Mobile Applications for Type 2 Diabetes Risk Estimation: a Systematic Review

Effects of Smartphone-Based Remote Interventions on Dietary Intake, Physical Activity, Weight Control, and Related Health Benefits Among the Older Population With Overweight and Obesity in China: Randomized Controlled Trial

BACKGROUND

Traditional health management requires many human and material resources and cannot meet the growing needs. Remote medical technology provides an opportunity for health management; however, the research on it is insufficient.

OBJECTIVE

The objective of this study was to assess the effects of remote interventions on weight management.

METHODS

In this randomized controlled study, 750 participants were randomly assigned to a remote dietary and physical activity intervention group (group DPI), remote physical activity intervention group (group PI), or control group (group C). At baseline (time 1), day 45 (time 2), and day 90 (time 3), data were collected, including data on dietary intake, physical activity, indexes related to weight control, and health benefits.

RESULTS

A total of 85.6% (642/750) of participants completed the follow-up. Compared with group C, group DPI showed a significant decrease in energy intake (-581 vs -82 kcal; P<.05), protein intake (-17 vs -3 g; P<.05), fat intake (-8 vs 3 g; P<.05), and carbohydrate intake (-106.5 vs -4.7 g; P<.05) at time 3. Compared with time 1, groups DPI and PI showed a significant decrease in cereal and potato intake (P<.05). Compared with time 1, the physical activity levels related to transportation (group PI: 693 vs 597 metabolic equivalent [MET]-min/week, group C: 693 vs 594 MET-min/week; P<.05) and housework and gardening (group PI: 11 vs 0 MET-min/week, group C: 11 vs 4 MET-min/week; P<.05) in groups PI and C were improved at time 3. Compared with groups PI and C, group DPI showed a significant decrease in weight (-1.56 vs -0.86 kg and -1.56 vs -0.66 kg, respectively; P<.05) and BMI (-0.61 vs -0.33 kg/m² and -0.61 vs -0.27 kg/m², respectively; P<.05) at time 2. Compared with groups PI and C, group DPI showed a significant decrease in body weight (-4.11 vs -1.01 kg and -4.11 vs -0.83 kg, respectively; P<.05) and BMI (-1.61 vs -0.40 kg/m² and -1.61 vs -0.33 kg/m², respectively; P<.05) at time 3. Compared with group C, group DPI showed a significant decrease in triglyceride (-0.06 vs 0.32 mmol/L; P<.05) at time 2. Compared with groups PI and C, group DPI showed a significant decrease in systolic blood pressure (-8.15 vs -3.04 mmHg and -8.15 vs -3.80 mmHg, respectively; P<.05), triglyceride (-0.48 vs 0.11 mmol/L and -0.48 vs 0.18 mmol/L, respectively; P<.05), and fasting blood glucose (-0.77 vs 0.43 mmol/L and -0.77 vs 0.14 mmol/L, respectively; P<.05). There were significant differences in high-density lipoprotein cholesterol (-0.00 vs -0.07 mmol/L; P<.05) and hemoglobin A_1c (-0.19% vs -0.07%; P<.05) between groups DPI and C.

CONCLUSIONS

Remote dietary and physical activity interventions can improve dietary intake among participants with overweight and obesity, are beneficial for weight control, and have potential health benefits.

TRIAL REGISTRATION

Chinese Clinical Trial Registry ChiCTR1900023355; https://www.chictr.org.cn/showproj.html?proj=38976.

Remotely Conducted App-Based Intervention for Cardiovascular Disease and Diabetes Risk Awareness and Prevention: Single-Group Feasibility Trial

BACKGROUND

Cardiovascular disease and type 2 diabetes mellitus are two of the most prevalent chronic conditions worldwide. An unhealthy lifestyle greatly contributes to someone's risk of developing these conditions. Mobile health is an emerging technology that can help deliver health promotion interventions to the population, for example, in the form of health apps.

OBJECTIVE

The aim of this study was to test the feasibility of an app-based intervention for cardiovascular and diabetes risk awareness and prevention by measuring nonusage, dropout, adherence to app use, and usability of the app over 3 months.

METHODS

Participants were eligible if they were aged 45 years or older, resided in Australia, were free of cardiovascular disease and diabetes, were fluent in English, and owned a smartphone. In the beginning, participants received an email with instructions on how to install the app and a user guide. After 3 months, they received an email with an invitation to an end-of-study survey. The survey included questions about general smartphone use and the user version of the Mobile Application Rating Scale. We analyzed app-generated and survey data by using descriptive and inferential statistics as well as thematic analysis for open-text comments.

RESULTS

Recruitment took place between September and October 2021. Of the 46 participants who consented to the study, 20 (44%) never used the app and 15 (33%) dropped out. The median age of the app users at baseline was 62 (IQR 56-67) years. Adherence to app use, that is, using the app at least once a week over 3 months, was 17% (8/46) of the total sample and 31% (8/26) of all app users. The mean app quality rating on the user version of the Mobile Application Rating Scale was 3.5 (SD 0.6) of 5 points. The app scored the highest for the information section and the lowest for the engagement section of the scale.

CONCLUSIONS

Nonusage and dropouts were too high, and the adherence was too low to consider the intervention in its current form feasible. Potential barriers that we identified include the research team not actively engaging with participants early in the study to verify that all participants could install the app, the intervention did not involve direct contact with health care professionals, and the app did not have enough interactive features.

A Review of Mortality Risk Prediction Models in Smartphone Applications

Healthcare professionals in healthcare systems need access to freely available, real-time, evidence-based mortality risk prediction smartphone applications to facilitate resource allocation. The objective of this study is to evaluate the quality of smartphone mobile health applications that include mortality prediction models, and corresponding information quality. We conducted a systematic review of commercially available smartphone applications in Google Play for Android, and iTunes for iOS smartphone applications. We performed initial screening, data extraction, and rated smartphone application quality using the Mobile Application Rating Scale: user version (uMARS). The information quality of smartphone applications was evaluated using two patient vignettes, representing low and high risk of mortality, based on critical care data from the Medical Information Mart for Intensive Care (MIMIC) III database. Out of 3051 evaluated smartphone applications, 33 met our final inclusion criteria. We identified 21 discrete mortality risk prediction models in smartphone applications. The most common mortality predicting models were Sequential Organ Failure Assessment (SOFA) (n = 15) and Acute Physiology and Clinical Health Assessment II (n = 13). The smartphone applications with the highest quality uMARS scores were Observation—NEWS 2 (4.64) for iOS smartphones, and MDCalc Medical Calculator (4.75) for Android smartphones. All SOFA-based smartphone applications provided consistent information quality with the original SOFA model for both the low and high-risk patient vignettes. We identified freely available, high-quality mortality risk prediction smartphone applications that can be used by healthcare professionals to make evidence-based decisions in critical care environments.

A non-linear ensemble model-based surgical risk calculator for mixed data from multiple surgical fields

BACKGROUND

The misestimation of surgical risk is a serious threat to the lives of patients when implementing surgical risk calculator. Improving the accuracy of postoperative risk prediction has received much attention and many methods have been proposed to cope with this problem in the past decades. However, those linear approaches are inable to capture the non-linear interactions between risk factors, which have been proved to play an important role in the complex physiology of the human body, and thus may attenuate the performance of surgical risk calculators.

METHODS

In this paper, we presented a new surgical risk calculator based on a non-linear ensemble algorithm named Gradient Boosting Decision Tree (GBDT) model, and explored the corresponding pipeline to support it. In order to improve the practicability of our approach, we designed three different modes to deal with different data situations. Meanwhile, considering that one of the obstacles to clinical acceptance of surgical risk calculators was that the model was too complex to be used in practice, we reduced the number of input risk factors according to the importance of them in GBDT. In addition, we also built some baseline models and similar models to compare with our approach.

RESULTS

The data we used was three-year clinical data from Surgical Outcome Monitoring and Improvement Program (SOMIP) launched by the Hospital Authority of Hong Kong. In all experiments our approach shows excellent performance, among which the best result of area under curve (AUC), Hosmer-Lemeshow test ([Formula: see text]) and brier score (BS) can reach 0.902, 7.398 and 0.047 respectively. After feature reduction, the best result of AUC, [Formula: see text] and BS of our approach can still be maintained at 0.894, 7.638 and 0.060, respectively. In addition, we also performed multiple groups of comparative experiments. The results show that our approach has a stable advantage in each evaluation indicator.

CONCLUSIONS

The experimental results demonstrate that NL-SRC can not only improve the accuracy of predicting the surgical risk of patients, but also effectively capture important risk factors and their interactions. Meanwhile, it also has excellent performance on the mixed data from multiple surgical fields.

Android application for type 2 diabetes mellitus

The purpose of this study is to develop an interactive service for people with DM and design a product specifically for the Android operating system. This study employs a research and development (R&D) design. The research phase consisted of a situation analysis, data collection, product design, expert validation, product revision, product tryout, and final revision. The research instrument was a questionnaire. Data analysis involved descriptive quantitative in percentage form. The results of the validation from the Android application experts, which focused on display quality, technical concerns, audio, and video quality, were found at 93.75% (very acceptable). In addition, the results of the validity assessment from the material experts, including general health experts and nutritionists, reached 98.08% (very acceptable) and 86.54% (very feasible), respectively. Based on the results of expert validation, the application design is categorized as very acceptable for development as a DM interactive service product.

Communicating personalized risk of diabetes and offering weight reduction program choice: Recruitment, participation, and outcomes

OBJECTIVE

Low patient recruitment into diabetes prevention programs is a challenge. The primary aim of this study was to demonstrate that an increased recruitment rate can be achieved by communicating personalized risk of progression to type 2 diabetes, estimating risk reduction with weight loss, and offering program choice. Secondary aims included program participation rate, weight loss, and short-term decreased diabetes risk.

METHODS

In this single-arm study, persons with prediabetes from 3 primary care sites received a letter that communicated their personalized risk of progression to diabetes within 3-years, estimated risk reduction with 5, 10, 15 % weight loss, reported in pounds, and offered a choice of 5 free, 6-month, programs. A one-sided test was used to compare the recruitment rate against the maximum expected rate of (10 %).

RESULTS

Recruitment response rate was 25.3 % (81/328, 95 % CI=[20.0 %, 29.4 %]) which was significantly higher than expected (p < 0.0001). Overall, 65 % of participants completed >75 % of contacts. BMI, HbA1c, and diabetes risk (all p < 0.0001) improved at 6 months; BMI (p < 0.0001) and HbA1c (p < 0.05) improved at 12 months.

CONCLUSION

Recruitment response rate was better than expected.

PRACTICE IMPLICATIONS

Communicating personalized risk and reduction estimates with a choice of programs resulted in favorable outcomes, sustained at 1-year.

Issues in reporting of systematic review methods in health app-focused reviews: A scoping review

No guidelines exist for the conduct and reporting of manuscripts with systematic searches of app stores for, and then appraisal of, mobile health apps ('health app-focused reviews'). We undertook a scoping review including a systematic literature search for health app-focused reviews describing systematic app store searches and app appraisal, for apps designed for patients or clinicians. We created a data extraction template which adapted data elements from the PRISMA guidelines for systematic literature reviews to data elements operationalised for health app-focused reviews. We extracted the data from included health app-focused reviews to describe: (1) which elements of the adapted 'usual' methods of systematic review are used; (2) methods of app appraisal; and (3) reporting of clinical efficacy and recommendations for app use. From 2798 records, the 26 included health app-focused reviews showed incomplete or unclear reporting of review protocol registration; use of reporting guidelines; processes of screening apps; data extraction; and appraisal tools. Reporting of clinical efficacy of apps or recommendations for app use were infrequent. The reporting of methods in health app-focused reviews is variable and could be improved by developing a consensus reporting standard for health app-focused reviews.

Early detection of type 2 diabetes mellitus using machine learning-based prediction models

Most screening tests for T2DM in use today were developed using multivariate regression methods that are often further simplified to allow transformation into a scoring formula. The increasing volume of electronically collected data opened the opportunity to develop more complex, accurate prediction models that can be continuously updated using machine learning approaches. This study compares machine learning-based prediction models (i.e. Glmnet, RF, XGBoost, LightGBM) to commonly used regression models for prediction of undiagnosed T2DM. The performance in prediction of fasting plasma glucose level was measured using 100 bootstrap iterations in different subsets of data simulating new incoming data in 6-month batches. With 6 months of data available, simple regression model performed with the lowest average RMSE of 0.838, followed by RF (0.842), LightGBM (0.846), Glmnet (0.859) and XGBoost (0.881). When more data were added, Glmnet improved with the highest rate (+ 3.4%). The highest level of variable selection stability over time was observed with LightGBM models. Our results show no clinically relevant improvement when more sophisticated prediction models were used. Since higher stability of selected variables over time contributes to simpler interpretation of the models, interpretability and model calibration should also be considered in development of clinical prediction models.

Determining minimum set of features for diabetes mobile apps

Purpose

Interest in mobile health applications (apps) for diabetes self-care is growing. Mobile health is a promising new treatment modality for diabetes, though few smartphone apps have been designed based on a proper study and prioritization. The aim of this study was to determine a minimum set of features for diabetes mobile apps.

Methods

This study was conducted in three steps: 1.A review of the literature to collect all available features, 2. Assessing the validity of suggested features by Content Validity Index (CVI) and Content Validity Ratio (CVR), 3. Examining the importance of features by Friedman test.

Results

We retrieved all features of available mobile apps for type 2 diabetes, which are suggested and discussed in literature and compiled as a single list comprising of 33 features. Then, a survey of expert's opinion produced a set of 23 final minimum features which includes all types of tracking, mealtime tagging, food database, diet management, educational materials, healthy coping, reducing risks, problem solving, Email, color coding, alerts, reminder, target range setting, trend chart view, logbook view, numerical indicators view, customizable theme, preset notes, and custom notes. According to the mean rank which indicates the priority of each feature, the most important one was blood glucose tracking (with 16.71 mean rank) and the least important feature was the numerical indicators like such as standard deviation or average (with 6.50 mean rank).

Conclusions

The present study is the first step towards the development of our mobile apps for people with type II diabetes, and highest the essential features that are required for an optimal self-care comprehensively.

eHealth and mHealth interventions in the treatment of fatigued cancer survivors: A systematic review and meta-analysis

OBJECTIVES

To (1) evaluate existing eHealth/mHealth interventions developed to help manage cancer-related fatigue (CRF); and (2) summarize the best available evidence on their effectiveness.

METHODS

A comprehensive literature search of PubMed, MEDLINE, EMBASE, and the Cochrane Library up to November 2016 was conducted. Study outcomes were extracted, tabulated, and summarized. Random effects meta-analyses were conducted for the primary outcome (fatigue), and the secondary outcomes quality of life and depression, yielding pooled effect sizes (r), and 95% confidence intervals (CI).

RESULTS

For eHealth interventions, our search of published papers identified 9 completed studies and 6 protocols for funded projects underway. No studies were identified for mHealth interventions that met our inclusion criteria. A meta-analysis of the 9 completed eHealth studies revealed a statistically significant beneficial effect of eHealth interventions on CRF (r = .27, 95% CI [.1109 - .4218], P < 0.01). Therapist-guided eHealth interventions were more efficacious then self-guided interventions (r = .58, 95% CI: [.3136 - .5985, P < 0.001). Small to moderate therapeutic effects were also observed for HRQoL (r = .17, 95% CI [.0384 - .3085], P < 0.05) and depression (r = .24, 95% CI [.1431 - .3334], P < 0.001).

CONCLUSIONS

eHealth interventions appear to be effective for managing fatigue in cancer survivors with CRF. Continuous development of eHealth interventions for the treatment of CRF in cancer survivors and their testing in long-term, large-scale efficacy outcome studies is encouraged. The degree to which mHealth interventions can change CRF in cancer survivors need to be assessed systematically and empirically.

Quality of Mobile Phone and Tablet Mobile Apps for Speech Sound Disorders: Protocol for an Evidence-Based Appraisal

Background

Although mobile apps are readily available for speech sound disorders (SSD), their validity has not been systematically evaluated. This evidence-based appraisal will critically review and synthesize current evidence on available therapy apps for use by children with SSD.

Objective

The main aims are to (1) identify the types of apps currently available for Android and iOS mobile phones and tablets, and (2) to critique their design features and content using a structured quality appraisal tool.

Methods

This protocol paper presents and justifies the methods used for a systematic review of mobile apps that provide intervention for use by children with SSD. The primary outcomes of interest are (1) engagement, (2) functionality, (3) aesthetics, (4) information quality, (5) subjective quality, and (6) perceived impact. Quality will be assessed by 2 certified practicing speech-language pathologists using a structured quality appraisal tool. Two app stores will be searched from the 2 largest operating platforms, Android and iOS. Systematic methods of knowledge synthesis shall include searching the app stores using a defined procedure, data extraction, and quality analysis.

Results

This search strategy shall enable us to determine how many SSD apps are available for Android and for iOS compatible mobile phones and tablets. It shall also identify the regions of the world responsible for the apps’ development, the content and the quality of offerings. Recommendations will be made for speech-language pathologists seeking to use mobile apps in their clinical practice.

Conclusions

This protocol provides a structured process for locating apps and appraising the quality, as the basis for evaluating their use in speech pathology for children in English-speaking nations.