The knowledge base of the mobile app was based on French national recommendations for healthy eating, physical activity levels, and limiting alcohol and tobacco consumption [22-25]. We analyzed these recommendations and extracted the necessary variables for the assessments of each behavior (eg, for healthy eating, we extracted 23 variables, each with 7 possible answers). We then combined the variables and their responses to establish a list of patient behaviors. For each patient behavior, we calculated a qualitative score extending from “very good behavior” to “very bad behavior” (based on guidelines). We then associated personalized recommended actions with these behavior scores (eg, “do the equivalent of 30 minutes of physical activity per day”) and digital health interventions (eg, “activity tracker for self-monitoring of daily activity”). The identified patient behaviors and their associated qualitative scores for risky behavior, personalized recommendations, and digital health interventions were then implemented within the app.

The mobile app was ergonomically designed according to usability principles [26,27], and the graphical interface was implemented with Gluon mobile technology [28]. The navigation system was based on a 3-step process: (1) assessment of each risk factor with specific questionnaires, (2) automatic assessment, by the app, of risky behavior, and visualization of recommended and personalized actions, and (3) suggestions for digital health interventions. The interface was designed according to the following usability principles [26,27]: simplicity, naturalness, and effective use of language by using concise, appropriate, and understandable language; consistency, minimizing cognitive load, and efficient interactions by dividing the navigation into a 3-step process and adding a clear taskbar and meaningful colors; and effective information presentation, by limiting the amount of text and using comprehensible icons (eg, a wineglass icon was used to represent alcohol consumption). The consideration of usability principles in app design can help reduce the user’s cognitive workload and increase confidence in the app [29].

We carried out an evaluation of the quality of the app by asking patients identified as potential future users to test the app. We organized 20 online sessions, each with 1-5 users. Each session began with a demonstration of the app and then testing of the app by the study participants. The opinions of the study participants were then collected in an electronic form and through an open discussion. The study protocol was validated by the appropriate Inserm ethics committee (CD/EB 20-023, 20-660, IRB000388, IORG0003254, FWA0005831).

Study participants were recruited online via a dedicated website and by word-of-mouth from April 15 to June 15, 2020. For participation, subjects had to be older than 18 years and not treated for CVD (eg, stroke, coronary artery disease). They also had to sign an online consent form and agree to participate without compensation.

The evaluation was performed online and lasted from 30 minutes to 1 hour, depending on the session. It proceeded via 5 successive steps:

For sociodemographic data and for each item of the uMARS scale, we calculated the percentage of each response. For the free discussion, data were analyzed, broken up, and combined into similar themes. Each theme was mapped to one of the categories of the uMARS scale (eg, engagement, functionality).

In total, 52 individuals tested and evaluated the app: 28 women and 24 men. The study participants were aged between 18 and 69 years and had diverse professional profiles. Most participants were familiar with digital devices and used a smartphone, computer tablet, or laptop computer at least once per week. Half of the participants declared that they had already used a connected object, mostly with an app, but only 13% (7/52) declared that they had already assessed their cardiovascular risk in this way (Table 1).

The mean total score for app quality was 4 on the uMARS scale (maximum mean score of 5, Figure 2).

For engagement, the mean uMARS score was 3.7 (SD 0.7). More than three-quarters of the study participants found the app interesting and appropriate for the target audience, and more than half found it fun and entertaining (Table 2, comment 1). One study participant said that the app was appropriate and accessible to everyone, including older adults who were less used to apps (Table 2, comment 2). However, fewer than half of the study participants found the app sufficiently interactive and customizable. Some study participants suggested making the app more interactive by adding a chatbot and the possibility of speaking to an expert, together with additional reminders and alerts concerning the recommended behavior changes. A few study participants also suggested delivering a daily preventive message to help educate users (Table 2, comment 3).

For functionality, the mean uMARS score was 4.4 (SD 0.6). Almost all the study participants appreciated the navigation system and found the app easy to use. More than 80% of the participants also appreciated the performance and design of the app (Table 2, comment 4). Study participants said that they found the app intuitive and very easy to use, although 1 person suggested adding a tutorial (Table 2, comment 5). None of the study participants declared having experienced technical problems during the evaluation.

For aesthetics, the mean uMARS score was 3.7 (SD 0.6). More than three-quarters of the study participants found the layout appropriate. However, less than half felt that the graphics were of sufficiently high quality and visually appealing. Some study participants found the design and colors appropriate for a health app, whereas others felt that the colors were insufficiently attractive (Table 2, comment 7). Some study participants suggested improving the aesthetics of the app by adding more 3D pictures, video animations, summaries, pictograms, (eg, arrows), and emojis (Table 2, comment 6).

For information, the mean uMARS score was 4.2 (SD 0.5). More than three-quarters of the study participants found the app content appropriate, relevant, concise, and comprehensive (Table 2, comment 8). Study participants said they found the content interesting, clear, and detailed. Almost all the study participants also appreciated the visual elements used to display information and had confidence in the credibility of the source used to build the app content (Table 2, comment 9). One study participant said that this app inspired more confidence than those available on app stores (Table 2, comment 10). Some study participants suggested improving the questionnaires used to assess cardiovascular risk by adding more accurate questions about family history, tobacco and alcohol use, and physical activities, for example, or by adding questions about drugs and making the eating habit forms easier to complete. Other study participants suggested adding more information about dairy products, sports activities, and recipes. One study participant suggested adding explanations about the workings of the human body and physiology (eg, heart physiology).

For subjective opinion about app quality, 92% (48/52) of the study participants said they would recommend the app to several people, 56% (29/52) of the study participants also declared that they would use the app more than 3-10 times in the next 12 months, and 59% (31/52) said that they would not be willing to pay for it. Finally, 88% (46/52) of the participants gave more than 3 stars for the app when asked to rate it (Table 2, comment 11) (Figure 3).

For perceived impact (Figure 4), 80% (42/52) of the study participants declared that the app had made them more aware of the importance of addressing health behavior, and 77% (40/52) said that it had helped to improve their awareness of their behavior. Some study participants explained that the automatic assessment of behaviors had improved their understanding of their lifestyles and attitudes. One study participant also said that this awareness could encourage changes in behavior to prevent health disorders (Table 2, comment 12). Overall, 67% (35/52) of the participants said that the app had changed their attitudes toward improving health behavior, and 65% (34/52) said that they were more motivated to change their habits by playing more sports or eating more healthily, for example. One study participant said that the app could help him to achieve his personal goals in terms of behavior changes. Furthermore, 61% (32/52) of the participants thought that the app would lead to improvements in health behavior, and 73% (38/52) thought that the app would encourage people to seek further help if required (Table 2, comment 13).

For global satisfaction, 55% (29/52) of the study participants were fully satisfied with the app, and only 12% (6/52) were not. The others were neutral. Almost all the study participants found that the recommendations delivered were clear.

For areas of improvement, some study participants suggested extending the app to cardiovascular follow-up. They suggested adding new functionalities for tracing behavioral changes such as functions for recording meals or an activity tracker for recording physical activity. They also suggested displaying the data collected through their other connected objects (eg, sleep data collected with connected watch) within the app. One participant also suggested adding a virtual avatar to assess progress in behavioral changes over time.