搜索结果：PLOS digital health

Despite growing investment in digital health for immunization, there is limited evidence on how gender is considered into design and implementation. In 2022, Ethiopia launched a Gender-intentional Digital Health Information Roadmap with the support of Gavi, the Vaccine Alliance; however, the operationalization of gender strategies remains unclear. This qualitative study explored gender inclusiveness in digital health for immunization in Ethiopia, drawing on document reviews, participatory workshops, key informant interviews, and stakeholder analysis across selected three regions in Ethiopia. Findings revealed that women dominate frontline roles as caregivers and vaccinators but remain underrepresented in supervisory, leadership, and digital system development positions. Although gender strategies and units exist, implementation is weak, with limited gender awareness and no intentional planning. In addition, a notable gender digital divide persists. Senior leadership role assignments, despite competence, depend on ability for women to balance responsibilities and travel with household duties. The routine health data systems, like District Health Information Software, version 2, lack sex-disaggregated immunization data, hindering gender-informed decision making. Enabling factors include national prioritization of gender equity, empowerment programs, and partner support, whereas barriers include cultural norms, inadequate advocacy, weak accountability, and insufficient resources. Embedding gender-transformative approaches is vital for equitable, inclusive digital health information and associated health system and immunization outcomes in Ethiopia.

Guided digital interventions in maternity and child health care may improve mental health outcomes by increasing user engagement and adherence. However, implementing such complex interventions remains challenging due to factors related to intervention characteristics, organizational dynamics, professionals' ompetencies, and user engagement. Understanding the experiences of those who deliver these interventions is essential to inform implementation efforts. This study explored midwives' and public health nurses' experiences of a guided digital intervention and the early implementation process three months post-training. Using a qualitative design, six focus groups were conducted with 32 professionals working in maternity and child health care services and participating in a multisite cluster-randomized trial of the universal digital intervention, Mamma Mia. Interviews were based on the Consolidated Framework for Implementation Research, and data were analyzed using reflexive thematic analysis. Three themes described how professionals made sense of implementing a guided digital intervention. "Broken Expectations - Navigating the Training-Practice Gap" reflected how abstract, decontextualized training left professionals' feeling unprepared once faced with real-world implementation. "Balancing Belief and Doubt - Navigating digital care in relational professions" captured tensions between valuing innovation and safeguarding professional values, as participants negotiated whether digital tools could align with relational care practices. "Learning by Doing and Support - Future Optimism" illustrated a gradual shift toward acceptance, as hands-on experience and collegial support helped professionals view the intervention as complementary rather than disruptive. These findings underscore that implementing digital interventions in relational care settings is not solely a technical process but a relational and interpretive one. Effective implementation should therefore support contextual adaptation, provide practical and collaborative training, foster trust, and ensure organisational alignment.

According to the World Health Organization, air pollution is the largest environmental risk to global health, with 99% of the world's population living in areas exceeding recommended guidelines. Providing real-time air quality information through mobile or web-based applications, alongside behavioural guidance, represents a key strategy for reducing individual exposure and improving population health. Such information is delivered at varying geographic and temporal scales and has become increasingly widespread and decentralised. However, there is limited understanding of the quality, characteristics, and potential health impacts of digitally available air quality information and messaging. This study addresses this gap through a systematic technology review of publicly available digital platforms ("channels") that share real-time local air pollution data. Using the UK as a case study, computational methods were applied to examine how data underpinning existing channels (websites, mobile applications, sensors, etc.) are generated and by whom. Systematic searches of Google, the Google Play Store (Android), and Apple App Store (iOS) were conducted using SerpAPI and predefined search terms. In total, 146 channels met inclusion criteria and were analysed. Channel metadata were used to identify trends in channel types, evolution over time, and emerging patterns across the digital information landscape. The review also compares differences in available information across channels but is limited to those accessible in the UK and excludes real-time social media data due to cost and access constraints. Finally, the study considers demographic and social factors influencing access to air quality information and its exposure-reducing benefits. These findings contribute to understanding the production and use of digital air quality information with global public health relevance.

Digitalisation offers promising solutions for improving efficiency and accessibility in healthcare. However, it is crucial to balance digital advancements with humanisation. The aim of this study is to support digital transformation in healthcare by striking a balance between the digital and human components, by proposing a roadmap for digital transformation, prioritizing people and organisational culture. An initial theoretical and contextual investigation guided the preliminary roadmap, addressing concerns and solutions from the literature. Then, an empirical study was conducted via interviews with Portuguese hospital managers, offering practical insights into initiatives for hospital digitalisation. These findings refined the roadmap with real-world perspectives. The roadmap is presented as a theoretically informed and empirically refined planning proposal; it has not yet been piloted in real-world settings or compared with existing implementation frameworks. Managers highlighted the human dimension as critical to successful digital transformation, reinforcing the need for such a roadmap, an integrated approach that considers technology, processes and people within organisational culture. Humanisation is treated as a measurable target, monitored through patient-experience and staff-experience indicators. Theoretically, this study advances the literature on change management and digital transformation in healthcare. Practically, it offers an exploratory roadmap to guide healthcare organisations in structuring, sequencing and monitoring digital transformation initiatives. Transferability to other contexts requires local tailoring and future piloting with broader stakeholder involvement.

Up to 40% of all cancers diagnosed could be prevented through the uptake of healthy lifestyle behaviours. This paper describes the co-design of a novel community-centric digital programme for cancer prevention and early detection support. Twenty-two community panel members partook in a series (i.e., three group-based sessions followed by one individual session) of iterative online workshops to co-design the programme. Based on established Design Thinking principles, the workshops aimed to i) identify barriers to access and use of existing cancer prevention information and support, ii) explore consumers' needs and preferences for information and support related to cancer risk reduction and early detection, iii) generate ideas for a digital solution to address those, and iv) seek panel members' feedback on a digital prototype developed in line with the insights from previous stages. The key barriers to accessing cancer prevention and screening support identified by the panel included limited availability and knowledge of cancer-specific resources alongside a paradoxical information overload regarding generic options. Panel members expressed concerns about information accuracy and relevance, while time constraints, financial limitations, and motivation deficits further impeded resource engagement. Regarding prototype preferences, participants prioritised accessibility, diversity, communication and connectedness, usability, and personal relevance as essential design elements. Digital solutions for cancer prevention and screening support should offer a customised experience, while catering for varying practical constraints and motivational challenges.

The spread of medical misinformation through conventional and online communication poses a public health risk, especially in countries like Lebanon, where the healthcare system is fragile. Its dangers were highlighted during the COVID-19 pandemic and other health emergencies, showing how inaccurate or misleading information can alter public health behavior. Despite growing awareness about the problem, no study in Lebanon has explored health misinformation from the perspective of key healthcare stakeholders. The aim of this study was to investigate the views of key stakeholders on the causes and implications of health-related misinformation in Lebanese media, and to provide recommendations to combat it. We conducted a qualitative study using semi-structured, in-depth interviews with nine elite interviewees from the media, health sector, government, professional syndicates, and legal system. Interviews were conducted between March 17 and 25, 2025 and thematically analyzed using an inductive-deductive approach. Data saturation was reached at the eighth interview. Trustworthiness was ascertained through triangulated coding, reflexivity, and adherence to established qualitative rigor criteria. Factors contributing to the spread of health misinformation included low health literacy, poverty, political pressure on the media, and a lack of regulations on digital platforms. WhatsApp, Facebook, and Instagram were identified as major channels for rapid dissemination. Although some Lebanese media outlets practiced voluntary fact-checking and consulted experts, these efforts remained intermittent and unenforced. Government responses were described as reactive, relying more on public credibility than legislation. Existing health legislation was seen as outdated, weakly enforced, and disconnected from digital realities. The COVID-19 pandemic exposed these vulnerabilities and highlighted the need for coordinated broadcasting approach and stronger expert oversight across media platforms. Medical misinformation in Lebanon is driven by structural weaknesses across media, health, and legal systems. Addressing it requires coordinated institutional action, improved health literacy, stronger collaboration between media and health professionals, and updated regulations suited to the digital era.

Adolescents living with HIV face numerous psychosocial challenges that increase their vulnerability to mental health problems. However, limited evidence exists on the prevalence and contributing factors among this population in the Teso region of Uganda. This study assessed the prevalence of mental health problems and associated factors among adolescents living with HIV in the Teso region. A cross-sectional study was conducted among adolescents aged 10-19 years receiving HIV care in selected high-volume health facilities in the Teso region. Data were collected using the Home, Education/Employment, Activities, Drugs, Sexuality, Suicide/Depression psychosocial assessment tool (HEADSS tool), uploaded onto Kobo Collect for digital data collection. The data were exported to Microsoft Excel, cleaned, and analyzed using STATA version 17. Descriptive statistics were used to summarize sociodemographic characteristics and estimate the prevalence of mental health problems. Logistic regression analysis was used to identify factors associated with mental health problems. The prevalence of mental health problems was 35.2%, with suicidal tendencies being the most common (31%). Factors significantly associated with mental health problems included staying with one parent (AOR = 0.71; 95% CI: 0.55-0.92; p = 0.001) and not working (AOR = 4.4; 95% CI: 1.66-11.62; p = 0.03). Mental health problems are prevalent among adolescents living with HIV in the Teso region. Supportive living arrangements were protective, while employment was associated with increased risk. Peer-led screening can aid early identification, emphasizing the need to integrate mental health services into adolescent HIV care, strengthen family and caregiver support.

Data sharing is essential for modern science, advancing transparency, reproducibility, and discovery. Emerging evidence shows that certain digital health data, particularly high-frequency accelerometry from body-worn sensors, carry re-identification risks even when de-identified. In 2023, the National Institutes of Health published its Data Management and Sharing (DMS) Policy formalizing a commitment to openness by requiring all funded researchers to share scientific data. However, the policy did not anticipate that raw motion signals collected by wearable devices can function as biometric identifiers. The WristPrint study demonstrated that a single day of raw accelerometry data could be used to re-identify individuals with 96% accuracy. Related research in gait detection shows that as few as ten steps may be enough to uniquely identify someone. These findings highlight gaps in current data sharing policies and the need for tailored guidance. We argue that policy updates, enforceable data use agreements, and educational initiatives are essential to align openness with protection. The path forward is not to retreat from data sharing but to share more wisely, safeguarding participant trust while sustaining scientific progress.

Dementia is an emerging public health challenge in low- and middle-income countries (LMICs), yet it remains underdiagnosed in rural Uganda, where primary health care (PHC) providers often lack tools, training, and data systems for early detection and management. Mobile health (mHealth) applications can enhance provider capacity, improve data capture, and strengthen feedback systems. This study explored the perspectives of PHC providers and District Health Teams (DHTs) to inform the co-design of eCap-CoDe, a mobile application for community-based dementia care in rural Uganda. We conducted in-depth interviews with 31 participants from two rural districts: 23 PHC providers (medical/clinical officers and nurses) and 8 DHT members. Participants were purposively sampled for diversity in cadre, experience, and facility type. Data were thematically analyzed using the Consolidated Framework for Implementation Research (CFIR), with four a priori domains, i.e., content, user experience, organizational, and service delivery, guiding coding and analysis. Content requirements - included dementia-specific screening and management tools, modular in-app training aligned with the WHO mhGAP Intervention Guide, and structured data capture integrated with Uganda's Health Management Information System (HMIS). User experience needs: emphasized simple, intuitive interfaces with dropdown menus, checkboxes, audio-visual decision support, and offline functionality to address connectivity gaps. Organizational requirements: prioritized interoperability with District Health Information System 2 (DHIS2), integration with supervisory workflows, and dementia-specific performance indicators. Service delivery needs: focused on real-time feedback loops, reducing duplicate documentation, and potential expansion to other common conditions to enhance utility and uptake. Co-designing mHealth tools with end-users ensures alignment with the realities of workflows, systems, and infrastructure. eCap-CoDe addresses capacity, data, and feedback gaps in rural dementia care and offers a scalable model for integrating digital tools into PHC in Uganda and similar LMICs. Pilot testing will assess the feasibility, usability, and impact before scaling up.

Currently, most research on digital transformation in higher vocational education conducts macro-level analyses, and few comprehensive studies specifically address higher vocational nursing education. Moreover, research is notably lacking on the perceptions of Chinese higher vocational nursing program directors, and the prevalent utilization of digital tools within this field remains underexplored. To understand the perceptions and experiences of higher vocational nursing program directors regarding the use of digital transformation tools in nursing education. This qualitative study employed semistructured interviews conducted with 14 program directors from higher vocational nursing programs. The interview data were transcribed, and Colaizzi's seven-step method was used for a subject analysis. The study employed Colaizzi's seven-step framework for inductive thematic analysis, with coding conducted collaboratively until data saturation was achieved. Analysis revealed four principal themes: (1) Recognition of digital education by program directors, where enthusiasm for enhanced engagement and transcended geographical barriers coexisted with unspoken doubts about resource quality and the authenticity of student participation; (2) Perceived status and impact of digital infrastructure, characterized by significant hardware investments yet inconsistent resource quality and utilization-revealing a fundamental gap between technological abundance and meaningful pedagogical integration; (3) Systemic Tensions and Barriers, exposing profound contradictions between policy aspirations and classroom realities, generational divides in educational philosophies among teachers, and persistent disconnects between technology's promise and its operational friction, all of which contributed to increased workloads and a pervasive sense of professional role dissonance; (4) Strategic Approaches to Digital Transformation, centered on increasing investment, fostering educator development, promoting resource sharing, and strengthening industry-education integration. Higher vocational nursing program directors recognize the value of digital education, yet its adoption is constrained by challenges in platform usability and a lack of supportive institutional incentives. Effective transformation requires systemic strategies that enhance the digital ecosystem's ease of use, align professional rewards with innovation, and promote cross-sector collaboration to develop pedagogically sound resources.

Digital technologies increasingly impact daily life. Digital wellbeing is emerging as a multidimensional construct encompassing emotional regulation, autonomy, social connection, and respectful online interactions. This study aimed to validate the English version of the 13-item Digital Wellbeing Questionnaire (DWBQ) in young adults. Methods: Data were collected via quota sampling on Prolific, an online crowdsourcing platform. The final sample comprised 1,853 young adults (US n = 933, UK n = 920), including 892 women, 871 men, and 90 non-binary participants., aged 18-25. Confirmatory factor analysis (CFA), measurement invariance testing (across gender and country), reliability analysis, and convergent validity assessments (with the Digital Flourishing Scale and the Digital Stress Scale) were conducted. CFA confirmed the four-factor structure (emotional resilience, agency, social connection, and communion), with good model fit (CFI = .969, TLI = .959, RMSEA = .058, SRMR = .038). The DWBQ showed strong internal reliability (α ≥ .80) and measurement invariance across both gender and country. Convergent validity was supported by positive correlations with digital flourishing and negative associations with digital stress. Sociodemographic factors and smartphone use patterns were significantly associated with DWBQ subscales, with gender and relationship status particularly influencing agency and emotional resilience. The English version of the 13-item DWBQ is a valid and reliable measure of digital wellbeing among young adults in the US and UK. It offers a concise tool for research, educational, and clinical use.

During outbreaks women struggle to access essential health services, including services for violence. Services may be disrupted or deprioritised, or women may avoid clinical settings. We conducted a scoping review to understand how health services for violence against women (VAW) were affected in low- and middle-income (LMIC) settings during recent outbreaks, and how women sought help following violence. We reviewed published academic literature reporting on primary research from LMIC settings during recent outbreaks (Ebola, Zika and COVID-19). Four databases were searched: Medline, Embase, Global Health, and Global Index Medicus. Thirty two papers met the inclusion criteria. Data were extracted using a thematic framework focusing on both the supply and demand for services. Experiences during COVID-19 were overrepresented, with no studies identified from other outbreaks. Research spanned 20 countries including a range of services and populations. In the face of lockdowns and reorientation of the health system towards COVID-19, VAW services were restricted or closed despite being essential. Many settings reported shifting services online or to telehealth platforms, raising concerns about digital access and safety, particularly when women accessed services from spaces shared with a violent partner. Some other adaptations included the use of community volunteers and the provision of cash assistance for survivors. Help-seeking varied, with some healthcare settings reporting increases and others decreases in the number of survivors presenting, likely reflecting fluctuating restrictions. Women experiencing violence often sought help from informal sources such as community leaders and family. Help-seeking was further constrained by the economic crisis accompanying COVID-19, including food insecurity and transportation challenges. To prepare for future outbreaks, research is needed to identify which services can be safely and equitably delivered online, and which require in-person provision, as well as to understand a broader range of emerging practices for adapting services to physical distancing, movement restrictions, and economic stress.

Despite Thailand's progress under the 1-3-7 malaria elimination framework, border districts such as Tha Song Yang in Tak Province continue to experience persistent transmission due to high population mobility, geographic constraints, and health system challenges. Understanding how local health systems respond to these pressures is critical for sustaining malaria elimination in complex border settings. This mixed-methods study applied the World Health Organization's Six Building Blocks framework to examine challenges and responses in malaria elimination in Tha Song Yang District. Qualitative data were collected through in-depth interviews with 24 key informants from district health offices, vector-borne disease units, malaria posts and clinics, hospitals, and local authorities. Quantitative data included household surveys assessing malaria-related knowledge, attitudes, and practices (n = 388), and secondary surveillance data on adherence to the 1-3-7 strategy from 2018 to 2022. Adherence to Day 1 and Day 3 activities improved steadily, reaching 96% and 100%, respectively, while Day 7 adherence declined sharply to 69% in 2022. The household survey showed high levels of malaria knowledge (75.77%) and positive attitudes toward prevention (94.59%), indicating that community awareness was strong despite ongoing transmission. Qualitative findings highlighted a surge of short-term migrants, workforce shortages, reduced domestic funding, and logistical barriers in remote areas as key constraints, while adaptive local responses-including local flexibility, committed leadership, and the use of the Malaria Information System and mobile communication platforms-helped sustain malaria control and surveillance activities. These findings demonstrate that malaria persistence in border areas is driven primarily by systemic and operational challenges rather than gaps in community awareness. Strengthening malaria elimination in similar border settings requires resilient health systems with sustained domestic financing, adaptive surveillance strategies, effective use of digital tools, and governance arrangements that account for population mobility and cross-border complexity.

Digital referral platforms can strengthen communication between primary and specialist care and improve continuity for people with COPD. However, adoption in Australian primary care remains uneven across jurisdictions. To explore general practitioners' (GPs) and practice managers' (PMs) perspectives on barriers and enablers to adopting digital referral systems for COPD management in Australia. A qualitative study was conducted with 16 participants (ten general practitioners and six practice managers) from urban, regional, and rural settings across five Australian states and territories. Semi-structured interviews (30-45 minutes) were conducted via Microsoft Teams and analysed using Braun and Clarke's reflexive thematic approach. Inductive coding attended to role and location, and two researchers independently reviewed coding and interpretations to enhance confirmability. The COPD used as an exemplar to examine the barriers, enablers and system gaps associated with digital referral systems. Four themes described current practice and needs. First, disrupted digital workflows: recurrent technical issues, limited interoperability with existing systems, and gaps in training reduced routine use. Second, fragmented communication: referrals often moved in one direction, with poor visibility of status and minimal feedback to primary care. Third, pragmatic enablers: auto-filled templates, transparent triage processes, and waiting time tracking reduced workload and uncertainty. Fourth, aspirations for integration: participants prioritised cross-sector interoperability, inclusive co-design, and real-time two-way messaging to support continuity, accountability, and timely care. Study participants described Australia's digital referral landscape as fragmented, inconsistently adopted, and hindered by weak feedback loops. Usability features that automate data entry and surface referral status show immediate value and may accelerate uptake. Realising system-level benefits will require nationally coordinated policy, minimum interoperability standards, and targeted investment in regional, rural, and under-resourced settings. These practice-informed priorities translate front-line experience into actionable design and policy levers, offering a roadmap for procurement, co-design, and evaluation of digital referral platforms for COPD and other chronic conditions.

Mental health disorders like depression and anxiety pose global challenges, requiring accurate, non-invasive detection methods. Classical modes of diagnosis are typically based on self-reported symptoms or clinical evaluation, which could be subjective and protracted in time. To address these limitations, this study proposes NeuroHAGWO-Net, an advanced artificial intelligence-based framework for automated mental health status detection using multimodal data. The proposed model integrates electroencephalogram (EEG) signals and behavioral textual data to enable early and reliable mental health screening. EEG signals are pre-processed with Empirical Mode Decomposition (EMD) for noise removal, while behavioral text data is transformed into embeddings using Bidirectional Encoder Representations from Transformers (BERT) models. The hybrid BiLSTM-CNN architecture captures temporal dependencies and spatial patterns in EEG data, enhanced by integrating behavioral embeddings for multimodal analysis. Features are selected using a novel Hybrid Ant-Grey Wolf Optimization (HAGWO) approach, combining Ant Colony Optimization (ACO) and Modified Grey Wolf Optimization (mGWO), respectively. The AI-based mental health detection is performed using NeuroVisionNet, integrating EfficientNetV2 and Temporal CNNs (T-CNNs). The model's performance is validated on two datasets: behavioral data and EEG signals data. On behavioral data, it achieves an accuracy of 0.9945, precision of 0.9874, sensitivity of 0.9935, specificity of 0.9915, F1-Score of 0.9909, Matthews Correlation Coefficient (MCC) of 0.9925, Negative Predictive Value (NPV) of 0.9905, False Positive Rate (FPR) of 0.0151, and False Negative Rate (FNR) of 0.0092. With its strong accuracy and efficiency in detecting mental health situations under diverse data modalities, NeuroHAGWO-Net Model proves to be a robust tool for early mental health screening and clinical support using modern optimization techniques and deep learning architectures.

The purpose of this scoping review was to identify, geographically map, and understand what and how novel digital technologies are being used for providing cognitive training and psychosocial interventions that aim at preserving cognitive and everyday function, and improving well-being, in dementia patients, and persons with Mild Cognitive Impairment (MCI). Studies were identified across seven databases (EMBASE, Scopus, CINAHL Complete, PubMed, Dimensions, APA PsychInfo) dating from 2012-Dec 2025, and one hundred and fourteen met the inclusion criteria. We followed the Joanna Briggs Institute guidelines and the framework. Screening was undertaken according to the PRISMA-SCR guidelines and involved 7 reviewers across stages. This review analysed 114 articles and uncovered 8 main categories of digital technology focused on intervention delivery methods that adhered to our inclusion/exclusion criteria. Of those, gamified technology, apps, and web-based approaches were most commonly studied. An emerging technology identified was virtual reality based delivery. There was great heterogeneity in the evidence of effectiveness of the technology-based interventions on cognitive, functional and wellbeing outcomes as reported in the studies included here. There was also great variability across the studies in how the interventions were designed and implemented and what and how many outcomes were measured. This review supports that the majority of the digital interventions for cognitive and psychosocial interventions are being developed and tested for administration under real-time supervision by a trained professional and not for unsupervised home environments. Notwithstanding, technology suitable for home use such as gamified tasks, apps and web/internet-based approaches were found to be the most prominent interfaces for digital interventions. Virtual reality (VR) based interventions, e.g., using simple headsets, must also be considered as intervention tools with high potential for home use mitigating issues of limited mobility through cognitive-physical training (combining physical - exergaming - and cognitive training).

Hand osteoarthritis is a leading cause of pain, disability, and reduced quality of life in older adults. Joint protection programs are recommended as a core component of self-management, but traditional delivery is limited by barriers to access. Digital programs can overcome these challenges for some people, but their reach and effectiveness depend on usability. We conducted a mixed methods usability study of a remotely delivered joint protection program designed for people with hand osteoarthritis. Twenty-three participants took part, recruited through purposeful sampling to ensure inclusion of groups often underrepresented in research. Usability was assessed using predefined task completion, browser-based eye-tracking, participant ratings, and think-aloud protocols, with iterative refinements applied between participants. Routine navigation tasks, such as navigating between different modules, accessing interactive activities, and viewing short videos, were consistently completed with high success. More complex interactive tasks, including drag-and-drop activities, scenario-based modules, and toggling videos to full screen, initially posed challenges. Over successive iterations, however, usability improved markedly, with later participants achieving near-perfect performance. Qualitative analysis revealed that participants valued clear language, short and focused videos, interactive elements, and the ability to proceed at their own pace, while raising concerns about excessive clicking, unclear instructions, and variation in age representation. Iterative refinements, including platform adjustments, clearer instructions and an introductory video, addressed these issues and contributed to improved performance. This study demonstrates that a remotely delivered, technology-enabled joint protection program for hand osteoarthritis is usable, accessible, and engaging across a diverse sample. Beyond refining the program itself, the study introduces a practical framework for iterative, equity-informed usability testing that can inform the design of future digital health interventions.

Large language models (LLMs) have recently gained prominence in healthcare content provision due to their numerous advantages. Despite these benefits, LLMs exhibit notable limitations in this domain. This study aimed to systematically identify the limitations of LLMs in provision of healthcare content. This study was a systematic review conducted in September 2025, including articles published in English between 2018 and 2025. Searches were performed in PubMed, Scopus, and the Cochrane Database of Systematic Reviews. Two independent evaluators screened the references and assessed quality of the selected studies using the Authority, Accuracy, Coverage, Objectivity, Date, and Significance (AACODS) checklist. Data were analyzed using Boyatzis's qualitative thematic approach with an inductive methodology, applying the input-process-output (IPO) model as the analytical framework. A total of 81 studies were included in the final analysis. The included studies were predominantly of high quality and demonstrated minimal risk of bias. The thematic analysis identified key themes: data limitations, dependence on input and prompt quality, accessibility issues, model design and architecture constraints, interaction challenges, response quality and comprehensiveness, and ethical, safety, and regulatory concerns. The study identified multiple limitations of LLMs in healthcare, with output issues being most common. In this regard, the most frequently cited limitation was the accuracy gap. However, these output issues were mainly resulted from flaws in input data, emphasizing the crucial role of input quality. The study also proposed strategies to address these challenges.

Assessing the risk of suicidal outcomes is challenging, particularly in older people. Smartphone-based digital phenotyping may help to monitor suicide risk through ecological momentary assessment (EMA) applications. In this real-world study, we investigated if age and other clinical factors were associated with participation in EMA at baseline, and with retention in EMA monitoring among patients at risk of suicide. Participation in EMA was determined by quantifying the installation of the MEmind mobile application in individuals involved in the SmartCrisis 1.0 and 2.0 studies. The patients were followed-up over a 6-month period. N&#x2009;=&#x2009;512 patients met inclusion criteria, of which 387 installed the MEmind application on their smartphone. While age as a continuous variable was not associated with using EMA at baseline, being aged older than 50 and being engaged in an intimate relationship were independently associated with longer participation in EMA (OR 2.070, 95%CI [1.054-4.066], and OR 2.103, 95%CI [1.076-4.110], respectively). In an exploratory survival analysis, we found that EMA retention increased with age (p&#x2009;<&#x2009;0.001). Feasibility of EMA seems warranted in older people at risk of suicide. Clinicians should be encouraged to offer EMA monitoring to older adults, as they commonly face limitations in their access to healthcare facilities.