In China, noncommunicable chronic diseases not only are a critical health problem but also impose a major economic burden. Globally, diabetes mellitus (DM) has been diagnosed in 415 million people, and DM and its complications are a global health emergency, accounting for 12% of global health expenditure [1]. The MOST survey revealed that 11.6% of adults in China (approximately 114 million people) have received DM diagnoses. The prevalence of DM has resulted in a substantial financial burden on medical systems, families, and societies. DM is more prevalent in older age groups and among people living in economically developed regions. Overall, DM-related expenditure in 2013 was approximately US $548 billion, accounting for 10.8% of total health expenditure worldwide. DM-related expenditure is expected to exceed US $627 billion by 2035 [2,3].

Self-management has become a critical approach for improving health outcomes among patients with DM, and market demand has emerged for related innovations in information technology (IT). Specifically, patients have expressed interest in cloud-based Internet of Things (IoT) measurement devices to record various types of structured and unstructured data (eg, blood glucose, blood pressure, body mass index, body fat, sleep quality, fatigue [4], depressive symptoms, working pressure, and social engagement frequency) [5-7]. Providers hope to integrate patient data into records of patient history, which can be used to derive engagement models of health promotion for increasing patient motivation and adherence [7-9].

The use of mobile health apps as a lifestyle intervention for improving diabetes outcomes has become increasingly common. In 1 study, patients who underwent this intervention exhibited continual improvement in glycemic control compared with baseline values for the duration of the program (with approximately 50% of participants losing at least 5% of their body weight over 6 months). However, the benefits of the transtheoretical model (TTM) [8,10] and behavioral change [4] were not sustained above the control group setting at 3 months after completion of the program [11].

Patients with diabetes should be encouraged to change their lifestyles and synchronize their personal data recorded inside and outside of hospitals [12]. Thus, patient-centered approaches should be developed for new dynamic behavioral engagement models [13] and as enhancements to existing health promotion models. Personalized approaches should focus on transtheoretical methods as well as patients’ self-management skills, wellness knowledge [9,14], education, psychosocial assessments, satisfaction, and rate of adherence with treatment [6,15].

Adherence exhibits significant beneficial effects on fatigue and social interaction for insulin-using patients with diabetes, emotional well-being for patients with myocardial infarction [4], and glycated hemoglobin (HbA_1c) for patients with myocardial infarction and diabetes. In addition, researchers have indicated that higher rates of adherence (especially to diet-related recommendations) are associated with improved health outcomes among insulin-using patients with diabetes [11,16].

Observational studies have reported that educational programs for patients with diabetes do not succeed unless they are intensive and continued over a long period with monitoring of behavioral changes and interventions [17-20].

The aim of this study was to identify the personalized influence factors of goal setting, self-execution activities, and self-efficacy for patients with type 2 diabetes. In addition, we aimed to analyze the association of the identified factors with patient-centered health promotion behavior and to encourage patients to engage in electronic word-of-mouth communication [21,22] to motivate their therapy partners [23,24].

In this study, we developed a cloud-based interactive health care management mobile app, interactive personalized management framework (IPMF), and adopted a platform as a service tier. The platform integrates and displays patient information on a 360-degree dashboard. The dashboard may be used to monitor patient health status and facilitate engagement with physicians. Interactive wellness education is also offered through the mobile-based app and cloud-based interactive service platform [16,25-27]. In addition to genetic and environmental factors, patient behavior is a critical factor associated with chronic disease. The effectiveness of disease management depends not only on patients’ clinical status but also on their lifestyle, disease knowledge, health literacy, beliefs, cognitive state, behavior changes, and emotional state [28-32].

Previous studies have demonstrated that health educators and care managers must understand a patient’s disease literacy level, personal characteristics, and readiness for action to determine effective treatment [1,33]. Care providers should create a service environment that enables patients to adopt appropriate behavioral changes [34,35].

Hence, in the developed system, the application programming interface engine uses a three-dimensional analysis tool to aggregate individual patient data and health status. The x-axis denotes disease type and indicates undiagnosed but high-risk and at-risk status, early diagnosed disease, self-perceived symptoms, and diagnosed disease. The y-axis denotes the severity of each symptom. Finally, the z-axis mainly displays the factors that may influence patients’ health in daily life, including family history, education, economic situation, lifestyle, and stress (see Figure 1).

This study used the TTM under the theory of planned behavior to monitor patients’ behavioral changes during the preintervention assessment, follow-up, and postintervention assessment periods [1,5].

Studies have indicated that IT and IoT measurement devices are valuable means of supporting disease management [9,33]. Studies have also reported the use of these technologies in the provision of evidence-based health education tailored to individual patients [16,36]. However, most relevant transtheoretical studies have not analyzed patients' self-knowledge or the effectiveness of patients’ self-care behavior for health management. Specifically, studies have not compared the results of health knowledgeability analyses with continuous data on personal health measurement data and health outcome.

The investigation comprised beginning (3 months), follow-up (18 months), and closure phases (3 months). Each outpatient had a different return visit time; therefore, for preassessment, the participants completed self-assessments using the mobile-based app throughout the initial phase (the open period was from months 1-6). Subsequently, each patient collaborated with a physician to set a personal goal and define a health program for the follow-up phase (from months 6-18). By the time they had completed 4 to 6 visits to the outpatient department, each participant had finished the postassessment of the closure phase (from months 18-24). To integrate behavioral activities into the health promotion service flow, the mobile apps were provided to the participants through a mobile tablet while they were waiting to see a doctor or health educator (see Figure 3).

Mobile tablets, such as iPads and mobile phone, also served as supporting tools for physicians and health educators to access patients’ integrated information. Patient compliance data were collected through the interface of the IPMF as the patient entered the start phase (1-6 months) for preintervention assessment and the closure phase (18-24 months) later for postintervention assessment. The patient self-assessment score was a taxonomy reference. The behaviors outlined in the American Association of Diabetes Educators (AADE) 7 Self-Care Behaviors program [14,34,37] were used as indicators. These behaviors comprise healthy eating, being active, blood-glucose and blood-pressure monitoring, medication use, problem-solving, use of healthy coping strategies, and risk reduction (see Figure 4).

Patient demographic data that were relevant to health promotion factors were provided by the hospitals providing care and inputted into the IPMF. Participants then received interventional DM education about the American Association of Diabetes Educators 7 Self-Care Behaviors (AADE7). Physicians and health educators conducted educational interventions according to patient’s interests and specific requirements to enhance the patient’s self-management abilities. The interactive mobile physician dashboard system improved patients’ willingness to engage in continual health education.

The patient intervention activity and integrated data were recorded in the physician dashboard system. Data were analyzed using IBM SPSS Statistics version 19.0. A paired t test was used for pre- and postintervention assessment analysis, and the t test and chi-squared test were used for group comparisons.

This study indicates that overall patient compliance rate increased after the IPMF-based educational intervention [4,5,12,26]. Participants who completed postintervention assessments were older than 60 years. This finding demonstrates that age was not an impediment to mobile device usage.

The use of IPMF to improve patient compliance may help patient engagement and customer relationships and retain patients [12,13,33]. The personalized interactive education approach used in this study was associated with greater overall compliance with recommended health behaviors [1,19,27,34,35].

In addition, education level was not a barrier to using a mobile device (such as an iPad) for wellness education. Mobile phones are increasingly prevalent, but only a few participants owned a mobile phone. This low mobile phone ownership may have been correlated with average participant age [25].

Key aspects of self-efficacy for patients include the ability to use a mobile app to self-monitor progress and to be accountable for goals set [33]. In this study, the IPMF system may be particularly helpful for T2D patients with hypertension or who are overweight.

From another perspective, older people may encounter problems in engaging with interventions to reduce the risk of type 2 diabetes; the strong representation of this group in this study sample may be helpful for addressing relevant concerns [3]. This study indicated the participants were willing to join and complete the study regardless of whether they owned mobile phone, indicating that older patients can benefit from IoT resources [30]. This finding indicates that the intervention stimulated behavior changes in participants with greater interest in self-management.

A strength of this study was the inclusion of individual personalized goals and an execution plan; this study was not treated as a research study. All patient feedback data came from the real world as opposed to being generated from a research setting.

A limitation inherent in mobile app interventions is that patient responses may differ from those of normal behavior. This analysis was limited after follow-up, and the findings may not be applicable to patients with type 2 diabetes with low motivation to use a mobile app for self-management.

The results of this study indicated that persistent personal health management is correlated with positive feedback from physiological data [18,19]. In recent years, considerations of ethnic differences have served as the key element for providers designing treatment plans [7,16]. For example, investigating Chinese people’s lifestyles—including elements such as diet, exercise, and stress-coping strategies—requires a wide variety of data obtained from user feedback [9,39].

Developers must address and acknowledge barriers to the use of mobile app interventions, such as low computer literacy and low health maintenance motivation [5,21,33].

This study found that (1) self-efficacy influences the factors of the IPMF intervention method and behavioral change model, and (2) a mobile app can be used to improve personalized care plans.

Despite the potential benefits of using a mobile app to support self-management of chronic diseases, we found that individual preferences can have a marked impact on personalized care plans and goal setting. As the mobile app for self-management in this study was developed and implemented rapidly, further work is necessary to produce a specific service that can enrich the quality of patient-centered care.

On the basis of this study, future related studies should consider the effects and continuity of behavioral changes resulting from interventions for enhancing self-efficacy. This research and the mobile interactive system will be revised and redeployed among an ethnic Asian population (eg, patients living in the United States with high motivation to use mobile apps to manage their personal health), and further evidence will be gathered to validate the effects of its use in health care self-management.