Hypertension, a leading contributor to global mortality, affects 1.28 billion adults worldwide, and its early detection is essential to prevent adverse outcomes. However, relying solely on blood pressure for hypertension detection proves insufficiently accurate. Recognizing this limitation, we have identified a range of multi-dimensional hemodynamics as novel biomarkers that can offer more comprehensive cardiovascular insights for detecting hypertension. Specifically, we propose Hyde, a cross-user hypertension detection system that leverages ballistocardiogram (BCG) signals to infer multi-dimensional hemodynamics and naturally detect hypertension without disrupting users' daily routines. Hyde features a multi-task, multi-branch, unsupervised domain adaptation learning framework, enabling simultaneous prediction of five hemodynamic biomarkers in the auxiliary regression task. This design addresses several challenges, including unlabeled target domain, single-source bias, and complete negative transfer. We implement a flexible optic fiber sensor mat to collect high-quality BCG signals and deploy it in a four-month clinical trial involving 85 subjects with different conditions. The results show that Hyde can accurately identify hypertension with an average accuracy of 97.65%, outperforming state-of-the-art approaches by 21.65% $\sim$ 38.16%. To further validate Hyde's reliability, we test the system with another 33 participants under natural, unconstrained sleep conditions, revealing consistent performance. These findings highlight Hyde's potential for long-term, non-contact hypertension assessment, offering valuable clinical implications for personalized healthcare and improving cardiovascular outcomes.

Common intensive care severity scores may be less accurate for critically ill patients with cardiovascular disease. We develop and validate a data-driven severity index for 30 day mortality prediction using routinely collected hospital data. We performed a retrospective multicenter cohort study. Model development used the Medical Information Mart for Intensive Care databases (39,458 admissions), split into training (10,457) and internal validation (29,001). External validation used a large United States multicenter intensive care database (9,222) and an independent Guangdong Second Provincial General Hospital cohort (235). We compared machine learning models with three established severity scores. We evaluated discrimination, calibration, clinical utility, and model explanations. Here we show that a gradient boosted decision tree model achieves the highest discrimination, with areas under the receiver operating characteristic curve of 0.853, 0.802, and 0.853 in the internal and two external cohorts, versus 0.724 to 0.765 for established scores. Calibration is good (standardized mortality ratio 1.00; 95% confidence interval 0.98 to 1.02) and remains consistent across age, sex, and race subgroups. Decision curve analysis shows a higher net benefit than conventional scores across thresholds of 0.20 to 0.60. Key contributors include comorbidity burden, urine output, activity status, respiratory rate, and minimum oxygen saturation. A gradient boosted decision tree-based severity index improves discrimination, calibration, and clinical utility compared with conventional intensive care scores and generalizes across external cohorts. The model explanations support interpretable risk stratification for critically ill cardiovascular disease. Many people with serious heart and blood vessel disease are treated in intensive care units. Doctors often use standard scoring tools to estimate a patient’s risk of dying, but these tools may be less accurate for this group. In this study, we aimed to build a more reliable risk score using information that is routinely recorded in electronic health records. We trained and tested several computer models on a large intensive care database and then checked the best model in two independent patient groups from different hospitals. We find that the best model predicts 30 day death risk more accurately than the standard scores and remains reliable across different patient subgroups. This could help clinicians identify high risk patients earlier and support better planning of care. Future work should test how it performs in real time clinical use. Ren et al. develop and validate an XGBoost-based acute illness severity index for critically ill cardiovascular patients using multicenter ICU data and compare it with APSIII, OASIS, SOFA and machine learning baselines. The index shows better discrimination, calibration and net benefit across validations, with SHAP highlighting key drivers of risk.

BackgroundIn China, the HIV prevalence among men who have sex with men (MSM) is still rising, with a large proportion of people never tested for HIV in this group. HIV self-testing (HIVST) offers a user-empowered approach in expanding testing coverage. Peer navigation, recognized for its effectiveness in supporting MSM in HIV care, is proposed here as a potential enhancement to HIVST. This study aims to evaluate peer navigation in promoting digital secondary distribution of HIVST among MSM in China.MethodsWe plan to recruit 400 participants (indexes) identified as key nodes in the network using our previously developed algorithm, RiskRank, via the BlueD platform. RiskRank is a method to prioritize nodes for targeted interventions by incorporating their topological features on the multilayer complex networks and considering the underlying epidemic dynamics. The eligible participants will be randomized into the intervention group (peer navigation) and the control group in a 1:1 allocation ratio. In the intervention group, 20 peer navigators will be recruited and trained. Each peer navigator will provide the peer navigated intervention to 10 participants (indexes). The peer navigation process consists of three modules ([BHSD], [SSDE], and [CFPS]). In the control group, we will implement the standard HIVST secondary distribution we have developed before. The index will distribute the HIVST kits to the alters, leveraging their existing social connections to try to make the kits reach those who may not have access to traditional testing services. All index participants will be requested to complete a baseline survey and a 3-month follow-up survey. Both indexes and alters will complete a survey upon returning the results, by taking a photo of the used kits with the unique identification number.DiscussionHIV testing rate remains to be below the desired levels among MSM in China. Creative approaches like peer navigation are essential to enhance HIV testing uptake among key populations. The findings of the trial can offer valuable scientific evidence and insights into promoting the secondary distribution of HIV self-testing (HIVST) to reach key populations not yet reached by current testing services.Trial registrationThe study has been registered with the Chinese Clinical Trial Registry ChiCTR2400093985; Date of Registration: 16 December, 2024; Protocol Version: R3.

Purpose:To systematically evaluate the application of artificial intelligence (AI) techniques in X-ray sensor-based coronary angiography for cardiovascular disease (CVD) diagnosis, mapping publication trends, geographic and topical hotspots via bibliometric analysis, and critically reviewing disease-specific AI methodologies and performance to inform future research and clinical integration. Non-angiographic inputs were considered only when angiography served as the reference standard or when the algorithm was explicitly integrated into an angiography-based workflow. Methods:A two-part approach was undertaken. In Part I, we performed a bibliometric analysis of English-language original research and reviews published between 1 June 2010 and 1 June 2025, retrieved from Web of Science, Scopus, and PubMed. Records (n = 123) were screened using a PRISMA flowchart and analyzed with CiteSpace v6.3.R1 to identify annual publication trends, country contributions, co-authorship networks, and keyword clusters. In Part II, we conducted a structured literature review of the AI methods reported in these studies, organizing findings by three major clinical categories-acute myocardial infarction, ischemic cardiomyopathy, and unstable angina-and extracting model architectures, data sources, and diagnostic performance metrics (accuracy, sensitivity, specificity, and AUC). Results:Bibliometric analysis revealed three publication phases: a formative period (2010-2017) with <3 papers/year; rapid growth (2018-2021) culminating in a peak of 28 papers in 2022; and sustained interest into 2025. The United States (n = 39) and China (n = 34) led contributions, and keyword clustering highlighted central themes around "artificial intelligence," "coronary artery disease," and "computed tomography angiography." In disease-specific review, convolutional neural networks (CNNs) and CNN-LSTM hybrids predominated, achieving AUCs from 0.724 to 0.997: for acute myocardial infarction detection, accuracies of 90%-95% and AUCs up to 0.99; for ischemic cardiomyopathy differentiation, accuracies of 75%-98% and AUCs up to 0.93; and for unstable angina prediction, overall accuracies of 89%-95%. Classical machine-learning models (XGBoost and random forest) also showed robust performance (AUC 0.77-0.94). Key challenges include dataset heterogeneity, limited multicenter validation, and model interpretability. Conclusion:AI, particularly deep-learning frameworks, substantially enhances the accuracy and efficiency of CVD diagnosis via X-ray coronary angiography. However, current evidence is constrained by small single-center datasets, limited external validation, inconsistent leakage safeguards, and scarce calibration/decision-curve reporting. To advance clinical adoption, future efforts should emphasize large-scale, multicenter validation studies, development of explainable AI models, and seamless integration into cardiology workflows.

Background:Low physical activity (LPA) is associated with cardiovascular and cerebrovascular pathologies. This study aimed to assess the prevalence of several noncommunicable diseases relating to LPA. Methods:Using the 2021 Global Burden of Disease data set, we modelled LPA-related disease burdens across 204 countries and territories, quantifying mortality counts, age-standardised mortality rates, and disability-adjusted life years (DALYs) for five noncommunicable diseases. We conducted multivariable stratification analyses to assess variations by gender, age, and sociodemographic index (SDI) quintiles. We used age-period-cohort modelling to project burden trajectories, while applying counterfactual decomposition frameworks to delineate synergistic interactions between LPA and risk factors. Results:We found that LPA accounted for 555 101 related deaths globally in 2021 across the five studied pathologies, mostly among individuals aged 60-94 years. Association between LPA-related disease burden and SDI followed a U-shaped distribution across regions and diseases. Among individuals aged 60-89 years, LPA-related deaths were significantly higher in women than in men, indicating a disproportionate burden on elderly females. Ischaemic heart disease (IHD) trends stabilised in low- and middle-SDI regions but declined significantly in high-SDI regions, underscoring global health disparities. From 2007 to 2011, LPA DALYs and mortality risk ratios for IHD, stroke, and lower extremity peripheral arterial disease declined from >1 to <1, whereas diabetes mellitus exhibited an opposite trend, highlighting LPA's persistent and significant impact on diabetes-related morbidity. Demographic shifts and epidemiological transitions were primary drivers of LPA-related disease burden across five pathologies. In high-SDI regions, epidemiological changes predominated, whereas population growth was a key factor in low- and middle-SDI regions. Synergistic interaction of these factors with LPA is projected to substantially amplify future disease burden. Conclusions:Physical activity should be increased among elderly women to address health risks associated with LPA. Likewise, urgent public health interventions are needed for LPA-related diabetes. As IHD burden rises in low- and middle-SDI regions, vascular disease care strategies require optimisation. Moreover, high-SDI regions should strengthen nationwide physical activity promotion, while low- and middle-SDI areas must enhance healthcare infrastructure and manage population growth to reduce LPA-related disease burdens.

Background The cardiovascular-kidney-metabolic (CKM) syndrome, first proposed by the American Heart Association (AHA) in 2023, represents a groundbreaking conceptual framework that integrates these three interrelated conditions into a unified clinical entity. Despite growing research on its prevalence, risk factors, and clinical management, the regional burden of CKM syndrome remains poorly characterised. To address this gap, we aimed to estimate the prevalence of CKM syndrome, providing critical insights into the regional impact of this novel disease definition. Methods In this study, we conducted literature retrieval in both English (PubMed, Web of Science and Wiley Online Library) and Chinese databases (CNKI and Wangfang), as well as the journal official websites (e.g., American Heart Association (AHA) and American Society of Nephrology (ASN)) from database inception until January 20, 2025, followed by an update search until April 1, 2025. Grey literature such as posters and preprint articles, and citations from the identified reviews were also searched for. Cross sectional and cohort studies were included without language limitation. Studies employing other study designs or were done in people who were not representative of the general population (e.g., people with specific diseases) were excluded. Summary data were obtained from included studies. The primary outcomes were the prevalence of CKM syndrome and its different stages (stages 0-4) among general population. The combined prevalence was obtained with Freeman-Tukey Double Arcsine Transformation method. The estimated annuls percentage change (EAPC) was employed to explore the trend of CKM syndrome. This study is registered with PROSPERO (CRD420251037912). Findings From 2,708 identified 2,708 related articles, 28 studies with 29 datapoints, encompassing 1,561,209 individuals, were included. The overall pooled prevalence of CKM syndrome (Stages 1-4) in the general population was 0.88 [95% CI 0.86-0.91]. This estimate was 0.85 [95% CI 0.76-0.91] in a sensitivity analysis selecting one representative study per database to test the magnitude of potential duplicate bias. The combined prevalence of stages 1, 2, 3 and 4 was 0.23 [95% CI 0.19-0.27], 0.46 [95% CI 0.41-0.51], 0.08 [95% CI 0.05-0.11], 0.07 [95% CI 0.04-0.12], respectively, displaying as the Stage 2 patients were the majority of CKM syndrome. The EAPC of CKM syndrome in the period of 1991-2021 was (-0.55% [95% CI -0.90 to 0.21], p=0.0024), displaying a significant decreased trend. Stratified by countries, the pooled estimates were 0.91 [95% CI 0.90-0.93] for USA, 0.90 [95% CI 0.87-0.93] for China, and 0.77 [95% CI 0.69-0.84] for other countries (UK, Italy and South Korea). CKM syndrome prevalence demonstrated an increasing trend with a higher proportion of males (male/female ratio <0.98) and with increasing mean age (up to 56.5 years). Statistically significant disparities were observed across social development index (SDI) level, data source and countries. Interpretation This study provides the pooled regional prevalence of CKM syndrome in the general population; these findings are valuable for understanding the current burden of CKM syndrome and facilitating more research into the clinical management and prevention. While a slight decreasing temporal trend was observed based on the included studies, the relatively high combined prevalence suggests more epidemiological research into missing regions, such as Africa and South America, to verify this finding. ### Competing Interest Statement The authors have declared no competing interest. ### Funding Statement L. L was supported by the InnoHK Project at the Hong Kong Centre for Cerebro-cardiovascular Health Engineering (COCHE). J.D.Z was supported by HKU Seed Fund for New Staff Basic Research (No. 103034014) and HKU Daniel and Mayce Yu Medical Development Fund for Research Start-Up (No. 200010837). K.T receives a Chair in Family and Community Medicine Research in Primary Care at UHN and a Research Scholar Award from the Department of Family and Community Medicine at the University of Toronto. ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes The summary table of extracted data from the included studies is provided in Table 1 and Table 2. All datasets generated and analyzed, including the search strategy, data extracted, and quality assessment, are available in the Article and on request from the corresponding author (JDZ, jdzhou{at}hku.hk).

Chronic ultraviolet (UV) exposure is the primary cause of skin photoaging, leading to wrinkles, pigmentation changes, and loss of dermal elasticity. This systematic review and network meta-analysis evaluated the efficacy and safety of topical compounds for treating skin photoaging. A comprehensive search identified 23 RCTs with 3905 participants, comparing anti-aging agents. Bayesian network meta-analysis showed isotretinoin, retinol, and tretinoin significantly improved fine wrinkles, with isotretinoin ranked highest. Tazarotene was most effective for coarse wrinkles, while glycolic acid reduced roughness. Tretinoin and retinol were superior for hyperpigmentation. Safety analysis indicated tretinoin had the most favorable profile, whereas tazarotene and glycolic acid had higher adverse event risks. Isotretinoin and tretinoin emerged as the most balanced treatments across efficacy and safety. These findings provide evidence-based guidance for clinical decision-making in anti-photoaging therapy and underscore the potential for these agents to be integrated into routine dermatologic practice, particularly for patients seeking effective and well-tolerated topical interventions. However, limitations included limited racial diversity, potential commercial bias, and variability in dermatological assessments. These findings provide evidence-based guidance for clinical decision-making in anti-photoaging therapy.

Differentiating acute exacerbation of chronic obstructive pulmonary disease (AECOPD) from acute heart failure (AHF) is clinically challenging due to overlapping symptoms, especially in resource-limited settings lacking radiological/ultrasonographic tools. This study developed an eXtreme Gradient Boosting (XGBoost) model for differential diagnosis using Database: Medical Information Mart for Intensive Care (MIMIC) and two Chinese hospital cohorts, comparing it with a guideline-based model and applying Shapley Additive Explanations (SHAP) analysis to identify key biomarkers. The XGBoost model showed high discriminatory performance (area under the curve [AUC]: 0.94-0.98 across development/validation, outperforming the guideline-based model's AUC of 0.53) with consistent accuracy across age/sex subgroups. Key biomarkers included NT-proBNP and total bilirubin. This robust model enables rapid, accurate differential diagnosis in resource-constrained emergency settings.

Integrating medical challenges with deep learning algorithms has become a prominent area of current research. In recent years, the generative capabilities of denoising diffusion models have garnered significant attention, while the application of deep learning methods to assist preoperative surgical planning has demonstrated considerable promise. Concurrently, we observe that chronic total occlusion (CTO) procedures present considerable technical challenges and carry a high incidence of complications. Obtaining preoperative information regarding the occluded segment's pathway while minimizing contrast agent exposure could enhance surgical success rates and safety. Addressing this concept, this paper proposes a clinically interactive Region of Interest (ROI)-guided conditional diffusion restoration method. This approach achieves ROI-specific restoration through conditional diffusion modelling and inpainting. To overcome the scarcity of CTO samples, we constructed a training dataset by annotating normal coronary arteries and simulating occlusions. Ultimately, while maintaining pixel-level consistency in non-ROI regions, we successfully reconstruct terminal branch structures. This reduces reliance on exploratory angiography and fluoroscopy, providing reference for surgical pathway planning and offering potential assistance in CTO surgical procedures.

This study unveils a cutting-edge camera-based system for the automated monitoring of hand hygiene practices within healthcare settings. Utilizing advanced computer vision and machine learning technologies, our system employs three strategically placed synchronized cameras around a wash basin. These cameras capture the handwashing process from multiple perspectives, allowing for detailed analysis of hand movements including finger and wrist dynamics. The extracted skeletal coordinate data are processed by a Gesture Category Model (GCM), which automatically identifies handwashing gestures. The model is rigorously trained on a dataset comprising video recordings from 55 healthcare professionals, focusing on the World Health Organizations seven-step hand-washing protocol. Furthermore, we introduce a Counting Algorithm to quantify the frequency and duration of each gesture, coupled with a Quality Assessment Model (QAM) that evaluates compliance with hand hygiene standards. The systems precision and its strong correlation with expert annotations highlight its potential to significantly enhance hand hygiene compliance and reduce healthcare-associated infections.

The integration of large language models (LLMs) into public health systems promises significant improvements in disease surveillance, clinical communication, and health education. However, their deployment raises fundamental ethical and security risks-ranging from data privacy breaches to algorithmic bias and misinformation. This survey systematically reviews recent literature to elucidate these challenges, categorize emerging risks, and assess current evaluation and governance mechanisms. Our findings highlight the urgent need for interdisciplinary frameworks that embed ethical principles throughout the LLM lifecycle to ensure trustworthy and equitable AI integration in health care.

BACKGROUND:Invasive fungal disease (IFD) is characterized by its capacity to rapidly escalate to life-threatening conditions, even when patients are hospitalized. However, the precise prognostic significance of baseline clinical characteristics related to the progression outcome of IFD remains elusive. METHODS:A retrospective cohort study spanning a duration of 10 years was conducted at two prominent tertiary teaching hospitals in Southern China. Patients with proven IFD were queried and divided into serious and non-serious groups based on the disease deterioration. To establish robust predictive models, patients from the first hospital were randomly assigned to either a training set or an internal validation set, while patients from the second hospital constituted an external test set. To analyze the potential predictors of IFD deterioration and identify independent predictors, the study employed the least absolute shrinkage and selection operator (LASSO) method in conjunction with binary logistic regressions. Based on the outcomes of this analysis, a predictive nomogram was constructed. The performance of the developed model was thoroughly evaluated using the training set, internal validation set, and external test set. RESULTS:A total of 480 cases from the first hospital and 256 cases from the second hospital were included in the study. Among the 480 patients, 81 cases (16.9%) experienced deterioration, and out of those, 45 (55.6%) cases resulted in mortality. Seven independent predictors were identified and utilized to construct a predictive nomogram. The nomogram exhibited excellent predictive performance in all three sets: the training set, internal validation set, and external test set. The area under the receiver operating characteristic curve (AUC) for the training set was 0.88, for the internal validation set was 0.91, and for the external test set was 0.90. The Hosmer-Lemeshow test and Brier score indicated a high goodness of fit for the model. Furthermore, the calibration curve demonstrated a strong agreement between the predicted outcomes from the nomogram and the actual observations. Additionally, the decision curve analysis exhibited that the nomogram provided significant clinical net benefits in predicting IFD deterioration. CONCLUSIONS:The study successfully identified seven independent predictors and developed a predictive nomogram for early assessment of the likelihood of IFD deterioration.

Alzheimer's disease (AD) is a progressive neurodegenerative disease which is continually increasing in prevalence and is attracting more and more attention. Based on the traditional diagnostic framework, imaging biomarkers have shown great prospect in diagnosis. Positron emission tomography (PET) imaging, as a novel biomarker of AD, evaluates the progress and changes at the molecular level and develops many radiotracers corresponding to the hallmark biological targets. Compounds labeled with radioactive elements are transported to specific regions or combine with specific substances such as amyloid β(Aβ), paired helical filaments (PHFs) and neurofibrillary tangles (NFTs), which makes different radioactive uptake in different brain regions of AD. This review will set forth 18F-FDG PET imaging, Aβ-PET imaging, Tau-PET imaging, neuroinflammatory PET imaging, neurotransmitter PET imaging and some other emerging PET imaging. In clinical practices, PET performed with other medical imaging tools shows a great prospect.

This systematic review and meta-analysis aimed to assess the effectiveness of home-based programmes to prevent hospital admissions compared with traditional hospital-based care for older adults. Health outcomes analysed included readmission rates, mortality, and length of treatment. Data from 15 studies were synthesised using Review Manager (version 5.4), and heterogeneity was assessed using forest plots and I2statistics. Subgroup analyses were performed for randomised controlled trials and for specific patient groups, such as those with cardiovascular and respiratory disease. The results suggest that hospital at home programmes may reduce the risk of readmission (risk ratio = 0.76, 95 % CI 0.58 to 1.01, P = 0.05), especially for patients with respiratory diseases (risk ratio = 0.53, 95 % CI 0.39 to 0.73, P = 0.00007), with no significant differences in mortality or treatment duration between groups.

Current risk assessment models for predicting ischemic stroke (IS) in patients with atrial fibrillation (AF) often fail to account for the effects of medications and the complex interactions between drugs, proteins, and diseases. We developed an interpretable deep learning model, the AF-Biological-IS-Path (ABioSPath), to predict one-year IS risk in AF patients by integrating drug-protein-disease pathways with real-world clinical data. Using a heterogeneous multilayer network, ABioSPath identifies mechanisms of drug actions and the propagation of comorbid diseases. By combining mechanistic pathways with patient-specific characteristics, the model provides individualized IS risk assessments and identifies potential molecular pathways involved. We utilized the electronic health record data from 7859 AF patients, collected between January 2008 and December 2009 across 43 hospitals in Hong Kong. ABioSPath outperformed baseline models in all evaluation metrics, achieving an AUROC of 0.7815 (95% CI: 0.7346-0.8283), a positive predictive value of 0.430, a negative predictive value of 0.870, a sensitivity of 0.500, a specificity of 0.885, an average precision of 0.409, and a Brier score of 0.195. Cohort-level analysis identified key proteins, such as CRP, REN, and PTGS2, within the most common pathways. Individual-level analysis further highlighted the importance of PIK3/Akt and cytokine and chemokine signaling pathways and identified IS risks associated with less-studied drugs like prochlorperazine maleate. ABioSPath offers a robust, data-driven approach for IS risk prediction, requiring only routinely collected clinical data without the need for costly biomarkers. Beyond IS, the model has potential applications in screening risks for other diseases, enhancing patient care, and providing insights for drug development.

Background:Cognitive impairment, indicative of Alzheimer disease and other forms of dementia, significantly deteriorates the quality of life of older adult populations and imposes considerable burdens on families and health care systems worldwide. The early identification of individuals at risk for cognitive impairment through a convenient and rapid method is crucial for the timely implementation of interventions. Objective:The objective of this study was to explore the application of machine learning (ML) to integrate blood biomarkers, life behaviors, and disease history to predict the decline in cognitive function. Methods:This approach uses data from the Chinese Longitudinal Healthy Longevity Survey. A total of 2688 participants aged 65 years or older from the 2008-2009, 2011-2012, and 2014 Chinese Longitudinal Healthy Longevity Survey waves were included, with cognitive impairment defined as a Mini-Mental State Examination (MMSE) score below 18. The dataset was divided into a training set (n=1331), an internal test set (n=333), and a prospective validation set (n=1024). Participants with a baseline MMSE score of less than 18 were excluded from the cohort to ensure a more accurate assessment of cognitive function. We developed ML models that integrate demographic information, health behaviors, disease history, and blood biomarkers to predict cognitive function at the 3-year follow-up point, specifically identifying individuals who are at risk of experiencing significant declines in cognitive function by that time. Specifically, the models aimed to identify individuals who would experience a significant decline in their MMSE scores (less than 18) by the end of the follow-up period. The performance of these models was evaluated using metrics including accuracy, sensitivity, and the area under the receiver operating characteristic curve. Results:All ML models outperformed the MMSE alone. The balanced random forest achieved the highest accuracy (88.5% in the internal test set and 88.7% in the prospective validation set), albeit with a lower sensitivity, while logistic regression recorded the highest sensitivity. SHAP (Shapley Additive Explanations) analysis identified instrumental activities of daily living, age, and baseline MMSE scores as the most influential predictors for cognitive impairment. Conclusions:The incorporation of blood biomarkers, along with demographic, life behavior, and disease history into ML models offers a convenient, rapid, and accurate approach for the early identification of older adult individuals at risk of cognitive impairment. This method presents a valuable tool for health care professionals to facilitate timely interventions and underscores the importance of integrating diverse data types in predictive health models.

BackgroundOccupational burnout is a type of psychological syndrome. It can lead to serious mental and physical disorders if not treated in time. However, individuals tend to conceal their genuine feelings of occupational burnout because such disclosures may elicit bias from superiors. This study aims to explore a novel method for estimating occupational burnout by elucidating its links with social, lifestyle, and health status factors.MethodsIn this study 5,794 participants were included. Associations between occupational burnout and a set of features from a survey was analyzed using Chi-squared test and Wilcoxon rank sum test. Variables that are significantly related to occupational burnout were grouped into four categories: demographic, work-related, health status, and lifestyle. Then, from a network science perspective, we inferred the colleague’s social network of all participants based on these variables. In this inferred social network, an exponential random graph model (ERGM) was used to analyze how occupational burnout may affect the edge in the network.ResultsFor demographic variables, age (p < 0.01) and educational background (p < 0.01) were significantly associated with occupational burnout. For work-related variables, type of position (p < 0.01) was a significant factor as well. For health and chronic diseases variables, self-rated health status, hospitalization history in the last 3 years, arthritis, cardiovascular diseases, high blood lipid, breast diseases, and other chronic diseases were all associated with occupational burnout significantly (p < 0.01). Breakfast frequency, dairy consumption, salt-limiting tool usage, oil-limiting tool usage, vegetable consumption, pedometer (step counter) usage, consuming various types of food (in the previous year), fresh fruit and vegetable consumption (in the previous year), physical exercise participation (in the previous year), limit salt consumption, limit oil consumption, and maintain weight were also significant factors (p < 0.01). Based on the inferred social network among all airport workers, ERGM showed that if two employees were both in the same occupational burnout status, they were more likely to share an edge (p < 0.0001).LimitationThe major limitation of this work is that the social network for occupational burnout ERGM analysis was inferred based on associated factors, such as demographics, work-related conditions, health and chronic diseases, and behaviors. Though these factors have been proven to be associated with occupational burnout, the results inferred by this social network cannot be warranted for accuracy.ConclusionThis work demonstrated the feasibility of identifying people at risk of occupational burnout through an inferred colleague’s social network. Encouraging staff with lower occupational burnout status to communicate with others may reduce the risk of burnout for other staff in the network.

Abstract Background The global population of adults aged 60 and above surpassed 1 billion in 2020, constituting 13.5% of the global populace. Projections indicate a rise to 2.1 billion by 2050. While Hospital-at-Home (HaH) programs have emerged as a promising alternative to traditional routine hospital care, showing initial benefits in metrics such as lower mortality rates, reduced readmission rates, shorter treatment durations, and improved mental and functional status among older individuals, the robustness and magnitude of these effects relative to conventional hospital settings call for further validation through a comprehensive meta-analysis. Methods A comprehensive literature search was executed during April–June 2023, across PubMed, MEDLINE, Embase, Web of Science, and Cumulative Index of Nursing and Allied Health Literature (CINAHL) to include both RCT and non-RCT HaH studies. Statistical analyses were conducted using Review Manager (version 5.4), with Forest plots and I 2 statistics employed to detect inter-study heterogeneity. For I 2 > 50%, indicative of substantial heterogeneity among the included studies, we employed the random-effects model to account for the variability. For I 2 ≤ 50%, we used the fixed effects model. Subgroup analyses were conducted in patients with different health conditions, including cancer, acute medical conditions, chronic medical conditions, orthopedic issues, and medically complex conditions. Results Fifteen trials were included in this systematic review, including 7 RCTs and 8 non-RCTs. Outcome measures include mortality, readmission rates, treatment duration, functional status (measured by the Barthel index), and mental status (measured by MMSE). Results suggest that early discharge HaH is linked to decreased mortality, albeit supported by low-certainty evidence across 13 studies. It also shortens the length of treatment, corroborated by seven trials. However, its impact on readmission rates and mental status remains inconclusive, supported by nine and two trials respectively. Functional status, gauged by the Barthel index, indicated potential decline with early discharge HaH, according to four trials. Subgroup analyses reveal similar trends. Conclusions While early discharge HaH shows promise in specific metrics like mortality and treatment duration, its utility is ambiguous in the contexts of readmission, mental status, and functional status, necessitating cautious interpretation of findings.

This review captured how digital strategies support social network approaches to promote HIV testing. Overall, 29 studies were identified by searching PubMed and Embase for studies published up to June 2023. Existing studies revealed three types of digital strategies (social media (n = 28), online information channels (n = 4), and multifunctional digital platforms (n = 4)) split into four major modes of digital strategy-supported social-network-based HIV testing promotion: 1) Online outreach and recruiting, 2) gathering and identifying key populations for HIV testing, 3) communicating and disseminating online HIV testing health interventions, and 4) assisting and facilitating HIV testing uptake and distribution. Social network approaches supported by digital strategies yielded advantages in HIV testing education and distribution, which increases HIV testing coverage among key populations. Studies are needed on how to facilitate the use of digital strategies for social network-based HIV testing, as well as how to integrate them with existing HIV testing approaches.

This study investigates the predictive utility of Google search queries for forecasting influenza-like illness (ILI) in compulsory education schools in Macau. The increasing availability of online data offers a novel approach to health surveillance, potentially improving the timeliness and accuracy of ILI predictions in educational settings. We employed three machine learning models: extreme gradient boosting (XGBoost), least absolute shrinkage and selection operator (LASSO), and ridge regression (Ridge), to forecast the ILI-caused absence rate in kindergartens, primary schools, and middle schools one and two weeks in advance in Macau. The covariates for these models include Google search queries and historical ILI data. This is Macau's first study to apply machine learning methodologies and Internet big data to the surveillance of ILI in educational institutions. Our approach offers a feasible and computationally efficient method for forecasting ILI in Macau's compulsory education schools. This methodology could be adapted for use in other regions with limited influenza data resources, providing a valuable tool for public health planning and response.