BACKGROUND
In China, diabetes has become much more common over the past three decades. The latest statistics on chronic diseases in China (2013) as well as the attendant risk factors, as reported by JAMA (2017), reveal that among Chinese adults, the standardized prevalence of diagnosed and undiagnosed diabetes is 10.9% (95% CI: 10.4%–11.5%) and that of prediabetes is 35.7% (95% CI: 34.1%–37.4%). By 2020, an estimated 155 million (10.9% of the population) will have a diabetes diagnosis, and App507 million (35.7%) will have a prediabetes diagnosis. Diabetes, as a chronic disease, requires medical treatment and self-management—such as adjustments to diet and physical activity. Because of the large clinical workload and hectic work schedule of Chinese health care providers, who primarily focus on treatment, crucial elements in health education are neglected and many patients do not have a personalized self-management plan designed by their health care provider. Furthermore, health care providers find it difficult to empathize with their patients emotionally and provide them prompt mental health care. In addition, health care providers, being busy, tend to be unavailable to calibrate their patients’ self-management plan and address problems as soon as they occur. ICT-based diabetes management tools have thus emerged as substitutes for health care providers, providing solutions tailored to the needs of each patient. China, in particular, has become the most promising market for self-management tools.
OBJECTIVE
To summarize the functions of representative ICT tools in China; analyze these tools’ effectiveness with reference to the literature; compare these AppAppstools with respect to their functions qualitatively and quantitatively; and finally, discuss the status, attendant risks, and further research directions regarding diabetes online communities.
METHODS
The most commonly used diabetes self-management Applications (DMAs) were searched for using the term “diabetes” in Apple’s App Store and MyApp (a widely used Android App store in China). The Apps’ basic information and main functions were then recorded. In the page for each App in the App stores, two investigators individually analyzed 99 randomly selected user comments, where keyword modifiers were coded. Each user comment was assigned a score, which was then aggregated. User comments were also coded with respect to 11 App functions. The scores for each App function and for the App as a whole Apps were analyzed using the Kruskal–Wallis test. Common online forums (Baidu Tieba, Sweet Home, and Zhihu) were selected according to user volume, number of posts, and depth of discussion. Social media platforms (WeChat and QQ) were selected according to download volume and number of related articles. The basic information and main functions of the forums and social media platforms were also collected and compared with those of the DMAs. The last section of this article reviews the literature on the effectiveness of the specific Apps analyzed here.
RESULTS
The functions of the ICT tools differed in their form, quality, and advantages offered. Peer communication is the common feature among the Apps. Welltang and DNurse were perceived most and least favorably overall, respectively (p = .024). Regarding App functions, Control Diabetes had the highest score for functions pertaining to diet, blood glucose management, and family support. The Apps did not significantly differ with respect to diabetes education and ancillary functions. Regarding diet management functionality, Control Diabetes scored the highest, primarily because it can recommend recipes, whereas Tang Tang Quan scored the lowest (p = .000). For drug management functionality, Da Tang Yi scored the highest, and Tang Tang Quan scored the lowest (p = .013). All Apps have a drug recording function, and Da Tang Yi scored the highest primarily because only it can remind users to take their medication. For blood glucose management functionality, Control Diabetes scored the highest, significantly higher than DNurse (p = .007) and Tang Tang Quan did (p = .013), possibly because of Control Diabetes’s ability to measure the user’s blood glucose and remind them to do so. For peer communication functionality, Tang Tang Quan was scored much higher than the other Apps (p = .000). We attribute this to the App being developed by the Tang Tang Quan online community, which has a larger user base and greater App development experience. Users trust Tang Tang Quan because it affords peer interaction. For doctor consultation functionality, Welltang scored the highest, whereas DNurse scored the lowest (p = .000). For family support functions, Control Diabetes scored the highest, whereas Tang Tang Quan scored the lowest (p = .026), possibly because of Tang Tang Quan’s complex binding process. Regarding functions classified as “other,” Tang Tang Quan and DNurse differed significantly (P = .027). “This version of Tang Tang Quan is well-optimized” and “operation is easy” are the most frequently used phrases in the comments. Users are deterred from using an App because of privacy concerns, lack of user friendliness, excessive advertisements, and inaccurate measurements by supporting equipment, which, in turn, reduce compliance. All ICT-based tools, except Baidu Tieba, Zhihu, and Da Tang Yi App, have been shown to improve diabetes self-management previously.
CONCLUSIONS
Each ICT-based tool has unique functionality that brings unique advantages to users in a given aspect of diabetes self-management. The supplementary functions of DMAs (e.g., recommending recipes and reminding users to take their medication or measure their blood glucose) distinguished the DMAs from each other in the function score. Although there are many ICT-based tools in the market, only a small proportion of them have been demonstrated to be effective, and the quality of research is poor. Therefore, high-quality randomized controlled trials are needed to establish the criteria for the design of ICT-based tools.
To Control Diabetes Using Machine Learning Algorithm and Calorie Measurement Technique