Mobile toolbox (MTB) remote measures of executive function and processing speed: development and validation.

OBJECTIVE: The ability to remotely monitor cognitive skills is increasing with the ubiquity of smartphones. The Mobile Toolbox (MTB) is a new measurement system that includes measures assessing Executive Functioning (EF) and Processing Speed (PS): Arrow Matching, Shape-Color Sorting, and Number-Symbol Match. The purpose of this study was to assess their psychometric properties. METHOD: MTB measures were developed for smartphone administration based on constructs measured in the NIH Toolbox® (NIHTB). Psychometric properties of the resulting measures were evaluated in three studies with participants ages 18 to 90. In Study 1 (N = 92), participants completed MTB measures in the lab and were administered both equivalent NIH TB measures and other external measures of similar cognitive constructs. In Study 2 (N = 1,021), participants completed the equivalent NIHTB measures in the lab and then took the MTB measures on their own, remotely. In Study 3 (N = 168), participants completed MTB measures twice remotely, two weeks apart. RESULTS: All three measures exhibited very high internal consistency and strong test-retest reliability, as well as moderately high correlations with comparable NIHTB tests and moderate correlations with external measures of similar constructs. Phone operating system (iOS vs. Android) had a significant impact on performance for Arrow Matching and Shape-Color Sorting, but no impact on either validity or reliability. CONCLUSIONS: Results support the reliability and convergent validity of MTB EF and PS measures for use across the adult lifespan in remote, self-administered designs.

Consideration of inequalities in effectiveness trials of mHealth applications - a systematic assessment of studies from an umbrella review.

BACKGROUND: The growing use of mobile health applications (apps) for managing diabetes and hypertension entails an increased need to understand their effectiveness among different population groups. It is unclear if efficacy and effectiveness trials currently provide evidence of differential effectiveness, and if they do, a summary of such evidence is missing. Our study identified to what extent sociocultural and socioeconomic inequalities were considered in effectiveness trials of mobile health apps in diabetic and hypertensive patients and if these inequalities moderated app effectiveness. METHODS: We built on our recent umbrella review that synthesized systematic reviews (SRs) of randomized controlled trials (RCTs) on the effectiveness of health apps. Using standard SR methodologies, we identified and assessed all primary RCTs from these SRs that focused on diabetes and/or hypertension and reported on health-related outcomes and inequality-related characteristics across intervention arms. We used the PROGRESS-Plus framework to define inequality-related characteristics that affect health opportunities and outcomes. We used harvest plots to summarize the subgroups (stratified analyses or interaction terms) on moderating effects of PROGRESS-Plus. We assessed study quality using the Risk of Bias 2 tool. RESULTS: We included 72 published articles of 65 unique RCTs. Gender, age, and education were the most frequently described PROGRESS-Plus characteristics at baseline in more than half of the studies. Ethnicity and occupation followed in 21 and 15 RCTs, respectively. Seven trials investigated the moderating effect of age, gender or ethnicity on app effectiveness through subgroup analyses. Results were equivocal and covered a heterogenous set of outcomes. Results showed some concerns for a high risk of bias, mostly because participants could not be blinded to their intervention allocation. CONCLUSIONS: Besides frequently available gender, age, and education descriptives, other relevant sociocultural or socioeconomic characteristics were neither sufficiently reported nor analyzed. We encourage researchers to investigate how these characteristics moderate the effectiveness of health apps to better understand how effect heterogeneity for apps across different sociocultural or socioeconomic groups affects inequalities, to support more equitable management of non-communicable diseases in increasingly digitalized systems. REGISTRATION: https://osf.io/89dhy/ .

Using the Mobile Toolbox in child and adolescent samples: A feasibility study.

Cognitive research with developmental samples requires improved methods that support large-scale, diverse, and open science. This paper offers initial evidence to support the Mobile Toolbox (MTB), a self-administered remote smartphone-based cognitive battery, in youth populations, from a pilot sample of 99 children (Mage = 11.79 years; 36% female; 53% White, 33% Black or African American, 9% Asian, and 15% Hispanic). Completion rates (95%-99%), practice performance (96%-100%), internal consistency (0.60-0.98), and correlations with similar NIHTB measures (0.55-0.77) provide the first evidence to support the MTB in a youth sample, although there were some inconsistencies across measures. Preliminary findings provide promising evidence of the MTB in developmental populations, and further studies are encouraged.

Quality Assessment of Digital Health Apps: Umbrella Review.

BACKGROUND: With an increasing number of digital health apps available in app stores, it is important to assess these technologies reliably regarding their quality. This is done to mitigate the risks associated with their use. There are many different guidelines, methods, and metrics available to assess digital health apps with regard to their quality. OBJECTIVE: This study aimed to give a holistic summary of the current methods and "condition agnostic" frameworks that are broadly applicable for the quality assessment of all digital health apps. METHODS: A systematic search of literature was conducted on 4 databases: Scopus, PubMed, ACM Digital Library, and IEEE Xplore. We followed the PICOS (Population, Patient, or Problem; Intervention; Comparison; Outcomes; and Study Design) and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) methodologies when conducting this umbrella review. The search was conducted on January 26, 2024, for review articles published between 2018 and 2023. We identified 4781 candidate papers for inclusion; after title and abstract screening, 39 remained. After full-text analysis, we included 15 review articles in the full review. RESULTS: Of the 15 review articles, scoping reviews were the most common (n=6, 40%), followed by systematic reviews (n=4, 27%), narrative reviews (n=4, 27%), and a rapid review (n=1, 7%). A total of 4 (27%) review articles proposed assessment criteria for digital health apps. "Data privacy and/or security" was the most mentioned criterion (n=13, 87%) and "Cost" was the least mentioned criterion (n=1, 7%) for the assessment of digital health apps. The Mobile App Rating Scale was the most frequently used framework for quality assessment of digital health apps. CONCLUSIONS: There is a lack of unity or consolidation across identified frameworks, as most do not meet all the identified criteria from the reviewed articles. Safety concerns associated with the use of digital health apps may be mitigated with the use of quality frameworks.

Challenges and Alternatives to Evaluation Methods and Regulation Approaches for Medical Apps as Mobile Medical Devices: International and Multidisciplinary Focus Group Discussion.

BACKGROUND: The rapid proliferation of medical apps has transformed the health care landscape by giving patients and health care providers unprecedented access to personalized health information and services. However, concerns regarding the effectiveness and safety of medical apps have raised questions regarding the efficacy of randomized controlled trials (RCTs) in the evaluation of such apps and as a requirement for their regulation as mobile medical devices. OBJECTIVE: This study aims to address this issue by investigating alternative methods, apart from RCTs, for evaluating and regulating medical apps. METHODS: Using a qualitative approach, a focus group study with 46 international and multidisciplinary public health experts was conducted at the 17th World Congress on Public Health in May 2023 in Rome, Italy. The group was split into 3 subgroups to gather in-depth insights into alternative approaches for evaluating and regulating medical apps. We conducted a policy analysis on the current regulation of medical apps as mobile medical devices for the 4 most represented countries in the workshop: Italy, Germany, Canada, and Australia. We developed a logic model that combines the evaluation and regulation domains on the basis of these findings. RESULTS: The focus group discussions explored the strengths and limitations of the current evaluation and regulation methods and identified potential alternatives that could enhance the quality and safety of medical apps. Although RCTs were only explicitly mentioned in the German regulatory system as one of many options, an analysis of chosen evaluation methods for German apps on prescription pointed toward a "scientific reflex" where RCTs are always the chosen evaluation method. However, this method has substantial limitations when used to evaluate digital interventions such as medical apps. Comparable results were observed during the focus group discussions, where participants expressed similar experiences with their own evaluation approaches. In addition, the participants highlighted numerous alternatives to RCTs. These alternatives can be used at different points during the life cycle of a digital intervention to assess its efficacy and potential harm to users. CONCLUSIONS: It is crucial to recognize that unlike analog tools, digital interventions constantly evolve, posing challenges to inflexible evaluation methods such as RCTs. Potential risks include high dropout rates, decreased adherence, and nonsignificant results. However, existing regulations do not explicitly advocate for other evaluation methodologies. Our research highlighted the necessity of overcoming the gap between regulatory demands to demonstrate safety and efficacy of medical apps and evolving scientific practices, ensuring that digital health innovation is evaluated and regulated in a way that considers the unique characteristics of mobile medical devices.

Examining labelling guidelines for AI-based software as a medical device: A review and analysis of dermatology mobile applications in Australia.

In recent years, there has been a surge in the development of AI-based Software as a Medical Device (SaMD), particularly in visual specialties such as dermatology. In Australia, the Therapeutic Goods Administration (TGA) regulates AI-based SaMD to ensure its safe use. Proper labelling of these devices is crucial to ensure that healthcare professionals and the general public understand how to use them and interpret results accurately. However, guidelines for labelling AI-based SaMD in dermatology are lacking, which may result in products failing to provide essential information about algorithm development and performance metrics. This review examines existing labelling guidelines for AI-based SaMD across visual medical specialties, with a specific focus on dermatology. Common recommendations for labelling are identified and applied to currently available dermatology AI-based SaMD mobile applications to determine usage of these labels. Of the 21 AI-based SaMD mobile applications identified, none fully comply with common labelling recommendations. Results highlight the need for standardized labelling guidelines. Ensuring transparency and accessibility of information is essential for the safe integration of AI into health care and preventing potential risks associated with inaccurate clinical decisions.

RegulEm, an unified protocol based-app for the treatment of emotional disorders: a parallel mixed methods usability and quality study.

BACKGROUND: Interest in mental health smartphone applications has grown in recent years. Despite their effectiveness and advantages, special attention needs to be paid to two aspects to ensure app engagement: to include patients and professionals in their design and to guarantee their usability. The aim of this study was to analyse the perceived usability and quality of the preliminary version of RegulEm, an app based in the Unified Protocol, as part of the second stage of the app development. METHODS: A parallel mixed methods study was used with 7 professionals and 4 users who were previously involved in the first stage of the development of the app. MARS, uMARS and SUS scales were used, and two focus groups were conducted. Descriptive statistical analysis and a thematic content analysis were performed in order to gather as much information as possible on RegulEm's usability and quality as well as suggestions for improvement. RESULTS: RegulEm's usability was perceived through the SUS scale scores as good by users (75 points) and excellent by professionals (84.64 points), while its quality was perceived through the uMARS and MARS scales as good by both groups, with 4 and 4.14 points out of 5. Different areas regarding RegulEm's usability and suggestions for improvement were identified in both focus groups and 20% of the suggestions proposed were implemented in the refined version of RegulEm. CONCLUSION: RegulEm's usability and quality were perceived as good by users and professionals and different identified areas have contributed to its refinement. This study provides a more complete picture of RegulEm's usability and quality prior analysing its effectiveness, implementation and cost-effectiveness in Spanish public mental health units.

Developing a governance process for apps can help enhance and change the skills base and visibility of health librarians.

The traditional qualifications and work of a health librarian may not, at first glance, seem like they readily lend themselves to the wider work of an organisation. Too often librarians are seen as experts in a small specialist field. However, as librarians, we know that at our core is extensive digital experience and knowledge as well as a core set of transferrable skills that can be adapted to meet the ever-changing needs of the organisation. This article describes how the library evidence team became part of a wider board project to develop a governance system for Apps. It also describes how the skills of librarians can be developed to work in this area and raise the profile of the team within the board.

Mobile Health-Based Health Literacy Weight Management Intervention Using Smart Devices for Adolescents: A Convergent Mixed-Methods Pilot Study.

This study aimed to develop and evaluate the feasibility of a mobile health-based health literacy weight management intervention program for adolescents during the COVID-19 pandemic. Therefore, a convergent mixed-method design was used. The "GO! GO! Smart Healthy" intervention development engaged multidisciplinary healthcare professionals, utilizing mobile health applications and smartwatches. Participants provided data on weight management, dietary habits, and health literacy via an mHealth app, while smartwatches collected objective information on physical activity and sleep. The 12-week pilot study was conducted from March to September 2021. It evaluated the program's feasibility, employing both quantitative and thematic analyses of focus group discussions for a comprehensive assessment. Of the 27 participants, 44.4% were overweight or obese, with insufficient physical activity reported. Thematic analysis identified three main themes: empowerment, self-awareness and self-regulation, and self-comparison and reflection. Adolescents engaged with smartwatches, demonstrated increased awareness, altruistic behavior, and satisfaction. After integrating mixed-method data, revisions ensured the program's alignment with participants' preferences. Despite pandemic disruptions, mHealth apps ensured uninterrupted research, showcasing technology's evolving role in healthcare and education. Iterative feedback with participants ensured program suitability, fostering effectiveness. Trial Registration: ClinicalTrials.gov: NCT04759716.

The use of mHealth apps to improve hospital nurses' mental health and well-being: A systematic review.

BACKGROUND: Nursing well-being has become a heightened focus since the COVID-19 pandemic. Nurses are leaving the profession early in their careers or retiring sooner than expected. Those who remain in the workforce report higher levels of burnout, anxiety, depression, and exhaustion. There is concern that there may be a shortage of at least half a million nurses by 2030. AIMS: This systematic review aimed to investigate the evidence of using a mental health promotion mHealth app to improve the mental health of hospital nurses. METHODS: A systematic search was conducted in CINAHL Plus with Full Text, MEDLINE with Full Text, Professional Development Collection, Psychology and Behavioral Sciences Collection, Sociological Collection, PsycInfo, Embase, and PubMed with search dates of January 2012-November 15, 2022. The mHealth intervention needed to be asynchronously delivered through a smartphone with hospital nurse participants to be included in this review. RESULTS: Of the 157 articles screened for this review, six were included. Primary outcome variables were anxiety, burnout, coping, depression, self-efficacy, stress, well-being, and work engagement. Intervention types included mindfulness-based interventions (MBIs), cognitive behavioral therapy (CBT), stress inoculation therapy (SIT), psychoeducation, and stress management. Anxiety, depression, well-being, and burnout improved with MBIs; depression improved with CBT; and anxiety and active coping improved with SIT. LINKING EVIDENCE TO ACTION: This review demonstrated promising findings in using mHealth apps to improve the mental health of hospital nurses. However, more randomized controlled trials with larger sample sizes may reveal which type of mHealth app and how much exposure to the intervention is more effective in improving specific mental health symptoms. Longitudinal follow-up is also recommended to study sustainability of the mental health improvements.

Validation of snoring detection using a smartphone app.

PURPOSE: Snoring is closely related to obstructive sleep apnea in adults. The increasing abundance and availability of smartphone technology has facilitated the examination and monitoring of snoring at home through snoring apps. However, the accuracy of snoring detection by snoring apps is unclear. This study explored the snoring detection accuracy of Snore Clock-a paid snoring detection app for smartphones. METHODS: Snoring rates were detected by smartphones that had been installed with the paid app Snore Clock. The app provides information on the following variables: sleep duration, snoring duration, snoring loudness (in dB), maximum snoring loudness (in dB), and snoring duration rate (%). In brief, we first reviewed the snoring rates detected by Snore Clock; thereafter, an ear, nose, and throat specialist reviewed the actual snoring rates by using the playback of the app recordings. RESULTS: In total, the 201 snoring records of 11 patients were analyzed. Snoring rates measured by Snore Clock and those measured manually were closely correlated (r = 0.907). The mean snoring detection accuracy rate of Snore Clock was 95%, with a positive predictive value, negative predictive value, sensitivity, and specificity of 65% ± 35%, 97% ± 4%, 78% ± 25%, and 97% ± 4%, respectively. However, the higher the snoring rates, the higher were the false-negative rates for the app. CONCLUSION: Snore Clock is compatible with various brands of smartphones and has a high predictive value for snoring. Based on the strong correlation between Snore Clock and manual approaches for snoring detection, these findings have validated that Snore Clock has the capacity for at-home snoring detection.

AI-based smartphone apps for risk assessment of skin cancer need more evaluation and better regulation.

Smartphone applications ("apps") with artificial intelligence (AI) algorithms are increasingly used in healthcare. Widespread adoption of these apps must be supported by a robust evidence-base and app manufacturers' claims appropriately regulated. Current CE marking assessment processes inadequately protect the public against the risks created by using smartphone diagnostic apps.

Societal Effects Are a Major Factor for the Uptake of the Coronavirus Disease 2019 (COVID-19) Digital Contact Tracing App in The Netherlands.

OBJECTIVES: Our study investigates the extent to which uptake of a COVID-19 digital contact-tracing (DCT) app among the Dutch population is affected by its configurations, its societal effects, and government policies toward such an app. METHODS: We performed a discrete choice experiment among Dutch adults including 7 attributes, that is, who gets a notification, waiting time for testing, possibility for shops to refuse customers who have not installed the app, stopping condition for contact tracing, number of people unjustifiably quarantined, number of deaths prevented, and number of households with financial problems prevented. The data were analyzed by means of panel mixed logit models. RESULTS: The prevention of deaths and financial problems of households had a very strong influence on the uptake of the app. Predicted app uptake rates ranged from 24% to 78% for the worst and best possible app for these societal effects. We found a strong positive relationship between people's trust in government and people's propensity to install the DCT app. CONCLUSIONS: The uptake levels we find are much more volatile than the uptake levels predicted in comparable studies that did not include societal effects in their discrete choice experiments. Our finding that the societal effects are a major factor in the uptake of the DCT app results in a chicken-or-the-egg causality dilemma. That is, the societal effects of the app are severely influenced by the uptake of the app, but the uptake of the app is severely influenced by its societal effects.

E-mental health applications for depression: an evidence-based ethical analysis.

E-mental health applications (apps) are an increasingly important factor for the treatment of depression. To assess the risks and benefits for patients, an in-depth ethical analysis is necessary. The objective of this paper is to determine the ethical implications of app-based treatment for depression. An evidence-based ethical analysis was conducted. The material was meta-reviews and randomized control studies (RCTs) on app-based treatment. Based on the empirical data, an ethical analysis was conducted using the 3-ACES-approach by Thornicroft and Tansella. Apps may empower autonomy, offer an uninterrupted series of contacts over a period of time, show evidence-based benefits for patients with subclinical and mild-to-moderate-symptoms, are easily accessible, may be used for coordinating information and services within an episode of care, and are on the whole cost-effective. Their risks are that they are not suitable for the whole range of severity of mental illnesses and patient characteristics, show severe deficits in the data privacy policy, and a big variability in quality standards. The use of apps in depression treatment can be beneficial for patients as long as (1) the usefulness of an app-based treatment is assessed for each individual patient, (2) apps are chosen according to symptom severity as well as characteristics like the patient's level of self-reliance, their e-literacy, and their openness vis-à-vis apps, (3) manufacturers improve their privacy policies and the quality of apps.

Comprehensively addressing postpartum maternal health: a content and image review of commercially available mobile health apps.

BACKGROUND: The United States is currently facing a maternal morbidity and mortality crisis, with the highest rates of any resource-rich nation. In efforts to address this, new guidelines for postpartum care suggest that mobile health (mHealth) apps can help provide complementary clinical support for new mothers during the postpartum period. However, to date no study has evaluated the quality of existing mHealth tools targeted to this time period in terms of sufficiency of maternal health information, inclusivity of people of color, and app usability. METHODS: Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) standards were used to review the peripartum apps from the Apple and Google Play stores in either the Health/Fitness, Medical, or Education categories. Apps were evaluated for extent and quality of maternal health information and inclusivity of people of color using an a priori coding scheme. App usability was evaluated using the Mobile Application Rating Scale (MARS) score. RESULTS: Of the 301 apps from the Apple and Google Play stores, 25 met criteria for final evaluation. Of the 30 maternal health topics coded for, the median number addressed by apps was 19.5 (65%). Peripartum behaviors were more frequently addressed than peripartum outpatient care topics and peripartum acute health risks. The coverage of maternal health information and inclusivity of people of color in app imagery both correlated positively with the MARS usability score of the app. Only 8 apps (32%) portrayed greater than 24% images of people of color- the percent of non-white Americans according to 2019 census estimates. There was no correlation between MARS usability score and number of app users, as estimated by number of ratings for the app available on the app store. In addition, apps with evidence-based maternal health information had greater MARS engagement, information, and aesthetics scores. However, presence of evidence-based information did not correlate with greater numbers of app users. CONCLUSIONS: Current commercially available peripartum apps range widely in quality. Overall current app offerings generally do not provide adequate maternal health information and are not optimally accessible to the target users in terms of inclusivity of women of color or app usability. Apps delivering evidence-based information and more usable design are more likely to meet these standards but are not more likely to be downloaded by users.

A review of iPhone and Android apps for cancer patients and survivors: assessing their quality, nutrition information and behaviour change techniques.

BACKGROUND: The present study reviewed the quality, nutrition content and behaviour change techniques (BCTs) of apps aimed at those with cancer. METHODS: The Apple App Store and Androids' Google Play were searched in March 2020. Apps were scored for accountability as per Silberg's standards, quality as per the Mobile Application Rating Scale, and BCTs using the CALO-RE Taxonomy. Nutrition content was summarised and a checklist developed from the European Society for Parenteral and Enteral Nutrition (ESPEN) cancer and nutrition guidelines and the World Cancer Research Fund (WCRF) guidelines for Cancer Survivors. RESULTS: Twelve apps were identified, mean (SD) accountability score was 2.7/8 (2.0) (range 0-6) and quality score was 2.9/5 (0.6) (range 1.7-3.7). Overall, 11 BCTs were used (range 0-8 per app). Nutrition content focussed on healthy eating and meeting energy needs. There was a lack of strategies for implementation and no indication of whether the advice was more suited for specific cancer types, stages or treatment. Limited reference was made to recommendations of ESPEN and the WCRF. A strong positive relationship between quality and number of BCTs was found (r = 0.805, n = 9, P = 0.01). CONCLUSIONS: Little nutrition information is currently included on publicly available apps aimed at those with a cancer diagnosis.

A Mobile App to Support Clinical Diagnosis of Upper Respiratory Problems (eHealthResp): Co-Design Approach.

BACKGROUND: The misuse of antibiotics is a global public health issue that fosters bacterial resistance and jeopardizes generational health. The development of validated tools such as web-based courses and mobile apps to enhance clinical decisions in upper respiratory infections is of great importance in reducing the incorrect use of antibiotics in these situations. OBJECTIVE: The aim of this study was to design and prevalidate the interface of a mobile app to assist and provide clinical support in the diagnosis of upper respiratory problems. We aimed to assess the adequacy and usability of the interface of the tool in the belief that it could be beneficial to health care delivery in the clinical decision setting. METHODS: Using a co-design approach that brought together professionals in interface design and experts in pharmacology and pharmacoepidemiology, the mobile app interface was evaluated through peer review sessions held by interface design professionals on a heuristic survey. The reviewers accessed a high-fidelity interactive mock-up of the interface and filled in a questionnaire to assess the dimensions of layout and visual design and navigation and tasks. The resulting feedback of this evaluation supported the redesign of the primary interface, which was assessed for the second time by 2 of the previously mentioned reviewers. RESULTS: With 4 as the highest score, the interface scored a mean of 3.16 (SD 0.45; median of the means 3.2) for layout and visual design and a mean of 3.43 (SD 0.33; median of the means 3.51) for navigation and tasks, reflecting an overall positive evaluation. The open-ended commentaries allowed us to better understand specific recommendations of the reviewers. Throughout this section, approximately 0.98 comments per parameter were registered, reflecting a high level of effectiveness of the chosen parameters in identifying potential problems. The resultant beta version of the interface, addressing the majority of the detected problems, was further assessed by 2 of the previous reviewers, validating the new design. Future tests with physicians and pharmacists will help assess credibility and user experience dimensions. CONCLUSIONS: Our study revealed that the designed interface is easy to interpret and use. Peer reviewers raised important issues that could be easily fixed and positively reassessed. As a result, the study enabled us to produce a new tool for interface usability assessment and a set of recommendations for developing mobile interfaces for clinical decision support systems in the scope of upper respiratory problems.

Lifelog Data-Based Prediction Model of Digital Health Care App Customer Churn: Retrospective Observational Study.

BACKGROUND: Customer churn is the rate at which customers stop doing business with an entity. In the field of digital health care, user churn prediction is important not only in terms of company revenue but also for improving the health of users. Churn prediction has been previously studied, but most studies applied time-invariant model structures and used structured data. However, additional unstructured data have become available; therefore, it has become essential to process daily time-series log data for churn predictions. OBJECTIVE: We aimed to apply a recurrent neural network structure to accept time-series patterns using lifelog data and text message data to predict the churn of digital health care users. METHODS: This study was based on the use data of a digital health care app that provides interactive messages with human coaches regarding food, exercise, and weight logs. Among the users in Korea who enrolled between January 1, 2017 and January 1, 2019, we defined churn users according to the following criteria: users who received a refund before the paid program ended and users who received a refund 7 days after the trial period. We used long short-term memory with a masking layer to receive sequence data with different lengths. We also performed topic modeling to vectorize text messages. To interpret the contributions of each variable to model predictions, we used integrated gradients, which is an attribution method. RESULTS: A total of 1868 eligible users were included in this study. The final performance of churn prediction was an F1 score of 0.89; that score decreased by 0.12 when the data of the final week were excluded (F1 score 0.77). Additionally, when text data were included, the mean predicted performance increased by approximately 0.085 at every time point. Steps per day had the largest contribution (0.1085). Among the topic variables, poor habits (eg, drinking alcohol, overeating, and late-night eating) showed the largest contribution (0.0875). CONCLUSIONS: The model with a recurrent neural network architecture that used log data and message data demonstrated high performance for churn classification. Additionally, the analysis of the contribution of the variables is expected to help identify signs of user churn in advance and improve the adherence in digital health care.

Digital Health Interventions for Cardiac Rehabilitation: Systematic Literature Review.

BACKGROUND: Cardiovascular disease (CVD) is the leading cause of death worldwide. Despite strong evidence supporting the benefits of cardiac rehabilitation (CR), over 80% of eligible patients do not participate in CR. Digital health technologies (ie, the delivery of care using the internet, wearable devices, and mobile apps) have the potential to address the challenges associated with traditional facility-based CR programs, but little is known about the comprehensiveness of these interventions to serve as digital approaches to CR. Overall, there is a lack of a systematic evaluation of the current literature on digital interventions for CR. OBJECTIVE: The objective of this systematic literature review is to provide an in-depth analysis of the potential of digital health technologies to address the challenges associated with traditional CR. Through this review, we aim to summarize the current literature on digital interventions for CR, identify the key components of CR that have been successfully addressed through digital interventions, and describe the gaps in research that need to be addressed for sustainable and scalable digital CR interventions. METHODS: Our strategy for identifying the primary literature pertaining to CR with digital solutions (defined as technology employed to deliver remote care beyond the use of the telephone) included a consultation with an expert in the field of digital CR and searches of the PubMed (MEDLINE), Embase, CINAHL, and Cochrane databases for original studies published from January 1990 to October 2018. RESULTS: Our search returned 31 eligible studies, of which 22 were randomized controlled trials. The reviewed CR interventions primarily targeted physical activity counseling (31/31, 100%), baseline assessment (30/31, 97%), and exercise training (27/31, 87%). The most commonly used modalities were smartphones or mobile devices (20/31, 65%), web-based portals (18/31, 58%), and email-SMS (11/31, 35%). Approximately one-third of the studies addressed the CR core components of nutrition counseling, psychological management, and weight management. In contrast, less than a third of the studies addressed other CR core components, including the management of lipids, diabetes, smoking cessation, and blood pressure. CONCLUSIONS: Digital technologies have the potential to increase access and participation in CR by mitigating the challenges associated with traditional, facility-based CR. However, previously evaluated interventions primarily focused on physical activity counseling and exercise training. Thus, further research is required with more comprehensive CR interventions and long-term follow-up to understand the clinical impact of digital interventions.