DIGITAL HEALTH最新文献

Background: Telehealth applications and mobile services have been growing in popularity. As their reach expands service across international boundaries, it remains unclear to what extent Chinese immigrants residing in the United States are using China-based medical applications and the factors impacting their uptake. Although transnational telehealth apps are beneficial in bridging cultural and linguistic gaps, they come with distinct risks and challenges that need to be further explored.

Objectives: The study had three aims: (1) estimate the prevalence of China-based telehealth app usage by Chinese migrants in the US, (2) identify factors associated with China-based telehealth apps utilization among Chinese migrants in the US, and (3) describe how Chinese migrants in the US are using and can use China-based telehealth apps remotely from the US.

Methods: Four focus groups (n = 17) and a cross-sectional survey (n = 227) were conducted among recent Chinese immigrants to the US (arrived in the past 10 years).

Results: Overall, 15% indicated usage of China-based telehealth apps while living within the US. Use of China-based telehealth apps while living in the US was associated with: higher perceived frequency of experiencing healthcare discrimination in the US (odds ratio (OR): 1.43, 95% CI: 1.14-1.80), younger age (OR: 7.86, 95% CI: 1.32-47.01), female sex (OR: 4.29, 95% CI: 1.50-12.23), living in a community with a large Chinese community (OR: 9.53, 95% CI: 1.90-47.79), and lack of medical insurance (OR: 51.59, 95% CI: 3.88-685.70). Some Chinese migrants living in the US are using China-based telehealth apps to consult with medical providers in China as their first line of medical consultation.

Conclusion: Findings suggest uptake of China-based telehealth are partially driven by negative experiences within the US healthcare system. These results are indicative of possible shortcomings in existing healthcare services that diminish the capacity to appropriately address the needs of immigrant communities and groups.

Objectives: Skin cancer represents a significant public health concern, and consistent sunscreen use reduces risk. With social media emerging as a dominant source of health information, unconventional video formats have gained increased prominence. Despite these shifts in health promotion practice, limited research has examined how digital message formats influence health-related beliefs. The objective of this study is to examine how different social media-based message formats (amateur video, professional video, written text, and control) and participant characteristics affect perceptions of skin cancer risk, sunscreen efficacy, and intentions to wear sunscreen to prevent skin cancer.

Methods: A national sample of white US adults (N = 538) were assigned to one of four digital message conditions in an online randomized controlled experiment. Participants completed a fully automated Qualtrics-based survey grounded in the Extended Parallel Process Model (EPPM). EPPM variables, audience perceptions of the message, confidence in identifying signs of skin cancer, and behavioral intentions to wear sunscreen were evaluated through a series of one-way analyses of variance (ANOVAs). Standard multiple regression analysis was used to assess associations between message assignment, demographic characteristics, and EPPM variables.

Results: Both amateur and professionally produced video formats significantly increased participants' beliefs in sunscreen's effectiveness compared to the control condition. The only significant audience perception differences were higher engagement and lower boredom ratings in the professional video condition compared to the amateur video and text-only conditions. Demographic variables including sex, generation, skin sensitivity, and education were significantly associated with differences in perceived threat and efficacy.

Conclusion: Findings indicate that both amateur and professional video-based social media messages can effectively promote sun safety. Public health campaigns aiming to reduce skin cancer risk may be strengthened by incorporating varied digital message formats and tailoring content to key demographic characteristics of the target audience.

Pacific Island Countries and Territories face unique challenges delivering equitable health services due to vast geographic dispersion, limited transport infrastructure, workforce shortages, and increasing climate risks. This article aims to highlight the urgent need for interoperable, standards-based digital health systems in the Pacific and to outline practical pathways for strengthening regional collaboration and resilience. Digital health is not merely a technological upgrade but an essential enabler of health equity and system sustainability. While telehealth and mobile health have gained traction, their full potential is undermined by fragmented, non-interoperable systems that restrict data flows, duplicate effort, and limit impact. Recent progress including national strategies in some Pacific countries and new regional frameworks championed by the Pacific Health Information Network demonstrates growing commitment to shared standards and people-centered solutions. A key pillar for digital health in the Pacific is the need to address language access capabilities thus building multilingual interoperability of systems for clinical safety, equity and effective public-health communication. Practical examples show how shared infrastructure and collaboration strengthen connected care and workforce development across dispersed islands. Achieving this transformation requires strong governance, climate-resilient systems, ethical data protections, and sustained partnerships. Countries, partners, and institutions are called to embed interoperability in national strategies, invest in local capacity, and advance regionally coordinated initiatives such as a Centre of Excellence for Telehealth and Interoperability further promoting language access and multilingual standards for shared health services and workforce. Together, these actions can build a digitally connected health ecosystem that bridges distances, strengthens resilience, and delivers equitable care for all Pacific communities.

Objective: This cross-sectional study aimed to analyze the content and quality of migraine-related videos on Chinese video-sharing platforms.

Background: In recent years, the escalating incidence and prevalence of migraine have imposed an increasing burden on individuals. Short-video platforms, such as TikTok and BiliBili, have demonstrated immense potential for disseminating health-related information. While a substantial number of migraine-specific videos are available on TikTok and BiliBili, their quality and reliability remain largely uncharacterized.

Method: On 24 August 2025, short videos related to migraine were gathered from BiliBili and TikTok via a comprehensive Chinese language search. Following the extraction of fundamental information, each video was evaluated using the Global Quality Score (GQS), the modified DISCERN tool (mDISCERN) score, and the Patient Education Materials Assessment Tool (PEMAT). Furthermore, Spearman's correlation analysis was employed to investigate the relationships among video variables, GQS, DISCERN, and PEMAT scores.

Results: TikTok demonstrated greater popularity than BiliBili, evidenced by higher numbers of likes, collections, comments, and shares. Overall, short videos on TikTok generally received superior scores across all evaluation metrics compared to those on BiliBili. Furthermore, it was observed that videos shared by Neurology Professionals consistently scored higher in GQS, mDISCERN, PEMAT-U, and PEMAT-A than those from other contributors. Spearman's correlation analysis indicated no significant association between video variables and GQS or mDISCERN scores.

Conclusions: The quality and reliability of migraine-related videos on both BiliBili and TikTok were found to be suboptimal. Notably, videos shared by Neurology Professionals tended to exhibit superior quality and trustworthiness. Therefore, individuals should exercise caution when consuming short-form video content.

Objective: To develop an interpretable stacking ensemble model for predicting in-hospital mortality in intensive care unit (ICU) patients with CKD and sepsis and to deploy it as a web-based tool for bedside clinical use.

Methods: Data were extracted from the MIMIC-IV 3.0 database and split into training and test sets at a 7:3 ratio. Feature selection was performed by combining the least absolute shrinkage and selection operator (LASSO) regression with the Boruta algorithm. Eight machine learning (ML) models were trained and optimized via ten-fold cross-validation and grid search. The two models with the highest area under the curve (AUC) in the training set were combined using a stacking ensemble strategy. SHapley Additive exPlanations (SHAP) were applied to improve interpretability. Model performance was compared with the SOFA score.

Results: A total of 5344 ICU patients with CKD and sepsis were included, with an in-hospital mortality rate of 19.1%. After feature selection, 16 variables were retained. In the training set, XGBoost and LightGBM performed best. The stacking model achieved an AUC of 0.757 on the test set, outperforming SOFA (AUC = 0.668). SHAP analysis identified age, Acute Physiology Score III, Simplified Acute Physiology Score II, and respiratory rate as the top predictors. The model was also deployed as a publicly accessible web application.

Conclusion: The stacking ensemble model demonstrated good discriminatory performance and interpretability for predicting in-hospital mortality in ICU patients with CKD and sepsis. Its web-based deployment provides a convenient platform for early risk assessment, although external validation is needed to confirm its broader applicability.

Objectives: Douyin (TikTok) has gradually emerged as an important channel through which the public obtains health information. This study aimed to evaluate the content, quality, and reliability of syphilis-related short videos on Douyin.

Methods: We conducted two rounds of data collection on Douyin, extracting video duration, engagement metrics, uploader identity, and video content for syphilis-related videos. Video quality and reliability were assessed using the Global Quality Score (GQS) and the modified DISCERN (mDISCERN) tool. Correlation analyses were performed between video metrics and quality scores.

Results: A total of 81 and 95 videos were included in the first and second rounds, respectively. Clinical manifestations were the most frequently discussed topic, whereas key information on diagnosis and prognosis was insufficiently covered. In the first round, the median GQS was 2.00 (IQR: 2.00-3.00), and the median mDISCERN score was 3.00 (IQR: 3.00-3.00). In the second round, the median GQS was 2.00 (IQR: 2.00-2.00), and the median mDISCERN score was 2.00 (IQR: 2.00-2.00). In both analyses, videos uploaded by healthcare professionals had higher GQS and mDISCERN scores than those uploaded by individual users (p < 0.05). No significant correlations were found between video engagement metrics and either GQS or mDISCERN scores in either round (p > 0.05).

Conclusions: The overall quality and reliability of syphilis-related videos on Douyin are suboptimal, and current content may not adequately meet public health information needs. Strengthening the involvement of healthcare professionals and improving content structure are essential to enhance the health education value of short video platforms.

Objective: Conventional scale-based diagnostic approaches are increasingly insufficient for addressing the growing mental health challenges among adolescents. Leveraging advances in artificial intelligence, this study aims to develop an accurate, efficient, and scalable model for early identification of adolescent depression risk using large-scale census data, and to identify key daily life factors associated with mental health outcomes.

Methods: Data were obtained from the 2021 National Survey of Children's Health, including 50,892 adolescents and 463 variables. Based on prior literature, 60 relevant variables were selected. Three progressively structured hypotheses concerning the relationships between adolescent depression and developmental environments were proposed. Machine learning models, including decision trees, XGBoost, support vector machines, and neural networks, were applied to predict depression risk. Mediation analysis was conducted to examine the pathways through which living conditions influence mental health.

Results: The optimal model demonstrated strong predictive performance, achieving an accuracy of 0.85 and an AUC exceeding 0.87. Feature importance analysis identified several key predictors. Mediation analysis indicated that living conditions exerted a direct effect of 0.225 on mental health, while physical activity and diet quality partially mediated this relationship.

Conclusion: Living conditions are critical indicators for early identification of adolescent depression risk. The use of nationwide census data enables timely screening and targeted intervention. Improving dietary habits and increasing physical activity may serve as effective preventive strategies for adolescent mental health disorders.

Background: Exposure to therapeutic landscapes has been consistently associated with reduced stress, improved affect, and enhanced emotion regulation among young adults. However, access to such environments is often limited on urban campuses where anxiety is prevalent. In response, this study conceptualizes the virtual therapeutic landscape (VTL) and proposes a design and evaluation model that translates therapeutic landscape theory into immersive virtual reality (VR). Methods: A three-stage mixed-methods design was employed. Semi-structured interviews (n = 18) were thematically analyzed to identify core experiential dimensions that informed VTL model development. Expert analytic hierarchy process (AHP) ratings (n = 12) yielded weights for domains and sub-criteria with acceptable consistency. On this basis, a standardized VTL exposure was administered to university students (n = 60), who completed psychometric questionnaires. Text-mined features from open-ended responses were integrated with these outcomes to refine the expert-weighted model. Results: The calibrated model produced a coherent weighting structure, with sensory experience receiving the highest weight among the experiential dimensions. Brief VTL exposure significantly reduced state anxiety (p < 0.001, d = 1.040) and negative affect (p < 0.001, d = 0.570) and increased subjective vitality (p < 0.001, d = 0.794). Text mining supported an architecture in which sensory-narrative coupling and low-friction interaction experience act as primary levers, while personalization experience moderates these effects. Conclusions: This study develops a design and evaluation model for VTL targeting anxiety and emotion regulation in university students. Brief VTL exposure has shown measurable psychometric change; long-term effects and variation across VTL types remain priorities for future research.

Objective: Neuropsychiatric complications following a stroke can impede recovery and reduce the quality of life. Current predictive methods for poststroke anxiety (PSA) are limited by inadequate feature selection and lack of interpretability. This study aimed to develop an interpretable machine learning model utilizing a wide range of clinical data to detect high-risk PSA patients early, enabling personalized interventions. Methods: This retrospective multicenter study included 238 stroke patients from 10 Chinese hospitals spanning from 1 January 2022 to 11 June 2025. Data encompassing demographic, clinical, biochemical, and psychosocial factors were gathered. Feature selection involved univariate analysis followed by least absolute shrinkage and selection operator (LASSO) regression. Seven machine learning models-logistic regression, extreme gradient boosting (XGBoost), light gradient boosting machine (LightGBM), random forest, decision tree, K-nearest neighbors, and stacking-were constructed and assessed using cross-validation. Feature importance was determined using SHAP (Shapley Additive Explanations), and a nomogram was developed based on the final model. Results: Among the 238 patients, 109 were diagnosed with PSA. In the test set, the logistic regression model exhibited the best performance, achieving an area under the curve (AUC) of 0.981, accuracy of 0.917, sensitivity of 0.867, specificity of 0.952, and an F1 score of 0.897. SHAP analysis identified recurrent stroke, income level, payment type, occupational stress, overwork, sleep quality, continuous drinking history, history of hypertension, diabetes, hyperlipidemia, hyperhomocysteinemia, white blood cell (WBC) count, total cholesterol (TC), low-density lipoprotein (LDL), fibrinogen (FIB), activated partial thromboplastin time (APTT), National Institutes of Health Stroke Scale (NIHSS) score, and Barthel index as crucial predictors. A nomogram incorporating the top 10 SHAP-ranked features was devised to assist in clinical decision-making. Conclusion: The machine learning model demonstrated high accuracy and interpretability in predicting PSA risk. Through the integration of SHAP analysis and nomogram visualization, it offers a practical tool for clinicians to recognize high-risk PSA patients and customize management strategies to improve poststroke outcomes.