Background: As health care continues to evolve with technological advancements, the integration of artificial intelligence into clinical practices has shown promising potential to enhance patient care and operational efficiency. Among the forefront of these innovations are large language models (LLMs), a subset of artificial intelligence designed to understand, generate, and interact with human language at an unprecedented scale. Objective: This systematic review describes the role of LLMs in improving diagnostic accuracy, automating documentation, and advancing specialist education and patient engagement within the field of gastroenterology and gastrointestinal endoscopy. Methods: Core databases including MEDLINE through PubMed, Embase, and Cochrane Central registry were searched using keywords related to LLMs (from inception to April 2024). Studies were included if they satisfied the following criteria: (1) any type of studies that investigated the potential role of LLMs in the field of gastrointestinal endoscopy or gastroenterology, (2) studies published in English, and (3) studies in full-text format. The exclusion criteria were as follows: (1) studies that did not report the potential role of LLMs in the field of gastrointestinal endoscopy or gastroenterology, (2) case reports and review papers, (3) ineligible research objects (eg, animals or basic research), and (4) insufficient data regarding the potential role of LLMs. Risk of Bias in Non-Randomized Studies—of Interventions was used to evaluate the quality of the identified studies. Results: Overall, 21 studies on the potential role of LLMs in gastrointestinal disorders were included in the systematic review, and narrative synthesis was done because of heterogeneity in the specified aims and methodology in each included study. The overall risk of bias was low in 5 studies and moderate in 16 studies. The ability of LLMs to spread general medical information, offer advice for consultations, generate procedure reports automatically, or draw conclusions about the presumptive diagnosis of complex medical illnesses was demonstrated by the systematic review. Despite promising benefits, such as increased efficiency and improved patient outcomes, challenges related to data privacy, accuracy, and interdisciplinary collaboration remain. Conclusions: We highlight the importance of navigating these challenges to fully leverage LLMs in transforming gastrointestinal endoscopy practices. Trial Registration: PROSPERO 581772; https://www.crd.york.ac.uk/prospero/

Background: Previous studies have shown that electrocardiographic (ECG) alarms have high sensitivity and low specificity, have underreported adverse events, and may cause neonatal intensive care unit (NICU) staff fatigue or alarm ignoring. Moreover, prolonged noise stimuli in hospitalized neonates can disrupt neonatal development. Objective: The aim of the study is to conduct a nationwide, multicenter, large-sample cross-sectional survey to identify current practices and investigate the decision-making requirements of health care providers regarding ECG alarms. Methods: We conducted a nationwide, cross-sectional survey of NICU staff working in grade III level A hospitals in 27 Chinese provinces to investigate current clinical practices, perceptions, decision-making processes, and decision-support requirements for clinical ECG alarms. A comparative analysis was conducted on the results using the chi-square, Kruskal-Wallis, or Mann-Whitney U tests. Results: In total, 1019 respondents participated in this study. NICU staff reported experiencing a significant number of nuisance alarms and negative perceptions as well as practices regarding ECG alarms. Compared to nurses, physicians had more negative perceptions. Individuals with higher education levels and job titles had more negative perceptions of alarm systems than those with lower education levels and job titles. The mean difficulty score for decision-making about ECG alarms was 2.96 (SD 0.27) of 5. A total of 62.32% (n=635) respondents reported difficulty in resetting or modifying alarm parameters. Intelligent module–assisted decision support systems were perceived as the most popular form of decision support. Conclusions: This study highlights the negative perceptions and strong decision-making requirements of NICU staff related to ECG alarm handling. Health care policy makers must draw attention to the decision-making requirements and provide adequate decision support in different forms.

Background: Though widely used, resting heart rate (RHR), as measured by a wearable device, has not been previously evaluated in a large cohort against a variety of important baseline characteristics. Objective: This study aimed to assess the validity of the RHR measured by a wearable device compared against the gold standard of ECG (electrocardiography), and assess the relationships between device-measured RHR and a broad range of clinical characteristics. Methods: The Project Baseline Health Study (PHBS) captured detailed demographic, occupational, social, lifestyle, and clinical data to generate a deeply phenotyped cohort. We selected an analysis cohort within it, which included participants who had RHR determined by both ECG and the Verily Study Watch (VSW). We examined the correlation between these simultaneous RHR measures and assessed the relationship between VSW RHR and a range of baseline characteristics, including demographic, clinical, laboratory, and functional assessments. Results: From the overall PBHS cohort (N=2502), 875 (35%) participants entered the analysis cohort (mean age 50.9, SD 16.5 years; n=519, 59% female and n=356, 41% male). The mean and SD of VSW RHR was 66.6 (SD 11.2) beats per minute (bpm) for female participants and 64.4 (SD 12.3) bpm for male participants. There was excellent reliability between the two measures of RHR (ECG and VSW) with an intraclass correlation coefficient of 0.946. On univariate analyses, female and male participants had similar baseline characteristics that trended with higher VSW RHR: lack of health care insurance (both P<.05), higher BMI (both P<.001), higher C-reactive protein (both P<.001), presence of type 2 diabetes mellitus (both P<.001) and higher World Health Organization Disability Assessment Schedule (WHODAS) 2.0 score (both P<.001) were associated with higher RHR. On regression analyses, within each domain of baseline characteristics (demographics and socioeconomic status, medical conditions, vitals, physical function, laboratory assessments, and patient-reported outcomes), different characteristics were associated with VSW RHR in female and male participants. Conclusions: RHR determined by the VSW had an excellent correlation with that determined by ECG. Participants with higher VSW RHR had similar trends in socioeconomic status, medical conditions, vitals, laboratory assessments, physical function, and patient-reported outcomes irrespective of sex. However, within each domain of baseline characteristics, different characteristics were most associated with VSW RHR in female and male participants. Trial Registration: ClinicalTrials.gov NCT03154346; https://clinicaltrials.gov/study/NCT03154346

Background: The maintenance of a healthy lifestyle significantly influences pregnancy outcomes. Certain pregnant women are more at risk of engaging in unhealthy behaviors due to factors such as having a low socioeconomic position and low social capital. eHealth interventions tailored to pregnant women affected by these vulnerability factors can provide support and motivation for healthier choices. However, there is still a lack of insight into how interventions for this target group are best designed, used, and implemented and how vulnerable pregnant women are best reached. Objective: This review aimed to identify the strategies used in the design, reach, use, and implementation phases of eHealth lifestyle interventions for vulnerable pregnant women; assess whether these strategies acted as facilitators; and identify barriers that were encountered. Methods: We conducted a search on MEDLINE, Embase, Web of Science, CINAHL, and Google Scholar for studies that described an eHealth intervention for vulnerable pregnant women focusing on at least one lifestyle component (diet, physical activity, alcohol consumption, smoking, stress, or sleep) and provided information on the design, reach, use, or implementation of the intervention. Results: The literature search identified 3904 records, of which 29 (0.74%) met our inclusion criteria. These 29 articles described 20 eHealth lifestyle interventions, which were primarily delivered through apps and frequently targeted multiple lifestyle components simultaneously. Barriers identified in the design and use phases included financial aspects (eg, budgetary constraints) and technological challenges for the target group (eg, limited internet connectivity). In addition, barriers were encountered in reaching vulnerable pregnant women, including a lack of interest and time constraints among eligible participants and limited support from health care providers. Facilitators identified in the design and use phases included collaborating with the target group and other stakeholders (eg, health care providers), leveraging existing eHealth platforms for modifications or extensions, and adhering to clinical and best practice guidelines and behavior change frameworks. Furthermore, tailoring (eg, matching the content of the intervention to the target groups’ norms and values) and the use of incentives (eg, payments for abstaining from unhealthy behavior) were identified as potential facilitators to eHealth use. Facilitators in the interventions’ reach and implementation phases included stakeholder collaboration and a low workload for the intervention deliverers involved in these phases. Conclusions: This scoping review offers a comprehensive overview of strategies used in different phases of eHealth lifestyle interventions for vulnerable pregnant women, highlighting specific barriers and facilitators. Limited reporting on the impact of the strategies used and barriers encountered hinders a complete identification of facilitators and barriers. Nevertheless, this review sheds light on how to optimize the development of eHealth lifestyle interventions for vulnerable pregnant women, ultimately enhancing the health of both future mothers and their offspring.

Background: Health education and promotion are recognized as effective strategies for fostering healthy ageing, reducing the disease burden, and addressing health inequalities, particularly when delivered through digital media. Primary care workers are often regarded as the key providers of these interventions. Despite the strong practical significance and substantial individual demand, the use of digital media for delivering health promotion practices was not widespread in China. One of the main challenges identified is the providers’ inadequate capacities. However, little is known about the digital capacity for health promotion among primary care workers. Objective: This study aimed to investigate the levels of digital capacity for health promotion and its associated factors among community health workers. Methods: A total of 1346 community health workers were recruited from across 47 communities in Shanghai, China, through cluster-stratified random sampling. The digital capacity for health promotion was measured using the revised version of the Digital Capabilities Framework. Web-based questionnaires were distributed to collect data from March 20 to March 29, 2024. Data were analyzed using descriptive statistics, independent t tests, one-way ANOVA, and linear hierarchical regression using Stata MP (version 17.0; StataCorp). Results: We included 1199 participants. Among them, 47.5% (570/1199) had high digital media use for more than 19.6 hours per week, whereas 31.8% (381/1199) demonstrated high digital media trust. The average level of digital capacity for health promotion was 16.71 (SD 2.94) out of 25 points. Demographics, digital media usage–related characteristics, perceived usefulness and usability, attitudes, and behaviors were significant predictors of the capacities, explaining 44.4% of the total variance. Master’s degree or above (β=.077; P=.013), perceived usability (β=.235; P<.001), attitudes toward digital media health promotion (β=.095; P=.002), and past digital media health promotion practices (β=.377; P<.001) had significantly positive associations with digital capacities for health promotion. However, senior (β=–.076; P=.008) or median (β=–.074; P=.01) titles had a significant negative association with capacity levels. Conclusions: A digitally capable workforce is required for primary health care systems to take full advantage of digital media health promotion. Therefore, solutions are necessary to achieve enhanced capacities among health professionals, including public health policy making, community empowerment, and individual practices.

Background: Despite the increasing popularity of mobile health (mHealth) technologies, little is known about which types of mHealth system engagement might affect the maintenance of antiretroviral therapy among people with HIV and substance use disorders. Objective: This study aimed to use longitudinal and detailed system logs and weekly survey data to test a mediation model, where mHealth engagement indicators were treated as predictors, substance use and confidence in HIV management were treated as joint mediators, and antiretroviral therapy adherence was treated as the outcome. We further distinguished the initiation and intensity of system engagement by mode (expression vs reception) and by communication levels (intraindividual vs dyadic vs network). Methods: Tailored for people with HIV living with substance use disorders, the mHealth app was distributed among 208 participants aged >18 years from 2 US health clinics. Supervised by medical professionals, participants received weekly surveys through the app to report their health status and medication adherence data. System use was passively collected through the app, operationalized as transformed click-level data, aggregated weekly, and connected to survey responses with a 7-day lagged window. Using the weekly check-in record provided by participants as the unit of analysis (N=681), linear regression and structure equation models with cluster-robust SEs were used for analyses, controlling within-person autocorrelation and group-level error correlations. Racial groups were examined as moderators in the structure equation models. Results: We found that (1) intensity, not initiation, of system use; (2) dyadic message expression and reception; and (3) network expression positively predicted medication adherence through joint mediators (substance use and confidence in HIV management). However, intraindividual reception (ie, rereading saved entries for personal motivation) negatively predicts medication adherence through joint mediators. We also found Black participants have distinct usage patterns, suggesting the need to tailor mHealth interventions for this subgroup. Conclusions: These findings highlight the importance of considering the intensity of system engagement, rather than initiation alone, when designing mHealth interventions for people with HIV and tailoring these systems to Black communities.

Traditionally, medical research is based on randomized controlled trials (RCTs) for interventions such as drugs and operative procedures. However, increasingly, there is a need for health research to evolve. RCTs are expensive to run, are generally formulated with a single research question in mind, and analyze a limited dataset for a restricted period. Progressively, health decision makers are focusing on real-world data (RWD) to deliver large-scale longitudinal insights that are actionable. RWD are collected as part of routine care in real time using digital health infrastructure. For example, understanding the effectiveness of an intervention could be enhanced by combining evidence from RCTs with RWD, providing insights into long-term outcomes in real-life situations. Clinicians and researchers struggle in the digital era to harness RWD for digital health research in an efficient and ethically and morally appropriate manner. This struggle encompasses challenges such as ensuring data quality, integrating diverse sources, establishing governance policies, ensuring regulatory compliance, developing analytical capabilities, and translating insights into actionable strategies. The same way that drug trials require infrastructure to support their conduct, digital health also necessitates new and disruptive research data infrastructure. Novel methods such as common data models, federated learning, and synthetic data generation are emerging to enhance the utility of research using RWD, which are often siloed across health systems. A continued focus on data privacy and ethical compliance remains. The past 25 years have seen a notable shift from an emphasis on RCTs as the only source of practice-guiding clinical evidence to the inclusion of modern-day methods harnessing RWD. This paper describes the evolution of synthetic data, common data models, and federated learning supported by strong cross-sector collaboration to support digital health research. Lessons learned are offered as a model for other jurisdictions with similar RWD infrastructure requirements.

Background: Patient organizations (POs) play a crucial role in supporting individuals with health conditions. Their activities range from counseling to support groups to advocacy. The COVID-19 pandemic and its related public health measures prompted rapid digital transformation efforts across multiple sectors, including health care. Objective: This study aimed to explore how POs digitally responded to pandemic-related circumstances, focusing on aspects such as the technologies used, positive outcomes, and challenges encountered. Methods: This scoping review followed the methodological guidance of the JBI (Joanna Briggs Institute) Scoping Review Methodology Group and adhered to the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) reporting guidelines. A systematic search of PubMed, the Web of Science Core Collection, and the WHO (World Health Organization) COVID-19 database, supplemented by a citation search approach, was conducted. The initial search was performed on November 10, 2022, and updated on November 8, 2023. Publications were eligible if they were published after November 30, 2019, and addressed pandemic-related digitalization efforts of POs, defined as nonprofit organizations with a focus on health-related support. A 2-step screening process was used to identify relevant literature. Data were extracted using a standardized table to capture aspects such as digital adaptation activities (eg, types of technologies implemented, positive outcomes, challenges, and facilitating factors) and coded inductively to identify similarities across included publications, and the findings were synthesized narratively. Results: The search and its subsequent update yielded 2212 records, with 13 articles included in this review. These articles revealed a range of PO services that were digitally adapted during the pandemic, with videoconferencing software emerging as the most commonly used tool (n=9 articles). The digital adaptation of group-based support activities was the most frequently reported transformation (n=9). Other adaptations included the digitalization of counseling services (n=3) and the delivery of information and education (n=3), including educational workshops, weekly webinars, and the dissemination of information through digital newsletters. While the use of digital formats, particularly for POs’ group activities, often increased accessibility by breaking down preexisting barriers (n=5), they also created new barriers for certain groups, such as those lacking digital skills or resources (n=4). Some participants experienced a loss of interpersonal aspects, like a sense of community (n=3). However, further findings suggest that the digital delivery of such group activities preserved essential interpersonal aspects (n=7) and a preference among some participants to continue digital group activities (n=4), suggesting the potential for sustainability of such options post the COVID-19 pandemic. Conclusions: The rapid digitalization efforts of POs demonstrate their adaptability and the potential of digital technologies to improve support services, despite some challenges. Future digitalization strategies should focus, among other things, on promoting digital literacy to ensure the accessibility and inclusiveness of digital services. Trial Registration: OSF Registries, https://osf.io/anvf4

Background: Mobile health (mHealth) refers to using mobile communication devices such as smartphones to support health, health care, and public health. mHealth interventions have their own nature and characteristics that distinguish them from traditional health care interventions, including drug interventions. Thus, randomized controlled trials (RCTs) of mHealth interventions present specific methodological challenges. Identifying and overcoming those challenges is essential to determine whether mHealth interventions improve health outcomes. Objective: We aimed to identify specific methodological challenges in RCTs testing mHealth interventions’ effects and develop consensus-based recommendations to address selected challenges. Methods: A 2-phase participatory research project was conducted. First, we sent a web-based survey to authors of mHealth RCTs. Survey respondents rated on a 5-point scale how challenging they found 21 methodological aspects in mHealth RCTs compared to non-mHealth RCTs. Nonsystematic searches until June 2022 informed the selection of the methodological challenges listed in the survey. Second, a subset of survey respondents participated in an online workshop to discuss recommendations to address selected methodological aspects identified in the survey. Finally, consensus-based recommendations were developed based on the workshop discussion and email interaction. Results: We contacted 1535 authors of mHealth intervention RCTs, of whom 80 (5.21%) completed the survey. Most respondents (74/80, 92%) identified at least one methodological aspect as more or much more challenging in mHealth RCTs. The aspects most frequently reported as more or much more challenging were those related to mHealth intervention integrity, that is, the degree to which the study intervention was implemented as intended, in particular managing low adherence to the mHealth intervention (43/77, 56%), defining adherence (39/79, 49%), measuring adherence (33/78, 42%), and determining which mHealth intervention components are used or received by the participant (31/75, 41%). Other challenges were also frequent, such as analyzing passive data (eg, data collected from smartphone sensors; 24/58, 41%) and verifying the participants’ identity during recruitment (28/68, 41%). In total, 11 survey respondents participated in the subsequent workshop (n=8, 73% had been involved in at least 2 mHealth RCTs). We developed 17 consensus-based recommendations related to the following four categories: (1) how to measure adherence to the mHealth intervention (7 recommendations), (2) defining adequate adherence (2 recommendations), (3) dealing with low adherence rates (3 recommendations), and (4) addressing mHealth intervention components (5 recommendations). Conclusions: RCTs of mHealth interventions have specific methodological challenges compared to those of non-mHealth interventions, particularly those related to intervention integrity. Following our recommendations for addressing these challenges can lead to more reliable assessments of the effects of mHealth interventions on health outcomes.