outweigh the risks. The panel converged strongly on professional limits but was more divided on the underlying theological and ethical objections. The most strongly endorsed competencies were judging AI's applicability and the boundaries of its use, safeguarding patient privacy and data, and assessing and mitigating risks. Subgroup analyses produced few robust differences. This study provides the first task-level map of expert opinion on AI in spiritual care, suggesting that current expert views are task-specific, future-oriented, and conditional. AI is seen as capable of assisting with administrative, informational, documentation, and research tasks. Direct relational care is widely regarded as a primarily human responsibility, and several tasks that experts judged AI capable of assisting with or substituting remain ethically complex. Experts converged on practical safeguards, including human oversight, evidence-based implementation, privacy protection, and limits on AI decision-making, and on required competencies, even as the theological and ethical rationales behind them remained contested. This suggests that professional guidance may be achievable at an early stage of AI adoption and can inform future research, guideline development, and curriculum design.
Assistive technology (AT) can significantly enhance functional abilities among persons with disabilities. Recently, a rapid advance in computer and Internet of Things (IoT)-based AT has been observed with the integration of Artificial Intelligence. The present scoping review explores a range of IoT-based AT for vision impairments, focusing on their functioning, reliability, and identifying the existing gaps and opportunities for future enhancement. We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. The search was conducted for the articles published between 2020 and March 2025. After the initial screening of both titles and abstracts of 705 articles, 18 full-text articles were selected for the review. It was found that the majority of the papers (38%) are related to smart canes/sticks. The findings indicate that several studies use Raspberry Pi and Arduino platform to make products cost-effective and flexible options for real-time processing. The accuracy of these devices ranges from 87.8% in navigation to 93% in voice control and 100% in object identification. Further, challenges such as usability testing, hardware limitations, dataset constraints, and environmental adaptability remain a matter of concern which need to be addressed in future innovation. Incorporating IoT technologies into vision-assistive devices has substantially improved accessibility. Future research should focus on enhancing real-time performance, AI-driven decision-making, and user-centric designs.
In addition to core behavioral symptoms, children and adolescents with autism spectrum disorder (ASD) frequently exhibit impairments in executive function and motor performance. Although virtual reality (VR)-based physical exercise interventions are increasingly used in ASD rehabilitation, evidence regarding their multidimensional effects remains limited. This study aimed to systematically review the effects of VR-based physical exercise interventions on behavioral outcomes, executive function, and motor performance in children and adolescents with ASD. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 guidelines were followed. PubMed (National Library of Medicine), Embase (Elsevier), Web of Science, Scopus (Elsevier), and other databases were searched from inception to May 7, 2026. Eligible studies included randomized and nonrandomized trials involving participants aged 6-18 years with ASD. The interventions consisted of VR-based exercise programs involving physical activity participation, including active video games, motion-sensing interactive training, and augmented reality-based exercise training. Risk of bias was assessed using Risk of Bias 2 (RoB 2; Cochrane Bias Methods Group) and Risk of Bias in Nonrandomized Studies of Interventions (ROBINS-I; Cochrane Bias Methods Group), and certainty of evidence was evaluated using the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) framework. Random-effects meta-analyses were conducted using Hedges g standardized mean differences (SMDs). A total of 15 studies involving 439 children and adolescents with ASD were included; among them, 9 studies were eligible for meta-analysis. Behavioral outcomes were reported in only 3 studies. Owing to substantial heterogeneity in study designs and assessment instruments, we did not conduct a meta-analysis for these outcomes. Current evidence suggests inconsistent effects on social interaction and stereotyped behaviors. VR-based physical exercise interventions may improve executive function (SMD 0.75, 95% CI 0.32-1.18; I²=0.0%; P=.01; GRADE moderate). However, the prediction interval crossed the line of no effect (-0.12 to 1.63). VR-based physical exercise interventions were associated with improvements in motor performance (SMD 1.08, 95% CI 0.08-2.08; I²=74.3%; P=.04; GRADE low). However, the wide prediction interval suggests substantial between-study variability (-1.31 to 3.47). Current evidence suggests a relatively consistent positive effect of VR-based physical exercise interventions involving physical activity participation on executive function in children and adolescents with ASD. However, substantial uncertainty remains regarding their effects on motor performance and behavioral outcomes. Unlike previous reviews that primarily focused on general VR interventions, social skills training, or single functional outcomes, this systematic review specifically examined the multidimensional effects of VR-based physical exercise interventions. The findings suggest that VR-based physical exercise interventions may be implemented as adjuncts to conventional exercise or rehabilitation programs rather than as stand-alone interventions. Future large-scale, high-quality randomized controlled trials with larger sample sizes, standardized intervention reporting, and long-term follow-up are needed to further clarify the optimal implementation conditions and underlying mechanisms of different forms of VR-based exercise training.
Digital health interventions (DHIs) are increasingly used to support antiretroviral therapy (ART) management among people living with HIV. However, existing systematic reviews have largely focused on single intervention types or limited outcomes, and few have integrated multiple DHI modalities across both behavioral and clinical end points. Additionally, previous evidence has rarely incorporated analytical approaches, such as prediction intervals (PIs) or trial sequential analysis (TSA), leaving uncertainty regarding the robustness and generalizability of findings. This systematic review aimed to evaluate the effectiveness of DHIs in improving ART-related outcomes among people living with HIV. We systematically searched PubMed, Cochrane Library, Embase, and Web of Science for randomized controlled trials (RCTs) published up to February 29, 2026. Eligible studies included people living with HIV receiving ART and evaluated DHIs, such as SMS, mobile apps, phone calls, adherence monitoring devices, multimedia education, or multicomponent interventions. Outcomes included viral suppression, CD4+ cell count, adherence, and retention. Random-effects meta-analyses were conducted using restricted maximum likelihood estimation with Hartung-Knapp-Sidik-Jonkman adjustment. Effect sizes were reported as risk ratios (RRs) or mean differences (MDs) with 95% CIs and 95% PIs. TSA was performed to assess the sufficiency of cumulative evidence. A frequentist network meta-analysis was conducted to compare the relative effectiveness of different DHIs. A total of 64 RCTs involving 22,286 participants were included. Compared with standard of care (SOC), DHIs improved subjective adherence (RR 1.13, 95% CI 1.04-1.23; 95% PI 0.82-1.57) and retention (RR 1.06, 95% CI 1.01-1.12; 95% PI 0.81-1.39). Viral suppression was modestly improved (RR 1.04, 95% CI 1.01-1.07; 95% PI 0.96-1.13), while no significant effect was observed for CD4+ cell count and objective adherence. TSA indicated sufficient evidence for viral suppression and subjective adherence but inconclusive evidence for other outcomes. In the network meta-analysis, SMS, mobile apps, and multicomponent interventions demonstrated statistically significant benefits versus SOC; however, all PIs crossed the null. Although phone calls ranked highest by surface under the cumulative ranking curve (SUCRA), differences between interventions were not robust. In contrast to previous systematic reviews that focused on single intervention types or limited outcomes, this systematic review provides a comprehensive synthesis of multiple outcomes across all types of DHIs, supporting their potential role as nonpharmacological strategies in HIV care. However, the wide 95% PI, together with a high risk of bias, small-study effects, and low to very low certainty of evidence based on GRADE (Grading of Recommendations Assessment, Development, and Evaluation), indicate substantial uncertainty regarding the true effects in future settings. Therefore, these findings should be interpreted with caution. These findings have important practical implications, as they may directly help inform the design of more targeted and context-specific digital interventions and highlight the need for further research to identify optimal implementation strategies in routine HIV care.
Digital surveys are increasingly integrated into clinical and public health research to capture patient-reported outcomes. However, concerns about fraudulent or duplicate responses threaten data integrity. Most of the literature on survey fraud focuses on open-access online recruitment, where bot-generated or anonymous entries are common, but far less is known about fraud patterns in clinic-linked, incentive-based surveys. Evaluations of the real-world implementation of fraud-deterrence strategies remain limited. This study evaluated whether implementing enhanced fraud-deterrence procedures in an incentive-based, clinic-linked, postprocedure survey reduced the prevalence of potentially fraudulent responses. Second, the study evaluated which indicators were most frequently triggered before and after implementation. This evaluation was conducted within the ADOPT (Alternatives to Dental Opioid Prescribing After Tooth Extraction) study. Eligible patients aged 12 to 25 years were recruited through QR-coded, clinic-distributed flyers and invitation cards. Participants completed a screening survey followed by an incentivized postprocedure survey between days 6 and 10 after tooth extraction. A midstudy protocol modification introduced enhanced fraud-deterrence measures in the screening process, including a phone number requirement, prohibition of email invitations, date of birth confirmation, and the use of a participant list with SMS text message invitations. For analysis, survey responses were categorized as "control" (before modification) or "intervention" (after modification). A 6-item scoring system assessing completion time outliers, submission time, repeated screeners, duplicated phone numbers, a blank recruitment source, and illogical response patterns was used to classify responses as potentially fraudulent (≥2 indicators). Sensitivity analyses evaluated thresholds from 1 to 3 indicators. A total of 573 survey responses were included, with 122 in the control cohort and 451 in the intervention cohort. The overall prevalence of potentially fraudulent responses (50/573, 8.7%) was lower than the rates reported in open-access online survey research, and the difference between the control and intervention cohorts was not statistically significant (15/122, 12.3% vs 35/451, 7.8%; P=.12). Fewer surveys in the intervention cohort were flagged for a blank recruitment source (16/451, 3.5% vs 15/122, 12.3%; P<.001) and completion of multiple screeners (29/451, 6.4% vs 16/122, 13.1%; P=.02). The frequency of duplicated phone numbers was higher in the intervention survey (82/451, 18.2% vs 3/122, 2.5%; P<.001), although this difference was not statistically significant when restricted to individuals who provided a phone number (82/451, 18.2% vs 3/46, 6.5%; P=.06). Sensitivity analyses showed consistent patterns across alternative thresholds, and subgroup analyses did not show overall differences in fraud rates based on age, sex, or recruitment location. Enhanced fraud-deterrence procedures did not statistically significantly reduce overall fraud prevalence in a survey setting using clinic-linked, QR-based recruitment with modest incentives. A transparent scoring system provides a replicable approach for assessing survey integrity and may be preferable to reliance on eligibility gating alone.
Health misinformation is a serious and growing concern, especially in the era of mass digitalization. However, the term lacks conceptual clarity, reducing our ability to build a reliable, replicable evidence base about how misinformation works and undermining our attempts to develop effective responses. There is, therefore, a need to examine how the term is used and to develop a coherent definition that better reflects people's information priorities, concerns, and understandings of the concept. This study aimed to surface common themes and debates around the concept of "misinformation" in contemporary English-language discourses about health. Specifically, we aimed to examine how people understand the problem of health misinformation (ie, its causes and consequences), how they perceive the relationship between "misinformation" and other problematic information, and any unresolved conceptual tensions. We conducted a 3-phase hybrid concept analysis following a framework from Schwartz-Barcott and Kim (2000), comprising (1) a theoretical, literature-based phase, (2) a fieldwork/primary interview phase, and (3) an integrative phase combining findings from the first 2 phases. This paper reports methods and findings from the first phase, an inductive, qualitative literature review and analysis. In this phase, we conducted a systematic search of recent literature on health misinformation (published between 2016 and 2022), selected a stratified random sample, and conducted inductive thematic analysis following the methods outlined by Rodgers (2000). A thematic, narrative summary of the findings is presented herein. Authors identified rapid technological change, information predators, and cultural issues (eg, social fragmentation, growing epistemic disagreement, and the loss of information authorities) as antecedents to health misinformation. They characterized health misinformation as a scourge, identifiable by its falseness or deceptiveness, the use of persuasive strategies, unscientific subjectivity, and its capacity to disrupt community consensus, and highlighted a broad range of potential consequences for individuals, communities, and social systems. However, there were also major areas of divergence and tension in this literature base. Specifically, there was disagreement about what kinds of problematic content constituted "misinformation," and how misinformation should be identified or adjudicated, especially in cases of evidentiary uncertainty or legitimate scientific disagreement. To our knowledge, this is the first review to examine both explicit and implicit definitions of misinformation and to center contemporary usages of the concept. The work identifies several key characteristics of health misinformation, as well as areas where further concept development is needed. The review highlights the need for explicit reporting around the operationalization of the term, implementation of standards for communicating evidentiary uncertainty, values-aligned and co-designed health communications, and continued concept development. Findings from this literature analysis will inform further stakeholder and integrative synthesis, with the goal of developing a more usable, inclusive, and responsive definition of health misinformation. RR2-10.12688/hrbopenres.13641.2.
The integration of mobile health apps (MHAs) into nursing practice is essential for improving efficiency, conserving resources, and enhancing patient care. To promote the acceptance of these apps in clinical care, it is important to assess current acceptance levels and identify influencing factors. This study investigated the acceptance of MHAs and related factors among nurses in hospitals affiliated with Kashan University of Medical Sciences, Iran. In this cross-sectional study, 250 nurses were selected via stratified random sampling in Kashan in 2022. Data were collected using the Nurses' Mobile Health Device Acceptance Scale (NMHDA-S) and the Probable Predictors Questionnaire. The data were analyzed using SPSS (version 16; IBM Corp), 1-way ANOVA, independent t test, Pearson correlation coefficients, and multiple linear regression. The findings showed that nurses had a mean acceptance score of 4.207 (SD 0.740) on a Likert scale from 1 to 7 (95% CI 4.115-4.299) regarding the use of MHAs. Multiple linear regression analysis indicated that 4 variables significantly predicted acceptance of MHAs (R²=0.230, F4,245=18.308, P<.001): interest in participating in relevant educational programs (β=0.392, P<.001), encouragement from health care professionals to use MHAs (β=0.133, P=.02), male participant (β=-0.116, P=.04), and duration of daily internet use (β=0.117, P=.04). Although MHA acceptance among clinical nurses in Kashan is moderate, adoption can be enhanced through targeted training, facilitated internet access, and gender-specific incentive policies, particularly for female staff.
Video-based psychotherapy (VBT) became an essential modality during the COVID-19 pandemic, enabling outpatient care despite social distancing measures. Yet little is known about how usage continued and acceptance evolved after the pandemic. Most existing research is cross-sectional, pandemic-focused, and rarely integrates technology acceptance with clinical process quality and therapist heterogeneity to explain sustained postpandemic VBT use. This study examined the postpandemic sustainability of VBT among German outpatient psychotherapists. Guided by UTAUT-T (unified theory of acceptance and use of technology for therapists), we investigated VBT use, acceptance-related predictors, perceived clinical process quality, and therapist acceptance profiles using cross-sectional and longitudinal perspectives. We conducted a repeated cross-sectional, partially longitudinal postal survey among licensed German outpatient psychotherapists during the COVID-19 pandemic (T1: July 2020-January 2021) and postpandemically (T2: March-May 2024). The final T2 sample included 296 psychotherapists; 117 participated in both waves. VBT sustainability was assessed through usage status and intensity. Technology acceptance was measured using the UTAUT-T, and clinical process evaluations were assessed using items based on Grawe's general change mechanisms. Analyses included regression models, longitudinal within-person tests, group comparisons, and person-centered cluster analysis. Postpandemic VBT use was reported by 68.2% (202/296) of psychotherapists. In retrospective comparisons, use remained above prepandemic levels but below the pandemic peak (4.4% [13/296] prepandemic; 81.1% [240/296] during the pandemic; χ22=357.2, P<.001, Kendall W=.64). In the longitudinal subsample, weekly VBT sessions declined from 5.81 during the first pandemic wave to 1.51 postpandemically (χ22=47.9, P<.001, Kendall W=.41). UTAUT-T constructs explained 66.6% of the variance in behavioral intention (R²=.666), with therapy quality expectation as the strongest predictor (β=.513, P<.001). Behavioral intention predicted postpandemic VBT use (odds ratio 6.56, 95% CI 4.31-9.99; P<.001) and usage intensity (β=.414, P<.001). Prior pandemic VBT use and regulatory awareness predicted postpandemic use beyond behavioral intention. Therapists rated VBT as less effective than face-to-face therapy (z=13.14, P<.001, r=0.77), and only 35.8% (106/294) perceived core therapeutic change mechanisms as equally supported. Cluster analysis based on UTAUT-T identified 3 clusters, associated with postpandemic VBT use (χ22=131.1, P<.001, V=.67) and significantly differing in pandemic VBT use, age, therapeutic approach, regulatory awareness and restrictions, and perceived equivalence of Grawe's therapeutic change mechanisms. This study extends previous pandemic-era and cross-sectional VBT research by examining sustained postpandemic use under more voluntary routine-care conditions and by integrating technology acceptance, clinical process quality, and therapist heterogeneity. The findings show that VBT has become a sustained but selectively used component of German outpatient psychotherapy rather than a universal replacement for face-to-face treatment. Postpandemic use is shaped by perceived clinical meaningfulness, prior experience, regulatory awareness, and therapist acceptance profiles. Implementation efforts should move beyond technical access and support clinically differentiated, profile-sensitive VBT use through targeted training, clear regulatory guidance, and shared decision-making with patients.
Multimorbidity involves heterogeneous disease combinations, treatment burden, competing priorities, and complex care pathways. Digital health interventions (DHIs) may support monitoring, self-management, and care coordination, but their effects on health-related outcomes remain uncertain. This systematic review and meta-analysis evaluated the effectiveness of DHIs on physiological, psychological, and functional outcomes in adults with multimorbidity, summarized implementation outcomes, and explored intervention-multimorbidity matching patterns. PubMed, Web of Science Core Collection, Embase, Cochrane Library, CINAHL with Full Text, Scopus, gray literature sources, trial registries, reference lists, and forward citations were searched through April 7, 2026. English-language randomized or cluster-randomized trials enrolled adults with 2 or more chronic conditions and compared a DHI with a comparator lacking the same digital component. Two reviewers independently screened studies, extracted data, and assessed risk of bias using the Cochrane Risk of Bias 2 tool. Random-effects meta-analyses used restricted maximum likelihood estimation, Hartung-Knapp adjustment, and Nagashima-Noma-Furukawa prediction intervals. Certainty of evidence was assessed using the GRADE (Grading of Recommendations, Assessment, Development and Evaluation) approach. In total, 36 reports (37 trials) were included. A total of 7 were rated as high risk of bias and 29 as having some concerns; none was rated as low risk overall. No statistically significant pooled effects were observed for glycated hemoglobin, blood pressure, mortality, hospitalization, or readmission, depression response, health-related quality of life, or pain-related functional impact or disability. For glycated hemoglobin, 6 studies (3992 participants) yielded a mean difference of -0.12 percentage points (95% CI -0.36 to 0.11; 95% prediction interval -0.70 to 0.43). For systolic blood pressure, 6 studies including 5596 participants yielded a mean difference of -3.40 mm Hg (95% CI -8.94 to 2.14; 95% prediction interval -19.15 to 13.03). Depression severity was the only outcome whose pooled 95% CI favored the intervention (8 studies; 1861 participants; standardized mean difference -0.49, 95% CI -0.82 to -0.16), but its prediction interval spanned benefit to harm (-1.53 to 0.51). Certainty was low or very low for all 7 GRADE-assessed outcomes. Implementation findings suggested feasibility, especially with monitoring, coaching, or clinician contact, but reporting was heterogeneous. Current evidence does not support consistent, transferable benefits of DHIs across most outcomes in adults with multimorbidity. Their real-world value may depend less on technology type than on alignment among intervention mechanisms, patient complexity, outcomes, and delivery context. Future DHIs should be adaptive, burden-sensitive, and workflow-integrated, linking digital data to patient priorities, clinician responses, and care coordination.
Men's health urology faces growing challenges driven by workforce shortages, rising disease burden, and persistent disparities in care. Despite an increasing prevalence of conditions like benign prostatic hyperplasia, prostate cancer, and urinary tract infections, more than half of U.S. counties lack a practicing urologist. Stigma, access, and uncertainty lead men to delay care further, resulting in higher morbidity, late-stage diagnoses, and unsustainable costs. To meet these challenges, this paper explores the transformative potential of agentic AI systems to drive Healthcare 5.0 in urology to create a more equitable, efficient, and proactive care system. We examine how AI can advance the quintuple aim of health care: enhancing patient experience, improving population health, reducing costs, increasing provider satisfaction, and promoting health equity. The paper introduces the concept of a suite of specialized AI agents, rooted in both currently in use and developing AI applications, that work collaboratively to support providers, patients, and health care administrators across the continuum of care. These agents not only improve efficiency, streamline workflows, and augment clinical reasoning, but also enable scalable, virtual-first care delivery systems. We articulate our view of the future urology patient journey, illustrating how AI agents can transform each step of the process to provide an improved, seamless experience for patients and providers while maintaining human-centered, personalized care. Finally, we outline critical future directions, such as data interoperability, regulatory frameworks, and inclusive design principles, to ensure that AI technologies are deployed safely, equitably, and in line with ethical regulations. Through the strategic implementation of agentic AI, we view the future of men's health urology as a model for innovation, driving better outcomes for patients and sustainable, meaningful care for providers.
Delirium is a serious condition characterized by an acute change in attention, arousal, and sleep disturbance. It is common in inpatients with Parkinson disease (PD) but is frequently missed or misidentified due to overlapping symptoms, such as cognitive impairment, hallucinations, and sleep disturbances. While clinical tools often measure only a snapshot of delirium, wearable devices could facilitate the identification and ongoing monitoring of delirium, including continuous assessment of activity and sleep patterns, which are frequently disrupted. Establishing feasibility is essential before wearable technologies can be implemented in routine clinical care. This study aimed to determine the feasibility and acceptability of using wearable devices in inpatients with PD, with and without delirium. Participants were recruited from an ongoing prospective cohort study comprising inpatients with PD. Delirium was diagnosed using the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) criteria and assessed daily. Participants were invited to wear Axivity AX6 devices attached to their lumbar region (lower back) and/or wrist for up to 7 days. Feasibility was assessed in terms of recruitment, device placement, nonsecurement, wear time, and compliance. Acceptability, practicality, and clinical constraints were recorded and compared between groups. Participants were predominantly older adults with advanced PD and high levels of frailty and cognitive impairment. The wearable device recruitment rate was 75.4% (46/61), comprising 68 admissions. Delirium was identified in 64.7% (44/68) of admissions. The wrist-worn device showed greater participant acceptability, with 98.5% (67/68) of participants wearing a device on the wrist compared to 38.2% (26/68) of participants wearing a lumbar device (25/68, 36.8%, wore both). Wrist placement was more practical, rated as "somewhat" or "very easy" to secure in 95.5% (64/67) of cases compared to 69.2% (18/26) for lumbar placement (P<.001). Clinical constraints such as injury or pain and low level of arousal were associated with lumbar nonsecurement. These findings indicate that wrist-worn devices are more practical and acceptable in acutely unwell patients. Wear time compliance for both placements tended to be lower in delirium cases but comparable overall (>83% for each location, P>.05). This is the first study to evaluate the feasibility of using wearable devices in inpatients with PD with and without delirium. Wearable devices were feasible, and the wrist-worn devices demonstrated greater participant acceptability and practicality, with fewer clinical constraints. These findings provide important guidance for the design and implementation of future digital health studies in this population and may ultimately support earlier recognition and management of delirium while enabling continuous, objective monitoring of delirium-related changes not captured by standard clinical assessments.
AI-powered chatbots offer new opportunities to enhance patient education; however, their integration may reshape patterns of information interactions and trust relationships among patients, caregivers, and nurses. Evidence remains limited on how these stakeholders perceive the value and risks of AI-powered chatbots, and on their potential effects on nurse-patient trust. This study explores patients', caregivers', and nurses' attitudes toward and experiences with integrating AI-powered chatbots into patient education and identifies perceived benefits, implementation challenges, potential effects on trust, and the supportive conditions required for safe integration. This qualitative study was conducted from April to July 2025. Patients and caregivers were recruited from a tertiary general hospital using maximum variation purposive sampling, while nurses were recruited through snowball sampling from 6 hospitals of varying tiers. Data were collected using a sociodemographic questionnaire and semistructured, in-depth interviews. Interview recordings were transcribed verbatim and analyzed using reflexive thematic analysis, with NVivo used for coding and theme development. Sociodemographic data were analyzed descriptively. A total of 60 participants were included: 29 patients, 17 caregivers, and 14 nurses. Four themes were identified: perceptions and maintenance of nurse-patient trust, conditional acceptance and practical needs, functional optimization and implementation safeguards, and nurses' role pressures and competency restructuring. All 3 stakeholder groups recognized the potential of AI-powered chatbots to address unmet information support needs in patient education but expressed reservations about their accuracy, personalization, and transparency. AI-powered chatbots were not perceived as a direct threat to nurse-patient trust. However, nurses were more sensitive to potential trust tensions, increased explanation burden, and expanded professional responsibilities, highlighting the need for competency restructuring. Stakeholder groups also differed in their perceptions of the conditions required to maintain nurse-patient trust. Limited digital health literacy and the digital divide affecting older patients were major barriers to integrating AI-powered chatbots into patient education. Patients, caregivers, and nurses were generally cautiously open to integrating AI-powered chatbots into patient education, although their assessments of benefits and risks differed by role. AI-powered chatbots may be best positioned as adjunctive information-support tools. Their safe use should be tailored to patient characteristics, information risk, and clinical context, with nurses' professional oversight and coordinated support across governance, technical, and clinical implementation levels.