Due to necessary scheduled maintenance, the JMIR Publications website will be unavailable from Wednesday, July 01, 2020 at 8:00 PM to 10:00 PM EST. We apologize in advance for any inconvenience this may cause you.
Who will be affected?
Readers: No access to all 28 journals. We recommend accessing our articles via PubMed Central
Authors: No access to the submission form or your user account.
Reviewers: No access to your user account. Please download manuscripts you are reviewing for offline reading before Wednesday, July 01, 2020 at 7:00 PM.
Editors: No access to your user account to assign reviewers or make decisions.
Copyeditors: No access to user account. Please download manuscripts you are copyediting before Wednesday, July 01, 2020 at 7:00 PM.
Background: Since the COVID 19 pandemic, health care and health information seeking have become increasingly digitally mediated. It remains unclear whether eHealth literacy is consistently associated with health behaviors across different behavioral functions and social contexts in the post COVID 19 era. Objective: To synthesize post COVID 19 evidence on the association between eHealth literacy and health behaviors and to examine whether this association varies by health behavior domain, country income context, and population age structure. Methods: We conducted a PRISMA 2020 compliant systematic review and meta analysis registered in PROSPERO (CRD4201009048). PubMed, Embase, and the Cochrane Library were searched from inception to January 28, 2026. Observational studies were eligible if they assessed eHealth literacy using a validated instrument with an explicit score, measured health behavior outcomes that could be classified as health decision making, health promotion, or health management, and collected data in 2020 or later or explicitly reported the timing of data collection. Odds ratios and correlation coefficients were synthesized separately using random effects models with Hartung Knapp adjustment. Funnel plots and trim and fill were used to assess small study effects. Subgroup differences were tested using between group heterogeneity statistics. Studies with non comparable outcomes were summarized narratively. Results: Twenty two studies met the inclusion criteria, including 15 studies contributing quantitative effect estimates and 7 studies summarized narratively. Overall associations were directionally positive, with substantial heterogeneity and sensitivity to small study effects. Behavioral domain was the most consistent source of between study variation across effect size frameworks. Income context moderated associations in the correlation based synthesis, whereas age structure did not show significant moderation. Narrative evidence was most consistent for health decision making outcomes, more mixed for health promotion outcomes, and more variable and generally weaker for health management outcomes. Conclusions: In post COVID 19 studies, eHealth literacy is generally associated with health behaviors, but the strength and consistency of this relationship vary across behavioral domains and settings. Future longitudinal and intervention research using more comparable behavior measures is needed to clarify directionality and to inform context tailored strategies for improving eHealth literacy and health behavior.
We have made changes to our privacy policy. The document has been restructured and updated for clarity and completeness. It includes more information about our cookies and cookie manager. It has also been updated for the new Karma features. If you have any questions or concerns please reach out to privacy@jmir.org
The Journal of Medical Internet Research (JMIR) is the pioneer open access eHealth journal, and is the flagship journal of JMIR Publications. It is a leading health services and digital health journal globally in terms of quality/visibility (Journal Impact Factor 6.0, Journal Citation Reports 2025 from Clarivate), ranking Q1 in both the 'Medical Informatics' and 'Health Care Sciences & Services' categories, and is also the largest journal in the field. The journal is ranked #1 on Google Scholar in the 'Medical Informatics' discipline. The journal focuses on emerging technologies, medical devices, apps, engineering, telehealth and informatics applications for patient education, prevention, population health and clinical care.
As an open access journal, we are read by clinicians, allied health professionals, informal caregivers, and patients alike, and have (as with all JMIR journals) a focus on readable and applied science reporting the design and evaluation of health innovations and emerging technologies. We publish original research, viewpoints, and reviews (both literature reviews and medical device/technology/app reviews). Peer-review reports are portable across JMIR journals and papers can be transferred, so authors save time by not having to resubmit a paper to a different journal but can simply transfer it between journals.
We are also a leader in participatory and open science approaches, and offer the option to publish new submissions immediately as preprints, which receive DOIs for immediate citation (eg, in grant proposals), and for open peer-review purposes. We also invite patients to participate (eg, as peer-reviewers) and have patient representatives on editorial boards.
As all JMIR journals, the journal encourages Open Science principles and strongly encourages publication of a protocol before data collection. Authors who have published a protocol in JMIR Research Protocols get a discount of 20% on the Article Processing Fee when publishing a subsequent results paper in any JMIR journal.
Background: In recent years, researchers have investigated machine learning (ML)–based approaches for the detection of left ventricular hypertrophy (LVH). However, the accuracy of ML in detecting LVH varies across different modeling variables and models. Systematic evidence is lacking in understanding how different ML approaches affect LVH detection accuracy.
Objective: The aim of this study is to systematically assess the diagnostic accuracy of these ML approaches to inform the development of artificial intelligence tools.
Methods: PubMed, Embase, Cochrane Library, and Web of Science were comprehensively searched up to November 12, 2025. The Prediction Model Risk of Bias Assessment Tool was used to evaluate the risk of bias. Subgroup analyses were performed based on ML model types and modeling variables (electrocardiogram [ECG], clinical features, and echocardiography). Only diagnostic 2×2 tables from validation sets were pooled for meta-analysis, with all statistical analyses performed using Stata.
Results: A total of 25 studies were included in the analysis. The performance of ML models varied with input data types and algorithms. A meta-analysis showed that ECG-based models, in comparison, exhibited a sensitivity of 0.76 (95% CI 0.66‐0.84) and a specificity of 0.84 (95% CI 0.78‐0.89). Echocardiography-based models had a sensitivity ranging from 0.71 to 0.94 and a specificity ranging from 0.67 to 0.96. The models based on clinical features had a sensitivity of 0.78 (95% CI 0.69‐0.85) and a specificity of 0.71 (95% CI 0.65‐0.76). A subgroup analysis of the ECG-based models revealed that the deep learning model produced a sensitivity of 0.71 (95% CI 0.60‐0.80) and a specificity of 0.79 (95% CI 0.65‐0.88).
Conclusions: ML demonstrates reasonably high accuracy in detecting LVH. However, these conclusions are derived from limited evidence. Meanwhile, the extreme heterogeneity reported in the meta-analysis requires more critical interpretation. Current conclusions regarding model accuracy should be interpreted with caution. Therefore, future research should focus on constructing high-performance ML models based on imaging data for LVH diagnosis.
Background: Insufficient physical activity among adolescents is a major global public health concern. Digital health interventions (DHIs) have gained increasing attention as a promising approach to promoting physical activity in adolescents. However, existing systematic reviews predominantly focus on single-intervention formats or specific study designs, while reviews that integrate multiple DHIs and diverse study designs remain scarce.
Objective: This systematic review aims to synthesize evidence from diverse DHIs and multiple study designs to assess their effectiveness in promoting physical activity among adolescents.
Methods: The review protocol was registered in PROSPERO (International Prospective Register of Systematic Reviews; CRD420251117923). This systematic review searched literature published between January 1, 2014, and June 30, 2025, across Web of Science, PubMed, EBSCO, Scopus, Embase, the Cochrane Library, ProQuest, and Google Scholar. The final search was completed on August 3, 2025. Using the PICOS (population, intervention, comparator, outcomes, and study design) framework, the review included adolescents aged 10-19 years and focused on evidence-based research promoting physical activity through DHIs. The review was limited to peer-reviewed English-language literature and excluded studies solely focused on measurement tools, those not evaluating intervention effectiveness, or those not involving adolescents. Two reviewers independently screened studies and extracted data. Research quality was assessed using the Joanna Briggs Institute tool. Findings were synthesized through narrative synthesis and qualitative content analysis.
Results: A total of 24 studies were included, involving approximately 12,183 adolescents. Study designs comprised 10 randomized controlled trials, 4 quasi-experimental studies, 3 quantitative research studies, 3 cross-sectional studies, and 4 mixed methods studies. Overall, 7 (29%) studies were of high quality, 16 (67%) were of moderate quality, and 1 (4%) was of low quality. Study populations included general adolescents as well as subgroups with specific health risks: insufficient physical activity (1/24, 4%), obesity or overweight (4/24, 17%), attention-deficit/hyperactivity disorder (1/24, 4%), cancer survivors (1/24, 4%), and at-risk youth (1/24, 4%). DHIs were categorized into 3 types: single-driver interventions (14/24, 58%), multimodal integrated interventions (7/24, 29%), and interaction-enhanced interventions (3/24, 13%). Most studies reported positive outcomes, including direct effectiveness (15/24, 63%), indirect effectiveness (8/24, 33%), and unclear effectiveness (1/24, 4%).
Conclusions: This systematic review synthesizes evidence from diverse research designs and multiple types of DHIs, offering a more comprehensive perspective than previous reviews focused on single designs or technological formats. The results indicate that DHIs generally enhance adolescent physical activity levels, although their effectiveness varies considerably across intervention types and study designs. The review fills key research gaps and highlights the critical role of intervention adaptability and implementation context. It also addresses practical concerns, including adolescents with special health conditions, digital health inequalities, and technology dependency. Despite limitations related to methodological quality and insufficient follow-up, this review provides important evidence to inform practical application, policy development, and the equitable promotion of DHIs to enhance adolescent physical activity. Against the backdrop of rising global adolescent physical inactivity and widening health disparities, it also outlines directions for future high-quality research.
Trial Registration: PROSPERO CRD420251117923; https://www.crd.york.ac.uk/PROSPERO/view/CRD420251117923
With the growing use of technology in qualitative data collection and analysis, there is an opportunity to gather rich and varied perspectives to improve health and well-being. However, large-scale qualitative datasets can be difficult to manage using traditional qualitative methods, and there are few examples of the application of large-scale qualitative analysis. In the context of digital health, large qualitative datasets are increasingly made up of short text segments, which need to be analyzed differently from lengthy transcripts from interviews or focus groups. Therefore, this tutorial describes the use of traditional qualitative methods to analyze a large corpus of qualitative text data. We use examples from a nationwide SMS text messaging poll of youth to highlight the opportunities to use this team-based analysis approach, which has been accessible and meaningful to youth researchers and novice qualitative researchers. These large-scale qualitative strategies may benefit novice researchers analyzing large volumes of qualitative data and short text segments, including SMS text messaging, social media posts, medical notes, and open-ended survey questions, among others.
Background: Approximately 14% of the adult population has tinnitus, and current treatments are often costly and time-consuming. Telerehabilitation might reduce treatment costs without compromising effectiveness.
Objective: Telerehabilitation is a quickly evolving research topic. Therefore, this systematic review update aims to give an overview of the research concerning the effectiveness of telerehabilitation interventions for self-management of tinnitus published between 2022 and 2025.
Methods: This systematic review adheres to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020) guidelines. PubMed, ScienceDirect, Scopus, Web of Science, and Cochrane Library were consulted for eligible studies concerning a study intervention of any possible form of self-management or telerehabilitation for adult patients with subjective tinnitus as a primary complaint. The risk of bias (RoB) and certainty of all included studies were assessed respectively by the Cochrane RoB2-tool and GRADE (Grading of Recommendations, Assessment, Development, and Evaluation) framework.
Results: In total, 24 papers were included, of which 6 studied multiple telerehabilitation forms. Internet-based cognitive behavioral therapy with guidance by a psychologist or audiologist was examined in 5 studies (n=619), self-help manuals in 1 study (n=10), technological self-help devices in 3 studies (n=286), smartphone apps in 13 studies (n=23,788), and other internet-based interventions in 5 studies (n=442). These rehabilitation categories were proven to be effective in decreasing tinnitus severity and relieving tinnitus distress as measured by tinnitus questionnaires.
Conclusions: The strength of this review is the gathering of recent studies on the very evolving topic of telerehabilitation for tinnitus. An important limitation of all included studies is that they raised some to great concerns of RoB. As a result, it is necessary to acknowledge that the overall certainty of the evidence ranged from low to moderate certainty. In addition, some crucial confounding parameters, such as the presence of hearing loss, hyperacusis, anxiety, depression, or sleeping problems, were not taken into consideration by all studies. This review gives an indication of the use of different telerehabilitation and self-management interventions for real-world clinical use, stating not only their possibilities but also their limitations. Overall, telerehabilitation was found to be effective in reducing tinnitus severity and distress. It forms a possible tool to improve the self-management capacities of the patient and the accessibility of tinnitus care as a replacement or an addition to in-person care. Nevertheless, barriers such as a lack of time, engagement, motivation, and openness of the patient, causing high dropout, should be taken into consideration. This review accentuated the shift from internet-based cognitive behavioral therapy to the growing interest in the use of smartphone apps, increasing the accessibility of the treatments even more.
Trial Registration: PROSPERO CRD 42021285450; https://www.crd.york.ac.uk/PROSPERO/view/CRD42021285450
Background: Adolescents perceive both immediate and long-term benefits and losses related to internet gaming, affecting their risk of internet gaming disorder (IGD). These perceptions could also be shaped and reinforced by IGD, indicating potential bidirectionality.
Objective: This study aimed to investigate the bidirectional relationships between perceived immediate and long-term benefits in 3 domains (mental health, social relationships, and personal achievement) and IGD, and between perceived immediate and long-term losses in 6 domains (mental health, sleep quality, academic performance, family relationships, social relationships, and personal achievement) and IGD.
Methods: A 12-month 2-wave prospective longitudinal study was conducted among junior middle school students who had played internet games in the past 12 months in Guangzhou and Chengdu, China, with a baseline survey (T1, December 2018) and the other identical follow-up survey conducted 1 year later (T2, December 2019). The participating schools were conveniently selected; all Grade 7 and 8 students were invited to self-administer the questionnaires in a classroom setting without the presence of the schoolteachers. The final sample size was 1173 students (mean age 12.5, SD 0.6 y; male: 693/1173, 59.1%). IGD was assessed by using the 9-item IGD checklist.
Results: Cross-lagged panel analysis (adjusting for background factors) showed (1) stronger perceived immediate benefits of mental health (=.08, 95% CI 0.01-0.15) and personal achievement (=.10, 95% CI 0.01-0.20) at T1 significantly predicted more IGD symptoms at T2; (2) more IGD symptoms at T1 significantly predicted stronger perceived immediate and long-term benefits of social relationships (immediate: =.09, 95% CI 0.03-0.15; long-term: =.11, 95% CI:0.05-0.17) and personal achievement (immediate: =.12, 95% CI 0.06-0.18; long-term: =.10, 95% CI 0.04-0.16) at T2; (3) more IGD symptoms at T1 significantly predicted stronger perceived immediate and future losses in mental health (immediate: =.09, 95% CI 0.03-0.15; long-term: =.08, 95% CI 0.02-0.14), sleep quality (immediate: =.10, 95% CI 0.04-0.16; long-term: =.13, 95% CI 0.07-0.19), academic performance (immediate: =.09, 95% CI 0.04-0.15; long-term: =.07, 95% CI 0.01-0.13), and family relationships (immediate: =.11, 95% CI 0.05-0.17; long-term: =.10, 95% CI 0.04-0.16) at T2, as well as perceived long-term losses in social relationships at T2 (=.08, 95% CI 0.02-0.14).
Conclusions: This study was innovative in integrating time perspective into both perceived benefits and losses of internet gaming, a cognitive dimension previously overlooked in literature. The current findings advance the field by revealing the unidimensional predictive effects of IGD on perceived immediate and long-term benefits and losses, with 2 exceptions of perceived immediate and long-term benefits of mental health and personal achievement conversely predicting IGD. These results contribute to the development of effective interventions: the cognitive components should go beyond the general pros and cons of gaming and target the potential temporal bias gamers hold.
Background: Artificial intelligence (AI)–enabled wearable devices are rapidly emerging in rehabilitation and motor function assessment for patients with Parkinson disease (PD). However, evidence remains fragmented, integration into nursing practice is limited, and comprehensive synthesis is lacking.
Objective: This study aimed to summarize studies on AI-enabled wearable devices for PD rehabilitation and motor function assessment, describing device types, monitored indicators, algorithms, and application characteristics, and identifying research gaps and barriers to clinical translation.
Methods: Guided by the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) framework, 9 databases (China National Knowledge Infrastructure, Wanfang Data, SinoMed, Cochrane Library, PubMed, Web of Science, CINAHL, Scopus, and Embase) were searched from inception to December 2025. Eligible studies were published in English or Chinese from January 1, 2020, onward and enrolled people with PD using noninvasive, body-worn AI-enabled wearable devices for rehabilitation, assessment, or monitoring. Dissertations and full conference papers were included, whereas preprints and conference abstracts were excluded. Methodological quality was appraised using the Mixed Methods Appraisal Tool, 2018 tool. Results were synthesized narratively and mapped to characterize devices, sensing modalities, algorithms, and evaluation methods.
Results: A total of 66 studies involving approximately 3579 participants were included. Wearable devices mainly comprised multisensor modules, smart insoles, and wrist-worn devices, with accelerometers being the most frequently used sensors. Data collection was predominantly passive, and most studies were conducted in laboratory or clinical settings using single- or short-term sessions. Internal validation approaches, particularly leave-one-out and k-fold cross-validation, were common, whereas external validation was rare, and reporting of calibration and clinical decision thresholds was limited. Sensitivity and accuracy were the most frequently reported performance metrics, highlighting substantial heterogeneity in analytical methods and outcome reporting.
Conclusions: This scoping review systematically synthesized evidence on AI-enabled wearable devices for motor function assessment and rehabilitation in PD, complemented by an evidence map and guided by a rehabilitation- and nursing-oriented perspective, and identified key translational gaps between proof-of-concept studies and real-world rehabilitation workflows. Compared with previous reviews that primarily focused on monitoring functions or device performance, this review places greater emphasis on rehabilitation applications and nurse-led translation into practice, and proposes a conceptual “challenges and opportunities” framework to inform the design, evaluation, and reporting of devices and algorithms, while further highlighting key considerations for workflow integration and the implementation of decision-support systems. These findings have practical relevance for advancing continuity of rehabilitation across clinical, home, and community settings, and may help guide nurses in delivering continuous monitoring, personalized follow-up, and timely intervention, thereby improving the efficiency and accessibility of rehabilitation management.
Background: Patients often struggle to understand standard hospital discharge letters, increasing the risk of medication errors and misunderstandings. According to cognitive load theory (CLT), complex, information-dense texts can overload working memory and impair comprehension. Artificial intelligence tools that generate patient-centered versions could help reduce extraneous cognitive load and bridge this gap. However, evidence for their effectiveness remains limited.
Objective: This study aimed to evaluate whether GPT-4 (OpenAI)–generated patient-centered letters improve standardized patients’ retention and understanding of safety-relevant medical information compared with standard hospital discharge letters, and to explore potential effects on cognitive load as described by CLT.
Methods: In this prospective, randomized, crossover study, 48 trained standardized patients received a conventional discharge letter for an assigned disease (out of 3) and its matching GPT-4–generated patient-centered letter. Participants read one version first, identified predefined safety-relevant “learning objectives,” and then repeated the task with the alternate version. The primary outcome was the proportion of learning objectives fully, partially, or not reported. In a secondary analysis, results were stratified by content field (Medication, Organization, Prevention of Complications, Lifestyle/Disease Management) and Bloom taxonomy level (“Remember,” “Understand”).
Results: The letter type significantly influenced comprehension (odds ratio [OR] 1.74, 95% CI 1.45-2.08; P<.001). Patient letters, compared with discharge letters, led to higher rates of fully (490/1073, 45.7% vs 413/1073, 38.5%) or partially (322/1073, 30% vs 287/1073, 26.7%) stated learning objectives and fewer omissions (261/1073, 24.3% vs 373/1073, 34.8%). Participants performed better on “Remember” than on “Understand” learning objectives, regardless of letter type (OR 3.33, 95% CI 1.96-5.88; P<.001). Compared with standard hospital discharge letters, patient letters consistently improved results at both cognitive levels (“Remember”: 278/545, 51% vs 242/545, 44.4%; “Understand”: 212/528, 40.2% vs 171/528, 32.4% fully stated). The effect of patient letters varied by content field (P<.001). The greatest improvements were observed for “Medication” (170/254, 66.9% vs 129/254, 50.8% fully stated) and “Organization” (78/158, 49.4% vs 62/158, 39.2% fully stated). Improvements in the content field “Prevention of Complications” were modest, and those for “Lifestyle/Disease Management” were even smaller across all conditions. A total of 24.3% (261/1073) of key information remained unrecognized.
Conclusions: In this explanatory study, GPT-4–generated patient letters improved comprehension of safety-relevant discharge information among standardized patients, particularly regarding medication and organizational aspects. However, they were less effective in supporting higher-order understanding, such as risk prevention or lifestyle management. These hypothesis-driven findings can be interpreted within a CLT framework and may motivate prospective evaluation of multimodal, iterative supports.
Trial Registration:
Background: Digital recruitment methods offer opportunities to address challenges in clinical research participation, particularly in neurology. However, the impact of digital approaches across socioeconomic and demographic groups remains inadequately understood.
Objective: This study investigates the influence of sociodemographic factors on recruitment and attrition in a remote neurological research cohort, mapping participation pathways and identifying disparities to inform inclusive digital strategies.
Methods: We conducted a nonexperimental, observational longitudinal cohort study at Mayo Clinic using patient-portal invitations between March and July 2024 as part of a remote speech capture study. Eligibility criteria included age 18 years and older, US residence, and English proficiency. Of 5846 invited patients, progression was tracked across checkpoints (invitation, eligibility screening, electronic consent, and task completion) using Epic (Epic Systems Corporation) to obtain demographic information, Qualtrics (Qualtrics, LLC) for screening, PTrax (a Mayo Clinic–developed Participant Tracking System) for consent tracking, and the recording platform. Socioeconomic context was assessed using the Housing-based Socioeconomic Status (HOUSES) index, where higher values indicate higher socioeconomic status, and the Area Deprivation Index (ADI), where higher values reflect greater neighborhood disadvantage. Data diagnostics included Anderson-Darling tests for non-normality and Little missing completely at random (MCAR) test to characterize missingness. Associations between participation outcomes and age, sex, urbanicity, and socioeconomic indices were examined using nonparametric tests. Exact values and 95% CIs are reported. Analyses were conducted using BlueSky Statistics (BlueSky Statistics, LLC) and the Python package.
Results: Overall, 415 out of 5846 participants (7.1%) completed all study requirements. Completers were older (median age 66.4, IQR 56.0-72.5; 95% CI 65.1‐67.6 years) than noncompleters (median age 62.8, IQR 47.5-72.7; 95% CI 62.2‐63.2 years; <.001). Participants from more socioeconomically disadvantaged neighborhoods were less likely to respond (invitation nonresponder median ADI 45.0, IQR 29.0-63.0 vs interested median ADI 42.0, IQR 27.0-59.0; <.001), and completers had slightly lower ADI ranks than noncompleters (median 41.0, IQR 27.0-56.0 vs median 44.5, IQR 28.0-62.0; =.04). Urban participants enrolled faster (median 32.0, IQR 9.0-58.0; 95% CI 31.0‐37.0 days) than rural (median 41.0, IQR 22.0-65.0; 95% CI 37.0‐49.0 days; =.01). Female participants responded slower (median 38.5, IQR 14.8-66.3; 95% CI 35.0‐41.0 days) than males (median 32.0, IQR 8.0-57.5; 95% CI 29.0‐38.0 days; =.01). No significant differences were observed for the HOUSES index, and device type was unrelated to completion or timelines. Missingness for key variables was completely at random (MCAR ²=3.45; =.24).
Conclusions: Digital recruitment does not overcome traditional barriers to participation and may introduce new disparities related to age, urbanicity, and neighborhood disadvantage. These findings inform inclusive digital research strategies, including multichannel outreach, age-specific engagement, and rural technical support. This study applies longitudinal pathway analysis to digital neurology recruitment, offering actionable insights for improving inclusivity in remote research.
Background: As Canada’s climate changes, extreme heat events have become more frequent, a trend that is expected to continue. Extreme heat can lead to several negative health outcomes, which disproportionately impact vulnerable populations. Evidence-based, equitable interventions are needed to inform and protect the public from the health effects. Effective communication can aid this effort to improve health outcomes by emphasizing the connection between health risks and climate change and empowering people to act. Machine learning has applications in understanding current attitudes, beliefs, experiences, and behaviors within the target audience for public health messaging. Machine learning analyses of social media data have elucidated user perceptions of heat events in the literature; however, research is limited with respect to social media user perceptions, beliefs, and behaviors related to extreme heat, particularly in the Canadian context. Analyzing Canadian social media discourse related to extreme heat will help to address this research gap and inform future research and communications to reduce the risks of extreme heat.
Objective: The purpose of this research is to better understand Canadian discourse and emotions related to extreme heat by examining social media (Reddit). Our objectives include (1) identifying common discussion topics, concerns, and questions related to extreme heat among Canadian Reddit users; (2) analyzing sentiment and emotional responses to extreme heat discussions; and (3) investigating the relationship between topics, sentiment, and engagement for posts.
Methods: We collected data using the Reddit application programming interface (API), retrieving posts from 30 Canada-specific subreddits between February 12, 2023, and February 11, 2024, based on a predefined set of heat- and climate-related keywords. Posts and comments were structured as hierarchical tree models, with text consolidated into documents for analysis. Topic modeling, sentiment analysis, and emotion analysis were conducted; engagement was assessed using net upvote scores to gauge community approval.
Results: The analysis of 607 Reddit posts from 15,366 users revealed that discussions about extreme heat were most frequently centered around the keyword “heat,” which appeared in 82.5% (n=501) of the posts and 81.1% (n=25,253) of the comments. Topic analysis identified key themes related to heating and cooling costs, weather records, air conditioning, and health impacts, while sentiment and emotion analyses showed varying levels of positivity and negativity across subreddits.
Conclusions: Our findings present an initial snapshot into Canadian perspectives and information needs about extreme heat in Canada. In our sample, discussions on Reddit about extreme heat in Canada are dominated by concerns over heating and cooling costs, weather patterns, and personal adaptation strategies, reflecting both practical and policy-related challenges. Additionally, sentiment and emotion analyses suggest significant regional differences in public perception, which may be useful for informing health and risk messaging initiatives to better protect Canadians from the adverse health effects of climate change.
Background: Since the COVID 19 pandemic, health care and health information seeking have become increasingly digitally mediated. It remains unclear whether eHealth literacy is consistently associated...
Background: Since the COVID 19 pandemic, health care and health information seeking have become increasingly digitally mediated. It remains unclear whether eHealth literacy is consistently associated with health behaviors across different behavioral functions and social contexts in the post COVID 19 era. Objective: To synthesize post COVID 19 evidence on the association between eHealth literacy and health behaviors and to examine whether this association varies by health behavior domain, country income context, and population age structure. Methods: We conducted a PRISMA 2020 compliant systematic review and meta analysis registered in PROSPERO (CRD4201009048). PubMed, Embase, and the Cochrane Library were searched from inception to January 28, 2026. Observational studies were eligible if they assessed eHealth literacy using a validated instrument with an explicit score, measured health behavior outcomes that could be classified as health decision making, health promotion, or health management, and collected data in 2020 or later or explicitly reported the timing of data collection. Odds ratios and correlation coefficients were synthesized separately using random effects models with Hartung Knapp adjustment. Funnel plots and trim and fill were used to assess small study effects. Subgroup differences were tested using between group heterogeneity statistics. Studies with non comparable outcomes were summarized narratively. Results: Twenty two studies met the inclusion criteria, including 15 studies contributing quantitative effect estimates and 7 studies summarized narratively. Overall associations were directionally positive, with substantial heterogeneity and sensitivity to small study effects. Behavioral domain was the most consistent source of between study variation across effect size frameworks. Income context moderated associations in the correlation based synthesis, whereas age structure did not show significant moderation. Narrative evidence was most consistent for health decision making outcomes, more mixed for health promotion outcomes, and more variable and generally weaker for health management outcomes. Conclusions: In post COVID 19 studies, eHealth literacy is generally associated with health behaviors, but the strength and consistency of this relationship vary across behavioral domains and settings. Future longitudinal and intervention research using more comparable behavior measures is needed to clarify directionality and to inform context tailored strategies for improving eHealth literacy and health behavior.
Background: Patient-reported outcomes measures (PROMs) have become an important tool in measuring a patient’s health status from their own perspective; however, they are typically measured using sta...
Background: Patient-reported outcomes measures (PROMs) have become an important tool in measuring a patient’s health status from their own perspective; however, they are typically measured using standardized questionnaires which do not account for each patient's unique experience of health. Recent improvements in Natural Language Processing (NLP) provide new possibilities to extract PROM scores from unstructured or free-text patient narratives; however, the feasibility and minimal data requirements needed to accomplish this task remain uncertain. Objective: To assess the practicality of transformer-based models for predicting EuroQol EQ-5D-3L scores from patient narratives and to evaluate minimum data requirements, narrative length and data augmentation effects. Methods: This proof-of-concept study used synthetically generated patient narratives to evaluate methodological feasibility. Three transformer models (BERT, BioBERT, DistilBERT) were fine-tuned for regression from patient narratives representing all 243 EQ-5D-3L health states. The performance of the models in various scenarios including a range of sample sizes (n=100–850), narrative length (100–1000 words), and data augmentation conditions were compared. The performance of the models was assessed through fivefold cross-validation and additional validation on datasets created by ChatGPT and DeepSeek. Results: Each model was able to predict EQ-5D-3L scores using each of the different configurations of data (n=100-850 patients; 100-1000-word narratives). However, optimal results were obtained when training the models with 100-word narratives derived from the largest number of people (n=850), where mean squared error=0.03 (95% CI: 0.02-0.04), mean absolute error=0.13 (95% CI: 0.13-0.15), explained variance=0.77 (95% CI: 0.64-0.77), and intraclass correlation coefficient=0.85 (95% CI: 0.81-0.87). Furthermore, it was found that the shorter narratives (100 words) performed better than longer narratives (100-1000 words). Additionally, the use of data augmentation improved the predictive performance. Conclusions: Transformer models show promise in predicting EQ-5D-3L PROM scores from synthetic patient generated narratives, with a minimum of 250 patients providing around 100-word narratives required for reliable performance. The work provides both a methodological basis and empirical standards for AI-based PROM systems. However, clinical implementation will require validation using real patient-authored narratives prior to adoption. If validated, the use of this approach could provide evidence to support the inclusion of a patient's experience as a narrative into standardized outcome measures and support patient-centred healthcare evaluations.
The COVID-19 pandemic has spurred an unprecedented collection of multi-omic and clinical data of patient cohorts. However, fragmented datasets, inconsistent terminology, and poor integration with exis...
The COVID-19 pandemic has spurred an unprecedented collection of multi-omic and clinical data of patient cohorts. However, fragmented datasets, inconsistent terminology, and poor integration with existing knowledge of molecular pathways hinder effective analysis. We developed SPOKE-C19, a web-based knowledge graph tool that integrates multi-omics and clinical data from the INCOV and RECOVER cohorts with SPOKE, a knowledge graph with over 27 million nodes of biomedical entities linked by 53 million relationship edges. The platform allows mapping of empirical data to the network of known mechanistic relationships, facilitating interpretation of novel associations observed in post-acute sequelae of SARS-CoV-2 infection (PASC) studies.
SPOKE-C19 offers an intuitive user interface for researchers to query biomedical variables and to link empirical associations observed in clinical or experimental studies to generic mechanistic pathways and to precomputed relationships extracted from large COVID studies. The comprehensively documented interactive COVID-C19 platform promotes “connecting the dots” between disjoint domains and encourages cross-disciplinary collaboration, which is particularly important given the multi-faceted nature of PASC.
Background: Family caregivers of children with chronic health conditions experience significant physical and mental health burdens, including burnout, anxiety, depression, fatigue, and sleep disturban...
Background: Family caregivers of children with chronic health conditions experience significant physical and mental health burdens, including burnout, anxiety, depression, fatigue, and sleep disturbances. Despite this growing need, validated digital mental health tools tailored specifically to this population remain limited. Conversational agents powered by artificial intelligence (AI) offer a promising avenue for delivering on-demand, personalized mental health support, yet evidence-based development and evaluation of such tools for family caregivers is lacking. Objective: This study aimed to systematically describe the iterative development of COCO (Caring of Caregivers Online), a conversational agent designed to deliver Problem-Solving Therapy (PST) integrated with Motivational Interviewing (MI) principles, and to evaluate its usability and preliminary effect on the emotional well-being of family caregivers of children with chronic health conditions. Methods: COCO was developed and refined across four phases. Therapeutic dialogues were grounded in PST and MI principles and informed by evidence-based caregiver personas. The Wizard-of-Oz (WOZ) method was used across phases to iteratively collect naturalistic dialogues and refine COCO's conversational design. Usability was assessed using the System Usability Scale (SUS) and the Post-Study System Usability Questionnaire (PSSUQ). Caregiver emotions were measured pre- and post-session using six subscales of the Positive and Negative Affect Schedule - Expanded Scale (PANAS-X). In the final phase, a large language model (LLM)-powered version of COCO was developed using GPT-4 with few-shot learning and evaluated using persona-based methods. Results: COCO achieved a mean SUS score of 75.6%, reflecting acceptable usability. Participants demonstrated significant improvements in negative affect, sadness, guilt, fatigue, and serenity following PST sessions (p ≤ 0.03). Analysis of MI techniques across all phases revealed progressive refinement in conversational quality, with the LLM-powered COCO achieving the highest density of MI techniques per turn (2.56) and greater balance across MI strategy types, particularly in seeking collaboration and reflection. Conclusions: COCO is a feasible, usable, and preliminarily efficacious conversational agent for supporting the mental health of family caregivers of children with chronic conditions. The iterative, human-in-the-loop development approach was instrumental in producing empathetic, therapeutically grounded responses. The systematic development and evaluation process described here can serve as a guide for similar conversational agent intervention studies. Future work will explore multi-agent architectures and retrieval-augmented generation (RAG) to further enhance personalization, controllability, and scalability toward clinical deployment.
Background: The 21st Century Cures Act information blocking regulations led to many health care providers (HCPs) altering policies to electronically release test results to patients immediately upon t...
Background: The 21st Century Cures Act information blocking regulations led to many health care providers (HCPs) altering policies to electronically release test results to patients immediately upon their availability. Objective: To understand how often patients view results in the patient portal before hearing from their HCP, and whether they are given the option to decide how results are communicated. Methods: Using data from the 2024 Health Information National Trends Survey on U.S. adults who received recent test results via patient portal (N=6,045), we examined whether patients viewed test results before hearing from their HCP, were given the option to decide how test results were communicated, and understood results viewed before hearing from an HCP. Results: 70% of patients who received results viewed them in their patient portal, most of whom viewed results before hearing from their HCP (58% overall). 28% of patients and 33% of portal users reported being given the option to decide whether they wanted to receive test results before hearing from their HCP. Two-thirds of patients understood results they viewed in their patient portal before hearing from their HCP (66%). Conclusions: While most patients viewed results before discussing with their HCP, only one-third reported being given the option to decide how results would be communicated and two-thirds of patients who viewed immediately released results understood their implications. Clearly presenting the option to decide when test results are communicated and incorporating patient preferences in portal communications could help empower patients and mitigate potential worry.
Background: Generative artificial intelligence (GenAI) tools powered by large language models (LLMs) are increasingly used by the public to seek health information. Unlike traditional web search, GenA...
Background: Generative artificial intelligence (GenAI) tools powered by large language models (LLMs) are increasingly used by the public to seek health information. Unlike traditional web search, GenAI systems generate conversational answers, which may influence how users assess credibility, manage uncertainty, and decide whether to verify information or consult clinicians. Evidence is needed to clarify facilitators, barriers, and user practices in GenAI-supported health information seeking. Objective: This systematic review synthesizes empirical research on consumer and patient health information seeking with GenAI/LLM tools, focusing on study contexts, adoption and use outcomes, facilitators and barriers, with implications for clinician-patient interactions. Methods: We conducted a review following PRISMA-ScR. Records were identified through database searching and screened using predefined eligibility criteria. Included studies were extracted using a structured form capturing study characteristics, GenAI tool type, health context, outcomes, as well as facilitators and barriers. Findings were synthesized using structured grouping aligned to the RQs. Results: The review included 27 studies. GenAI was used for symptom appraisal, condition understanding, treatment options, and care navigation. Facilitators emphasized convenience and clarity, including efficiency and access (29.6%, n=8), comprehensibility and presentation quality (40.7%, n=11), personalization and specificity (18.5%, n=5), and affective or interpersonal comfort (18.5%, n=5). Barriers were dominated by credibility and trust concerns (48.1%, n=13), particularly when accuracy cues or citations were missing or difficult to interpret. Additional barriers included perceived unsuitability for complex, urgent, or emotionally charged situations (18.5%, n=5), privacy or data security concerns (14.8%, n=4), limited prompting skills (7.4%, n=2), and modality or interaction constraints that hindered credibility assessment and information comparison (18.5%, n=5). Literacy-related capability was reported in 22.2% of studies (n=6), and verification-supporting features (e.g., visible sourcing, transcripts, and save/revisit/share functions) were reported in 18.5% of studies (n=5). Conclusions: GenAI is used for diverse health information needs, but reliance is shaped by trust, perceived risk, and verification capacity. Future research should improve reporting of tools and prompting conditions, standardize measures of reliance and verification, and evaluate use in higher-stakes and underserved contexts to inform safer design and public guidance.
