https://www.jmir.org/issue/feed 2025-01-01T11:30:03-05:00 JMIR Publications editor@jmir.org Open Journal Systems The leading peer-reviewed journal for digital medicine and health and health care in the internet age.  https://www.jmir.org/2026/1/e92863 2026-07-15T17:30:14-04:00 Anna Heilig Tobias Bonten Kasper Recourt M Elske van den Akker-van Marle

Background: The rapid expansion of remote patient monitoring (RPM) interventions highlights the need for their comparison and evaluation. Current evaluation frameworks often fail to capture a multistakeholder perspective. Traditional health technology assessment approaches emphasize health and economic outcomes, providing an incomplete picture of RPM’s broader impact. The Sextuple Aim, encompassing health outcomes, costs, patient and provider experience, equity, and sustainability, offers a more comprehensive approach. Yet, the specific aspects that are important for capturing each domain for diverse stakeholders, and that could form a core outcome set (COS) for RPM evaluation, remain undefined. Objective: This study aimed to identify the most important value aspects for evaluating RPM interventions from the Sextuple Aim domains. Methods: Value aspects relevant to evaluating RPM interventions within the Sextuple Aim framework were identified in a review of the literature, and this initial set was refined through stakeholder meetings. Subsequently, 6 stakeholder groups completed 3 Delphi rounds. Patients and informal caregivers (25‐30 respondents per group) were recruited via a patient federation. Health care providers (20‐30 experts), managers, insurers, and researchers (10‐20 experts per group) were recruited. In each round, participants rated the importance of aspects on a 5-point Likert scale. In round 2, participants ranked aspects within each domain to determine their relative importance; in round 3, they ranked aspects across domains. A multicriteria consensus rule was used to select aspects for the preliminary COS. The COS was finalized during a multistakeholder expert meeting with 7 experts. Results: Of the 171 respondents, 137 (80%) completed all Delphi rounds. The input set, formed through a literature search and meetings, contained 43 value aspects. The total set contained 47 aspects after the first round. After all 3 rating rounds, 33 of 47 (70%) value aspects met the predefined importance threshold. Ranking results from rounds 2 and 3 identified aspects prioritized within and across domains. In total, 26 (55%) aspects met the multicriteria consensus rule and formed the preliminary COS. During the final multistakeholder expert meeting involving subgroup exercises and moderated discussions, value aspects with conceptual overlap were merged, removed, or relocated. The final COS included 17 value aspects, with 1 additional emergent aspect identified during the expert meeting and included as a recommended item, which also resulted in full Sextuple Aim domain coverage. Conclusions: An RPM Sextuple Aim COS comprising 17 value aspects and 1 recommended aspect was developed, providing a multidimensional and multistakeholder evaluation set. This COS provides a standardized, comprehensive basis for evaluating RPM interventions and can be used to support the validation and refinement of existing RPM evaluation tools. Future research may consider operationalizing the COS and its recommended aspect into measurable items and assessing its feasibility and uptake in practice.

2026-07-15T17:30:14-04:00 https://www.jmir.org/2026/1/e92199 2026-07-15T17:30:14-04:00 Yi Wang Andrew N Crenshaw Rodrigo Reis

Using sentiment analysis of 44,216 public health–related posts from the FedNews Reddit forum (2020-2025), we found that fear and anger scores rose 32% and 63%, respectively, in 2025 over 2024. These results underscore the adverse mental health impact of federal workforce changes and funding cuts on public health employees.

2026-07-15T17:30:14-04:00 https://www.jmir.org/2026/1/e91842 2026-07-15T17:00:16-04:00 Rosalyn Schroeder Antonia Biggs Finley Baba Lauren J Ralph Amy Hagstrom Miller Colleen McNicholas Daniel Grossman

Background: Medication abortion accounts for the majority of abortions in the United States, driven in part by the growth in access to telehealth provision of medication abortion. While research indicates high patient satisfaction with telehealth medication abortion care, research on care preferences of patients who use medication abortion remains understudied. Understanding these preferences is essential to informing evidence-based policies that enable people to access person-centered abortion services that meet their values, preferences, and needs. Objective: The aim of this study is to compare the abortion care features rated as most important among patients receiving medication abortion via telehealth vs in-person care and to assess participants’ access to technologies required for telehealth care. We hypothesized that participants receiving telehealth care would place greater importance on features supporting limited in-person clinic interaction. Methods: From May 2021 to March 2023, we surveyed participants (≤70 d of gestation, English or Spanish speakers, aged ≥15 y) obtaining medication abortion at 4 organizations providing care in 6 US states. Participants rated the importance of 12 abortion care features (ie, getting care at home, convenience, cost, safety, effectiveness, and privacy) and described access to technologies required for telehealth. We used bivariable logistic and ordinal regressions with robust SEs to assess whether features rated as “extremely important” differed by the medication abortion model received (telehealth vs in-person). Results: Among 1017 patients approached, 876 were eligible, 583 participants enrolled, 487 initiated a survey, 477 (242 telehealth and 235 in-person) completed survey questions regarding access to technologies for telehealth, and 397 completed questions about abortion care features. Across groups, the features most often rated as “extremely important” included effectiveness (340/397, 85.6%), safety (326/397, 82.1%), timeliness (307/397, 77.3%), and privacy (165/211, 78.2%), with no significant differences (>.05) between groups. Compared to in-person participants, telehealth participants were more likely to report getting care at home (65.9% vs 43.5%; odds ratio [OR] 2.48, 95% CI 1.57‐3.89; <.001) and having a medication abortion (62.1% vs 51.1%; OR 1.58, 95% CI 1.11‐2.23; =.01) as extremely important. They were less likely to report having an in-person meeting with a clinician (16.1% vs 45.7%; OR 0.23, 95% CI 0.19‐0.26; <.001) and ultrasonography (15.2% vs 38.2%; OR 0.28, 95% CI 0.20‐0.40; <.001) as extremely important. Abortion care features rated as extremely important were sometimes discordant with the care received. Almost all participants had access to the technologies required for telehealth. Conclusions: While telehealth abortion services offer many features that people find important, the availability of both telehealth and in-person abortion care remains critical to ensuring that care aligns with patient preferences. In addition to efforts focused on expanding access to telehealth medication abortion services, advocates and policymakers should continue their work to ensure access to in-person care for those who need or prefer this model.

2026-07-15T17:00:16-04:00 https://www.jmir.org/2026/1/e95596 2026-07-15T16:45:11-04:00 Haoran Chen Shenglan Xiao Tong Wan Gui Li Yanhong Peng Zhimin Wang

Background: Hypertension is a major global health challenge, and effective health education is crucial for improving patients’ self-management. Traditional health education approaches are often limited by insufficient personalization, accessibility, and scalability. Artificial intelligence (AI), including natural language processing, machine learning, and large language models (LLMs), offers promising solutions to address these limitations. However, evidence regarding AI applications in hypertension health education has not been comprehensively synthesized. Objective: This scoping review aimed to summarize the current evidence on AI applications in hypertension health education, and identify research gaps to inform future research and practice. Methods: This review followed the Joanna Briggs Institute methodology and PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines. Six databases (PubMed, Embase, Web of Science, Cochrane Library, CINAHL, and Scopus) were searched from January 2015 to June 2026. Eligibility criteria were developed using the participant-concept-context framework. Two reviewers independently conducted study screening and data extraction. Study designs were classified using the Mixed Methods Appraisal Tool framework. Consistent with scoping review methodology, no formal quality assessment was performed. Findings were synthesized narratively and presented using evidence gap maps, tables, and figures. Results: A total of 24 studies from 11 countries were included, comprising 6 randomized controlled trials, 4 nonrandomized trials, 11 quantitative descriptive studies, and 3 mixed methods studies. Most studies were published between 2024 and 2026. In total, 3 AI application scenarios were identified: rule-based health education, data-driven adaptive health education, and generative AI–driven health education. Natural language processing was the most widely applied technology, and LLM-based applications increased rapidly after 2023. However, generative AI studies were predominantly proof-of-concept evaluations and lacked randomized clinical validation. Health education was rarely implemented as a standalone intervention and was typically embedded within multifunctional AI platforms. Outcomes were categorized using the Digital Health Scorecard Framework across 4 domains: technology, clinical, usability, and cost. Technical accuracy and blood pressure outcomes were the most frequently reported measures, whereas no study evaluated economic outcomes. Conclusions: This first scoping review of AI applications in hypertension health education identified a mismatch between rapid advances in generative AI and the limited availability of rigorous clinical evidence. Three major research gaps were identified: (1) the lack of standardized core outcome sets covering technical, behavioral, clinical, and implementation domains; (2) limited development of hybrid architectures integrating LLM with structured medical knowledge bases; and (3) the absence of evaluation frameworks that satisfy both regulatory and implementation requirements. AI appears most suitable as a complement to, rather than a replacement for, clinician-delivered education. Future research should prioritize rigorous clinical validation, economic evaluation, multicultural adaptation, and health literacy equity to ensure that AI-driven health education reduces rather than exacerbates disparities in hypertension control. Trial Registration: OSF Registries osf.io/4wv3f; https://osf.io/4wv3f/overview

2026-07-15T16:45:11-04:00 https://www.jmir.org/2026/1/e92347 2026-07-15T16:45:11-04:00 Iony D Ezawa Francisco N Ramos Steven D Hollon Gloria T Han Noah Robinson

Background: Depression is prevalent and debilitating. Although interventions exist, they are rarely delivered in accessible, scalable ways that retain their effectiveness. Cognitive behavioral immersion (CBI) is a coach-led cognitive behavioral skills program delivered in social virtual worlds that offers a potential solution. Objective: This parallel-group, web-based randomized controlled superiority trial compared CBI accessed via virtual reality headsets (CBI-VR) or flat-screen devices (CBI-FS) to a delayed access control. Methods: Inclusion criteria included a clinical level of depression symptoms, age ≥18 years, able and willing to give informed consent, access to a computer with an internet connection, and ability to speak and read English. Eligible participants were randomized using a random number generation script in a 1:1:1 ratio to conditions. CBI consisted of 8 weekly 1-hour groups led by coaches who taught cognitive behavioral skills. The intervention lasted 8 weeks; follow-up lasted 6 months. The primary outcome was depression symptoms; secondary outcomes were anxiety symptoms and quality of life. Recruitment and study procedures were conducted online. Outcomes were assessed through electronic self-report questionnaires. The study was unblinded. Hierarchical linear modeling was used to examine differences in rates of change among conditions. We explored the sense of presence as a potential mediator of intervention response. Results: Participants were recruited from February 2024 to January 2025; n=102 were randomized to each condition. Participants randomized to CBI-VR and CBI-FS attended an average of 5 intervention sessions. Primary analyses included all participants in the intent-to-treat sample that completed ≥2 outcome surveys to estimate within-person change (CBI-VR: n=98; CBI-FS: n=86; control: n=102). CBI-VR showed faster reductions in depressive and anxiety symptoms than either CBI-FS (depression: =.21; 95% CI 0.02-0.40; =.03 and anxiety: =.20, 95% CI 0.03-0.38; =.02) or the control (depression: =.31, 95% CI 0.13-0.48; <.001 and anxiety: =.18, 95% CI 0.01-0.34; =.03) across the 8-week intervention, with improvements largely maintained over the 6-month follow-up. CBI-VR also showed greater improvements in general quality of life (=−1.02; 95% CI −1.63 to −0.40; =.001) and psychological well-being (=−1.01, 95% CI −1.44 to −0.59; <.001) than the control from pre- to postintervention. The sense of physical presence in the environment was associated with CBI-VR’s effects on depression symptoms (=−0.85, 95% CI −1.71 to −0.15). No adverse effects occurred in any group. Conclusions: This study evaluated the efficacy of an innovative coach-led cognitive behavioral skills group delivered via VR. To our knowledge, our trial is the first to demonstrate that CBI delivered via VR is effective. These findings extend prior work on digital cognitive behavioral therapy by supporting CBI-VR as an effective and viable intervention package for depression and anxiety symptoms. These findings may help inform future research on suitable technology that can help bridge mental health care gaps. Trial Registration: ClinicalTrials.gov NCT06418997; https://clinicaltrials.gov/study/NCT06418997 International Registered Report Identifier (IRRID): RR2-10.2196/65970

2026-07-15T16:45:11-04:00 https://www.jmir.org/2026/1/e94777 2026-07-15T16:30:14-04:00 Rebecca Rhodin Rakel Eklund Anneli Silvén Hagström Rolf Gjestad Atle Dyregrov Josefin Sveen

Background: Adolescents who experience the loss of a family member are at increased risk of adverse mental health outcomes, yet many face barriers or may be reluctant to access in-person or group-based support. mHealth (mobile health) interventions can help address these barriers by offering flexible, accessible, and low-threshold support. Objective: This study evaluated the short- and long-term mental health effects of Alba – Youth in Grief, a preventive self-management mobile app for bereaved adolescents. The primary outcome was symptoms of prolonged grief, while secondary outcomes included grief reactions, personal growth, and symptoms of posttraumatic stress and depression. User-reported helpfulness and negative experiences were also examined. Methods: In an unblinded randomized controlled trial (ClinicalTrials.gov NCT06093113), 126 adolescents aged 12‐19 years who had lost a parent or sibling were allocated to either the unguided Alba app (n=61) or an active control condition receiving unguided web-based psychoeducation (n=65). Online self-assessments were conducted at baseline and at 2, 6, and 12 months. Participants generally demonstrated high levels of distress at baseline, with 40% (50/126) reporting symptoms indicative of probable prolonged grief disorder according to () diagnostic scoring rules. Mental health outcomes were analyzed using linear mixed models to examine changes over time between groups, while user experiences were examined using descriptive statistics and summative content analysis. Results: Intention-to-treat analyses showed moderate reductions in prolonged grief symptoms at 12 months among adolescents randomized to Alba compared with the control group, with no significant effects at the 2- and 6-month follow-ups. The app group also demonstrated greater reductions in grief reactions, posttraumatic stress symptoms, and depressive symptoms compared with controls, with the strongest effects observed at long-term follow-up. No effect on personal growth was demonstrated. Most participants reported the app as helpful, while a minority disclosed negative experiences such as sadness. Conclusions: Overall, the findings indicate that Alba may be beneficial in reducing mental health symptoms among bereaved adolescents and highlight its potential as a safe, acceptable, and scalable mHealth intervention. Trial Registration: ClinicalTrials.gov NCT06093113; https://clinicaltrials.gov/study/NCT06093113

2026-07-15T16:30:14-04:00 https://www.jmir.org/2026/1/e86045 2026-07-15T16:00:21-04:00 Nina Vitlov Nensi Bralić Tina Poklepović Peričić Daniel Garcia-Costa Emilia López-Iñesta Elena Álvarez-García Francisco Grimaldo Ana Marušić

Background: Summary of findings (SoF) tables are widely used in systematic reviews and clinical practice guidelines to present evidence about health care interventions in a concise and transparent format. Although developed to improve accessibility and interpretation of evidence, previous studies have shown that users often experience difficulties understanding statistical information, certainty ratings, and the relationships between outcomes and treatment effects. Limited research has explored how different groups of users cognitively interact with SoF tables while solving evidence interpretation tasks, particularly when table complexity increases. Objective: This study aimed to investigate how different types of users—Grading of Recommendations Assessment, Development, and Evaluation (GRADE) or Cochrane experts, practicing clinicians, and health care users—interact with SoF tables of varying complexity while answering intervention-related questions, and to examine differences in search behavior, navigation patterns, and task performance. Methods: We used the Read&Learn tool (ERI-Lectura and LAIA-UV) in an online, single-session study to evaluate participants’ interactions with SoF tables. Participants (n=120; 40 per group) accessed preselected SoF tables via a secure link and unique login. Participants completed tasks involving 4 SoF tables with increasing complexity. Specific table cells were blurred and had to be clicked to reveal information. Outcomes included the number of correct answers, total time spent on tasks, number of table cells visited, number of target and nontarget cells, and question-reading behavior. Results: Simpler SoF tables with a small number of outcomes and single target cells were correctly interpreted by most participants, regardless of their expertise. As table complexity and task demand increased, all participant groups demonstrated reduced performance and less efficient navigation patterns. Experts generally performed better than clinicians and health care users, particularly by spending less time reading nontarget cells and visiting fewer irrelevant table elements. Nevertheless, even experts experienced difficulties with tasks requiring synthesis and interpretation across multiple table cells. Questions requiring comparison and integration of information across outcomes resulted in the highest rates of incorrect responses in all groups. Heatmaps of the number of clicks and time spent on tables demonstrated that experts used more targeted search strategies, whereas clinicians and health care users explored a larger number of nontarget cells and spent more time navigating the tables. The most complex SoF tables produced the highest cognitive demands for all groups, suggesting that increasing element interactivity and information density substantially affect usability. Conclusions: SoF tables remain cognitively demanding even for experienced users of evidence synthesis. Increasing table complexity appears to reduce users’ ability to identify, interpret, and synthesize relevant information. These findings suggest that current SoF formats may impose substantial intrinsic and extraneous cognitive load, particularly for nonexpert audiences. Future development of SoF tables should prioritize clearer presentation of clearer outcomes, inclusion of absolute alongside relative effects, and interactive or user-centered formats that support evidence navigation and interpretation.

2026-07-15T16:00:21-04:00 https://www.jmir.org/2026/1/e93578 2026-07-15T15:30:04-04:00 Xi Wang Tian Shen Xi Chen Kejia He Yuxiang Chris Zhao

Background: Online health information–seeking (OHIS) behavior shapes health self-management, and eHealth literacy—the ability to seek, appraise, and apply electronic health information—is regarded as its key driver. Previous reviews aggregated heterogeneous outcomes, focused on measurement properties, or examined single clinical populations, without isolating the eHealth literacy–OHIS link. Objective: This study quantified the strength and heterogeneity of the eHealth literacy–OHIS association and identified its boundary conditions across generation, morbidity status, and information source credibility. Methods: Following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses), we searched PubMed, Embase, Web of Science Core Collection, PsycINFO, Psychology and Behavioral Sciences Collection, and Library, Information Science, and Technology Abstracts (LISTA) up to March 15, 2026 (PROSPERO [International Prospective Register of Systematic Reviews] CRD420251088300). Eligible studies enrolled participants, measured eHealth literacy with validated instruments, and assessed OHIS. Risk of bias used the modified Newcastle-Ottawa Scale. Correlations were Fisher z–transformed and pooled under a random-effects model with the Hartung-Knapp-Sidik-Jonkman correction; subgroups were age cohort, morbidity status, and source type. Heterogeneity was quantified with ² and τ²; a univariate meta-regression examined temporal trends, and certainty of evidence was rated using GRADE (Grading of Recommendations, Assessment, Development, and Evaluation). Results: Of 9249 nonduplicate records, 32 studies entered the qualitative synthesis, and 19 (20 effect sizes) the meta-analysis. The grand mean correlation was 0.27 (95% CI 0.15-0.38; <.001) but is of limited interpretive value given extreme heterogeneity (²=99%; τ²=0.064; 95% prediction interval −0.26 to 0.67). Correlations were stronger in non–Gen Z (k=12; r=0.39; 95% CI 0.27-0.50; <.001) than in Gen Z (k=8; r=0.07; 95% CI −0.06 to 0.20; =.23), in patients (k=3; r=0.58; 95% CI 0.01-0.86; =.049) than in nonpatients (k=17; r=0.22; 95% CI 0.11-0.32; <.001), and in professional (k=5; r=0.41; 95% CI 0.11-0.64; =.02) than in nonprofessional (k=14; r=0.21; 95% CI 0.06-0.35; =.01) sources. Meta-regression on collection year showed no significant temporal change (b=−0.005 per year; =.55), and neither the Egger test (=.60) nor trim-and-fill indicated small-study effects. Conclusions: The eHealth literacy–OHIS association is best understood through its boundary conditions, not the overall estimate. The association was robust in non–Gen Z and professional-source contexts but near-null in Gen Z, showing that the eHealth literacy scale’s behavioral predictive validity is cohort- and platform-dependent. Interventions for Gen Z and nonpatient populations should pair literacy training with motivational cues and professionally curated information environments. GRADE certainty was very low, underscoring the need for longitudinal, performance-based research. Trial Registration: PROSPERO CRD420251088300; https://www.crd.york.ac.uk/PROSPERO/view/CRD420251088300

2026-07-15T15:30:04-04:00 https://www.jmir.org/2026/1/e106387 2026-07-15T15:30:03-04:00 Noemi Buo Eleonora Rosi Silvia Busti Ceccarelli Mariarosa Ferrario Valerio Maiorca Erika Morandi Nicole Viganó Ivan Limosani Laura Falcone Massimo Molteni Paola Colombo 2026-07-15T15:30:03-04:00 https://www.jmir.org/2026/1/e89190 2026-07-15T14:45:11-04:00 Yuhan Zhao Rongrong Guo Yiqun Miao Yuan Luo Huiying Wang Ying Wu

Background: Large language models (LLMs) are increasingly embedded in conversational agents for cardiometabolic care. These systems could support self-management, but their behavior change content, delivery mechanisms, and implementation transparency are poorly understood. Objective: This scoping review mapped behavior change techniques (BCTs) used in LLM-driven conversational agents for cardiometabolic prevention and management, described how these techniques are delivered across static, rule-based, and generative mechanisms, examined LLM design, personalization, and safety reporting, and summarized user experience and behavioral or clinical outcomes. Methods: We searched PubMed, Web of Science, Embase, CINAHL, APA PsycInfo, IEEE Xplore, ACM Digital Library, arXiv, ClinicalTrials.gov, and the WHO International Clinical Trials Registry Platform for records published from January 1, 2020, to November 30, 2025. The final search was run on March 25, 2026, using this publication-date limit. Eligible studies reported a patient-facing text- or voice-based cardiometabolic conversational agent using an LLM or other transformer-based generative model. Two reviewers independently screened records and extracted data. BCTs were coded using the Behavior Change Technique Taxonomy v1; selected self-management BCTs were classified as static, rule-based or templated, or generative or context-aware. Empirical human-participant– or evaluator-based studies were appraised with the Mixed Methods Appraisal Tool, and a study-specific checklist assessed LLM implementation reporting transparency. Results: Thirty-eight studies were included; 19 involved empirical human-participant– or evaluator-based assessments, whereas 19 were technical and system-level evaluations, including framework-development, simulated-output, and proof-of-concept studies. Studies were concentrated in 2024‐2025. Instruction on how to perform behavior was identified in 30 of 38 (79%) studies, information about health consequences in 27 of 38 (71%) studies, and feedback and monitoring techniques in 19 of 38 (50%) studies. Most agents were positioned as educators or coaches targeting type 2 diabetes, obesity, or related cardiometabolic risk, and GPT-family models embedded in hybrid architectures with retrieval-augmented generation or rule-based components predominated. Generative outputs were used mainly for tailored explanations, risk information, and socioemotional responses, whereas self-monitoring, reminders, and structured interactions were more often rule-based or mixed-mode. Only 13 of 38 (34%) studies fully reported prompts or system messages, and 16 of 38 (42%) studies fully reported safety or oversight mechanisms. User evaluations reported good usability and perceived helpfulness, but behavioral or physiological outcomes were sparse and usually limited to pilot, short-term, or single-case designs. Conclusions: LLM-driven conversational agents for cardiometabolic care are proliferating but remain early-stage and methodologically heterogeneous. Current systems primarily use LLMs as educational and explanatory layers with “synthetic empathy” over rule-based data capture and safety functions, while behavior change content remains dominated by information provision and simple feedback. More rigorous comparative studies with longer follow-up are needed before firm conclusions can be drawn about sustained behavioral or clinical benefit. Trial Registration: OSF Registries osf.io/jw8vz; https://osf.io/jw8vz

2026-07-15T14:45:11-04:00