DIGITAL HEALTH最新文献

Objectives: Electronic medical records (EMRs) are increasingly recognized as a contributing factor to patient safety incidents. Clinicians' experiences can reveal EMR-related risks that may otherwise go unnoticed. This study explores EMR-related patient safety incidents reported by physicians across diverse care settings, institutions, and EMR products.

Methods: A national sample of Swiss physicians was surveyed online and asked whether they had experienced a patient safety incident related to EMR use within the previous four weeks. Free-text descriptions of incidents were analyzed thematically using a structured, multi-step procedure.

Results: Of the 1933 inpatient and outpatient physicians who completed the survey, 23.9% (n = 398) reported experiencing an EMR-related safety incident in the previous four weeks. Half of these incidents (49.7%) had not been formally reported (e.g. through critical incident reporting or IT channels). A total of 385 incident descriptions were analyzed, revealing seven emergent themes: (1) patient identification and selection errors (16.7%), (2) system reliability and performance issues (15.8%), (3) interoperability and system integration (8.8%), (4) usability, interface, and design problems (21.8%), (5) system errors and unexpected behavior (8.8%), (6) security and access control (2.6%), and (7) problems with order entry, decision support, alerting, and verification (25.2%). There were considerable differences in the patterns of events reported in relation to the used EMR system.

Conclusions: Physicians reported a broad range of EMR-related safety problems, particularly related to ordering functionalities and usability, many of which were not formally recorded. In addition to broader socio-technical strategies, such as user training, incident reporting, and alignment with clinical workflows, systematically incorporating clinicians' experiences into EMR design is required to guide advancements in patient safety.

Objective: Depression represents a significant global health challenge, further complicated by the multifaceted and complex nature of its diagnosis and treatment. This study explores the application of multiple feature selection (FS) methodologies combined with XAI (explainable artificial intelligence) method named SHapley Additive exPlanations (SHAP) to enhance predictive accuracy in depression classification models using large-scale national survey data.

Methods: Leveraging microdata from the National Mental Health Survey of Korea (2021), encompassing 5511 Korean adults, this research systematically evaluates how different FS-machine learning classifier combinations affect model performance and identifies nondiagnostic socioeconomic, psychological, and lifestyle factors associated with clinically diagnosed depression. By employing diverse FS methods (e.g., ReliefF, Markov Blanket, and Information Gain) across multiple machine learning classifiers, we systematically compare their performance across 12 classifiers.

Results: We demonstrate that optimal FS method selection depends on machine learning classifier architecture, with ReliefF excelling in Stacking (F2-score =0.9851) and Markov Blanket performing best in ExtraTrees and LightGBM (F2-score =0.9848, 0.9838). After excluding core diagnostic criteria variables to avoid circularity, our analysis reveals that social distress (loneliness), reluctance to seek professional help, quality of life measures, and physical health comorbidities emerge as highly influential nondiagnostic predictors.

Conclusion: Our findings advance the field by: (1) systematically demonstrating that FS method effectiveness varies by machine learning classifier type, (2) providing a dual-layer XAI framework combining FS with SHAP for comprehensive interpretability, and (3) identifying culturally specific risk factors in an underrepresented Asian population using high-quality face-to-face collected data. These contributions provide methodological guidance for researchers developing interpretable depression prediction models and offer clinically actionable insights for identifying at-risk individuals in Korean populations.

Digital technologies, such as mHealth interventions and integrated data management tools, are increasingly being developed and implemented to support patients and health care providers in low-resource, high tuberculosis (TB)-burden countries in initiating and proceeding through the TB care cascade (e.g., screening, testing, diagnosis, treatment). Yet, given the proliferation of these tools, there exists a need to synthesize what technologies are being used and where, as well as build a comprehensive understanding of their respective functionality and implementation considerations. The objectives of this systematic scoping review were: (1) to systematically identify literature on digital technologies for supporting the TB cascade in high TB-burden countries; and (2) to describe the facilitators and barriers to technology implementation. Four databases were systematically searched for published literature using a search hedge of terms related to TB, technology, and implementation. Two independent reviewers conducted screening of retrieved literature, data extraction, and data analysis. Eighteen digital technologies were identified, with 10 classified as backbone technologies and eight as add-in technologies. Three key implementation domains were identified: (1) Interoperability and Integration, (2) Digital Infrastructure, and (3) User Experience. Backbone technologies showed higher integration rates with National TB Programs and were more likely to be sustainably implemented. Key barriers to technology implementation included connectivity issues, inadequate user training, and complex multistakeholder integration processes. Included sources described how implementation success was influenced by the interplay between systems-level, technology-level, and user-level factors. Future research should prioritize implementation science approaches to facilitate technology adoption and use to support the TB care cascade.

Objective: Person-centred communication in memory clinics is essential, but often not optimal. This study aimed to develop a solution that supports people with cognitive complaints in expressing their needs and preferences during memory clinic consultations.

Methods: Following a human-centred design approach, co-researchers (n = 4 people with dementia) identified a problem statement. This problem was confirmed and elaborated upon via a questionnaire (n = 25) and focus group (n = 18) for triangulation purposes, and in co-design sessions with people with cognitive complaints (n = 3), care partners (n = 2), and clinicians (n = 3). These sessions informed prototype development in collaboration with a design agency. Usability and User eXperience (UX) testing were conducted with people with cognitive complaints (n = 30), care partners (n = 4), and clinicians (n = 17) via think-aloud sessions, interviews, questionnaires, and focus groups.

Results: Co-researchers emphasized the importance of clinicians gaining a holistic understanding of someone's life and circumstances, which was confirmed in the 'triangulation' questionnaire, focus group, and co-design sessions. Co-design resulted in a digital and analogue prototype of 'Helder in Gesprek' ('Clear in Conversation'), a tool to assist people with cognitive complaints in reflecting on what they wish to share with their clinician and facilitate communication during consultations. Usability testing revealed a generally positive attitude toward the prototypes, while also identifying areas for improvement, such as navigation, system feedback, understandability, distinguishable elements, and cognitive overload.

Conclusion: Our human-centred design approach informed the design and development of two prototypes of 'Helder in Gesprek'. Usability and UX testing provide directions for re-design and feasibility testing in a real-world setting.

Background: While the link between internet use and depressive symptoms in older adults is studied, research often overlooks the interdependent nature of couples. This study examines the longitudinal actor and partner effects of internet use on depressive symptoms among older couples, testing social participation as a key mediating mechanism.

Methods: Using a multistage, stratified probability sampling method, data were drawn from 4878 heterosexual married couples participating in the 2013, 2015, and 2018 waves of the China Health and Retirement Longitudinal Study. A longitudinal dyadic analysis was conducted using structural equation modeling to test an Actor-Partner Interdependence Mediation Model.

Results: For both husbands and wives, their own internet use was associated with lower depressive symptoms, a relationship fully mediated by their own increased social participation (actor-actor effects). Crucially, significant asymmetric partner effects emerged. A husband's internet use was associated with a substantial reduction in his wife's depressive symptoms (β = -0.959, p = .039), indicating a practically meaningful protective effect. This benefit operated both directly and indirectly by increasing the wife's social participation (β = -0.072, p = .026). However, a wife's internet use had no significant effect on her husband's depression.

Conclusions: The mental health benefits of digital engagement extend beyond the individual user to their spouse, operating through enhanced social participation. These findings underscore the importance of dyadic, gender-sensitive approaches when developing interventions to promote digital literacy and social engagement to improve well-being in later life.

Objective: To develop a clinical nursing decision support system for pressure injury and explore its application in managing pressure injury in postoperative cardiac surgery patients.

Methods: A multidisciplinary research team was formed to develop a clinical nursing decision support system. Key indicators, including wound assessment accuracy, wound treatment accuracy, pressure injury healing rate, pressure injury incidence, and defect rates in nursing records, were compared before and after the clinical nursing decision support system utilization. Count data were described using frequency and composition ratio (%), and comparisons were made using the chi-square test or Fisher's exact probability method. Measurement data following a normal distribution were described by mean and standard deviation, while non-normally distributed data were described by median and interquartile range. Independent sample t-tests and rank-sum tests were used for between-group comparisons. A significance level of α = 0.05 was set, with results considered statistically significant if P < 0.05.

Results: The clinical nursing decision support system implements an intelligent decision-making engine and interactive dashboard for human-computer interaction, enabling intelligent assessment and decision-making, re-evaluation reminders, interactive modules, intelligent auditing, and a three-level quality control system for pressure injury. After applying the clinical nursing decision support system, the pressure injury incidence in postoperative cardiac surgery patients decreased from 14.8% to 12.8%, with no statistically significant difference (P > 0.05). The pressure injury healing rate increased from 89.1% to 97.2%, wound assessment accuracy improved from 90.8% to 97.2%, and wound treatment accuracy increased from 88.3% to 96.5%. The defect rate in nursing records decreased from 15.3% to 7.7%, with all differences being statistically significant (P < 0.05).

Conclusion: This study successfully developed and implemented a clinical nursing decision support system for pressure injury management in postoperative cardiac surgery patients. These results confirm the system's clinical utility in standardizing pressure injury care, optimizing nursing workflows, and elevating documentation quality. The clinical nursing decision support system provides an effective tool for enabling evidence-based, personalized interventions and strengthening closed-loop quality control in pressure injury management.

Background: Early detection of diabetic foot complications is essential to prevent ulcers and amputations. Thermographic imaging offers a non-invasive method for identifying risk, but clinical interpretation has traditionally relied on human thermographers. Artificial intelligence (AI) may offer a more scalable and objective alternative.

Objective: To evaluate the diagnostic performance of an AI-powered thermographic screening tool in identifying risk for diabetic foot complications, compared to nurse-led clinical assessment.

Methods: We conducted a cross-sectional study of 100 adults with diabetes undergoing routine foot screening. For each participant, a smartphone-based thermal imaging device was first used to capture plantar images, from which the AI model generated risk scores (0-3). Second, a diabetic foot nurse performed a clinical examination and assigned the reference risk scores (0-3). Absolute temperature differences were computed from thermal images, and diagnostic accuracy metrics were calculated using the nurse assessment as the reference standard.

Results: The AI system demonstrated 100% sensitivity, 96.8% specificity, 66.7% positive predictive value, and 100% negative predictive value for detecting moderate-to-high risk cases. There was a strong correlation between AI and nurse scores (ρ = 0.973), and both assessors showed increasing temperature asymmetry with higher risk levels.

Conclusions: The AI model accurately detected all moderate-to-high risk cases flagged by the nurse, with high sensitivity and specificity. Its strong alignment with thermal data and consistent scoring suggest its value as a scalable and reproducible adjunct for diabetic foot screening. Further validation in longitudinal settings may support broader integration in remote and primary care environments.

Objectives: This study is the first to develop and evaluate a machine learning (ML) model for predicting pelvic adhesions based on ultrasound features, utilizing the SHapley Additive Explanations (SHAP) framework for interpretability analysis.

Methods: This prospective study included 220 patients who underwent laparoscopic surgery and preoperative ultrasound assessments at our hospital between April 2023 and June 2024. Patients were randomly assigned to training and validation sets. A Least Absolute Shrinkage and Selection Operator regression was used to identify independent risk factors, followed by incorporation into an Extreme Gradient Boosting prediction model. Model performance was evaluated using receiver operating characteristic (ROC) curves, calibration plots, and a decision curve analysis.

Results: The included patients were randomly divided into a training set and a validation set in a 7:3 ratio. The final model included four predictors-obstructed ovarian activity, surgical history, endometriosis, and gynecological inflammation-and demonstrated strong discriminatory performance, with an area under the ROC curve of 0.869 and 0.846 in the training and validation sets, respectively. The ML model demonstrated a sensitivity of 0.946 and a specificity of 0.597 in the training set, while in the validation set, it achieved a sensitivity of 1.000 and a specificity of 0.600. Calibration analyses showed good agreement between predicted and observed outcomes. The model exhibited high clinical utility. SHAP analysis revealed that endometriosis contributed most significantly to the predictions, followed by surgical history, obstructed ovarian activity, and gynecological inflammation.

Conclusions: The interpretable ML model developed in this study demonstrates strong predictive performance for assessing the risk of pelvic adhesions in patients prior to surgery. It can be utilized to accurately identify high-risk patients before the procedure, enabling the implementation of appropriate measures during surgery to reduce the occurrence of postoperative pelvic adhesions.

Objective: This mixed-methods investigation examined relationships between digital health literacy and self-efficacy among older adults, focusing on mediating pathways involving social support and life satisfaction and the moderating effect of health consciousness.

Methods: Quantitative data from 1016 community-dwelling older adults across six Chinese provinces were analyzed using structural equation modeling with bias-corrected bootstrapping procedures. A multi-stage probability sampling strategy ensured geographic and socioeconomic diversity. Complementary in-depth semi-structured interviews with 30 purposively selected participants were conducted to explore underlying mechanisms.

Results: Digital health literacy was significantly and directly associated with self-efficacy (β = 0.21, p < .05) and indirectly associated through social support (β = 0.10, 95% CI [0.06, 0.14]) and life satisfaction (β = 0.17, 95% CI [0.12, 0.22]). A sequential mediation pathway was identified whereby digital health literacy was related to social support, which in turn was associated with life satisfaction and subsequently with self-efficacy (β = 0.05, 95% CI [0.01, 0.09]). Health consciousness significantly moderated these relationships, with stronger associations among participants with high health consciousness (direct path: β = 0.32 vs. β = 0.09; sequential indirect path: β = 0.06 vs. β = 0.01). Integrated qualitative analysis revealed information-processing strategies, social validation processes, and goal-directed feedback loops that helped translate digital competencies into psychological resources.

Conclusion: Findings clarify mechanisms through which technological competencies relate to psychological resources in later life. Digital health interventions for older adults should therefore integrate social components, emphasize feedback systems that strengthen efficacy beliefs, and adapt content based on individual levels of health consciousness to optimize psychological benefits.

Uganda has rapidly digitised many health services, but persistent challenges in data governance - including fragmented systems, variable data quality, and the exclusion of vulnerable populations - hinder effective care and equity. This analysis reviews recent developments (2023-2025) in Uganda's digital health policy and practice, drawing on strategy documents, conference reports, and stakeholder input. It highlights how the COVID-19 pandemic accelerated innovation while exposing systemic weaknesses. For example, the Ministry of Health's (MoH) 2023 strategy explicitly targets data accessibility and integration, and the 2024 guidelines standardise management across the sector. Yet, execution gaps remain due to resource constraints and organisational silos. This article proposes an inclusive data governance framework with five pillars (inclusive governance, equity, interoperability, privacy, and capacity) and recommends concrete actions. By adopting these reforms, Uganda can transform its digital health systems into people-centred, equitable platforms that build trust, protect rights, and advance universal health coverage.