This study was structured as a mixed methods prospective cohort study over a 3-month period. In this study, we integrate quantitative app use data with rich qualitative insights from participants and podiatrists. The study focused primarily on the used qualitative methods, which were based on open-ended responses from surveys and narratives from interviews. The study focused on usefulness, user experience, and engagement with the Feetchecker app. In this study, we conceptualized engagement as “effective engagement” [25,26], in which we describe the total number of times the app was used, the duration (for how long the app was used), frequency (eg, patterns of use), and depth (eg, use of the knowledge clips) [27]. Additionally, we consider user experience as satisfaction, acceptability, credibility, perceived impact, and usability [28].

The Feetchecker app (Figure 1) was designed through an iterative, user-centered co-design process, including multistakeholder cocreation sessions with patients, therapists, and a variety of other health care professionals. This process is described in detail in Pelders et al [24]. This led to the final version of the Feetchecker app, which was developed into a fully functioning iOS and Android app [29]. To use the app, users need to use a specific patient number, which they obtain from the clinic that uses the Feetchecker app. In addition, patients need to enter their date of birth; this acts as an additional identity check. Data and pictures from the Feetchecker app are linked to the right patient record in the electronic health record (EHR) system so the podiatrist can identify the patient for follow-up when required. The core features include the 9-question feetcheck, which asks questions such as whether patients see a red spot, cut, or sense temperature differences (Multimedia Appendix 1), the picture-taking module, a timeline with the history of completed checks and pictures, curated knowledge clips created with podiatrists specialized in diabetes, setting reminders, and frequently asked questions. A more detailed description of the intervention and its features can be found in Multimedia Appendix 2. Podiatrists let patients decide their own check routines. Most podiatrists instruct patients to “use it when you experience a problem.”

The Feetchecker app does not fall under the European Medical Device Regulation [30], as the app has no intended medical purpose and does not perform any medical action (eg, advice and diagnostics) on the collected data different from storage, archival, and communication.

Ethics approval of the study was granted by the institutional ethical committee of the Fontys University of Applied Science and approved by the board of the clinic. All participants (patients and podiatrists) were provided with an information letter describing study details. Informed consent was obtained from all participants. This included consent to publish anonymized research data such as interview results and Feetchecker pictures. All study data were pseudonymized using a patient code to ensure confidentiality. Research data required for the study were password-protected and stored at the research institute’s secure storage facility, following General Data Protection Regulation standards. Study data were deidentified after the study was finished. Only the research team had access to the decryption key. The research team had no access to the EHR system of the clinic. Participants did not receive any form of compensation for participation. Throughout the study period, a dedicated phone number and email address were available for patients to contact the researchers for technical support or with any queries related to the study or the Feetchecker app.

The mixed methods data collection strategy consisted of demographic baseline characteristics of the patients, an intake and poststudy survey and semistructured interviews with patients and podiatrists, and finally, quantitative data from the app server data logs. All qualitative data were first analyzed in Dutch and subsequently translated into English for publication. Analyses of the qualitative data were done by 2 of the authors (RH and DP).

In total, we recruited 70 patients through the podiatry clinics, of which we could include 36 patients, who then signed the consent form and started the study. In total, 24 patients (male: n=20, female: n=3, unknown: n=1) with an average age of 71 (SD 8.6) years completed the exit survey and concluded the study. Detailed demographics per participant can be found in Multimedia Appendix 7. All patients successfully installed the Feetchecker app on their personal smartphones. Of all patients who completed the study, 10 consented to a follow-up interview, of which, we could reach 5 for an interview over the phone, which lasted ~20‐40 minutes. All patients had type 2 diabetes mellitus. Most patients had a high risk of developing foot complications, indicated by the Care Profile, which is a classification of care needs further explained in Multimedia Appendix 3. Of the 24 patients, we had 9 very high, 12 high, 1 medium, and 2 unknown risk profiles. The average study period for the patients was 111 (SD 19.9) days. Not every patient filled in the exit survey immediately after 3 months, causing their study period to be longer than 3 months, hence, the variability in study duration.

Reasons for dropping out were mostly due to not responding to invitations and reminders to complete either the start or exit survey and reminders for starting the study after recruitment. One patient indicated that they no longer wanted to participate due to a lack of digital literacy.

Additionally, we interviewed 5 experienced (8‐20 years) podiatrists, familiar with treating patients with diabetes, who helped recruit patients for the study and who had experience with the Feetchecker app.

This paper describes the results of a 3-month mixed methods evaluation study of the Feetchecker app with 24 patients with type 2 diabetes and at risk of DFU. We examined app usefulness, engagement, and overall user experience through a synthesis of quantitative and qualitative data.

Overall, this study found that the majority of patients perceived self-monitoring the health of their feet with Feetchecker as useful and empowering. This study showed that the app helped most patients play a more active role in checking for potential problems, and patients saw real value in sharing the Feetchecker pictures with podiatrists. Previous efforts in using pictures showed that people perceived this form of self-monitoring as valuable [19]. This study contributes to the literature by identifying an effective and accessible way of self-monitoring potential foot problems in people’s everyday routines, focusing on at-home use. The results showed that most patients frequently conducted feetchecks in their own home and recognized that the accessibility of a specialized podiatrist through an easy-to-use app was useful. While the majority of patients did have a look at the timeline and information features in the app, most did not actively use these over longer periods of time.

While engagement in using eHealth apps often drops off after the first week [32,33], we observed good retention over time and believe that most patients had effective engagement with the Feetchecker app; they were able to use the app in a way that fits their needs. Analysis of data logs from the app identified 3 main engagement patterns with the app: continuous, frequent, and no to little engagement. While some patients stopped their use of the app altogether, for others, a period of low use perhaps does not indicate “engagement failure,” as Ploderer et al [19] describe, but rather, a sign of changing needs over time, which is more in line with the conceptualization of effective engagement by Yardley et al [26]. How patients were instructed in using the app by their podiatrists is likely to be important to the level of engagement. Kilic and Karadağ [21] asked patients to use the app every time they measured blood glucose and send foot observations daily. Ploderer et al [19] asked patients to use MyFootCare each time they changed their wound dressing away from the clinic to introduce a routine to the task. In our study, podiatrists were free to recommend how patients should engage with the app. In practice, podiatrists often promoted steady checking routines (eg, every week) that fit patients’ needs and routines. We found that podiatrists did not mind that patients used the app less after a while. As patients learned how to inspect their own feet by using the app and then would keep doing so regularly (without the app), podiatrists understood that patients would use the app when necessary.

Research on promoting self-management practices for patients with diabetes showed that just handing patients flyers or a website is insufficient for engaging them in the intervention [34,35]. In our study, we find that having a podiatrist help explain the app in person and follow up if something is not right can help with starting and continuing use of the Feetchecker app.

In our results, we found that some patients experienced the picture-taking process as somewhat difficult due to either physical limitations or their digital skills. This sometimes results in blurry, under- or overexposed pictures. Potential solutions to this are shared in work from Anthony et al [36], who saw the best results in picture-taking by using tools such as a selfie-stick or asking another adult to help. However, Kolltveit et al [15] share that despite having health care professionals take the pictures, they found that a lack of space, light, and a purposeful working position seemed to impede the use of the software. Similarly, Swerdlow et al [20] showed promising results from their “Foot Selfie” system, but this required a specialized tool. Since all these systems have promising advantages, as well as disadvantages, we argue for a more deliberate approach that tailors the choice of intervention to the patients’ needs and capabilities. Furthermore, the necessary quality of pictures depends on the intended use. Pictures for monitoring wound healing might need to have a good and clear quality (eg, for the algorithm to identify edges), while a trained podiatrist can already use a simple homemade picture with lesser quality to judge whether a patient should be seen or not.

This study was based on a mixed methods evaluation with 24 patients and 5 podiatrists, which limits the conclusions we can draw on the effects of this intervention on actual long-term prevention of DFU. The quantitative results are used descriptively, as this study design was not intended to provide results on clinical outcomes or effects on DFU formation. Our strengths are in real-life deployment for 3 months in everyday situations, leading to high ecological validity. This was highly informative of app engagement, use cases, types of users that emerge, and how patients used the app in their everyday life.

While we did not observe any technical problems with the Feetchecker app system, one patient did not enter their patient number correctly in the app, which makes it impossible to link the output of the app to the EHR. For future work, we aim to work on more robust authentication methods, as well as providing patients with a leaflet where the podiatrist can write down the patient number.

While a big benefit of the Feetchecker app is the ability for remote monitoring by podiatrists through submitted pictures, the app has no 2-way communication. If a podiatrist wants to ask for a new picture, they need to call the patient. We aim to investigate the potential to use 2-way communication features to reduce unnecessary calls or visits to the clinic.

Many people with or at risk of DFU have difficulties with the usability of mobile apps [37]. Future studies should focus on improving the usability of the Feetchecker app to be more inclusive toward patients with visual impairments and limitations of physical mobility and should offer multiple language options. As Mannheim et al [38] recommend, future work should include this target group in the design process. Furthermore, Wang et al [39] suggest to further investigate optimal formats and the frequency of engaging patients as well as build better tailoring of reminders.

Similar to skin cancer screening apps such as Skinvision [40], we envision that the picture-taking feature of the Feetchecker app might be enhanced through the use of algorithms that can recognize whether an image is of sufficient quality for effective inspection by the podiatrist. Recent research has shown potential in using artificial intelligence systems to support triage or diagnostics [41]; however, these systems are not accurate enough to be implemented in real-world use [42,43]. While we see the advances of artificial intelligence–powered screening systems [44-46], we must not lose sight of having a human-in-the-loop to make sure people are being listened to, as we showed that this human connection was a primary positive result of this study. Future studies will also be needed to determine if the Feetchecker app can reduce the need for in-person clinic visits, the incidence of severe or infected DFUs, and eventually reduce the cost of medical care.

The Feetchecker app supported patients in self-monitoring risk factors associated with DFU through routine checks and quick contact with a health care professional in case of a potential issue. Patients generally described a positive user experience and considered the app helpful. While mHealth tools are not for everyone, user engagement for many patients was high and shows that such apps can offer support for patients able to use them. Furthermore, podiatrists appreciated how the pictures enabled accessible and quick contact with patients. Supporting health care professionals in integrating mHealth tools in their regular care is important toward adoption and sustained use. Future work needs to be done to better understand how mHealth can be leveraged to better integrate into the everyday routines of patients at risk of DFU. This is crucial to promoting self-monitoring of foot health, as it may shift the focus from managing wounds to effectively preventing the need for care in the first place. Innovative solutions like these are urgently needed to enhance current systems for the prevention and care of diabetic foot ulceration in a world where the number of patients with diabetes is increasing quickly.