This mixed methods cross-sectional study assesses UX outcomes of a chronic pain self-management mobile app, including acceptability through content quality assessments by users and use through global use metrics.

This study was submitted to the Geneva Cantonal Commission for Research Ethics and was deemed to fall outside the scope of the Swiss Federal Act on Research involving Human Beings (Human Research Act); therefore, formal ethics committee approval was not required. All participants were informed about the purpose of the study, and informed written consent was obtained prior to participation. Participation was voluntary, and participants could withdraw at any time without consequence. All data were collected and pseudonymized and handled in accordance with applicable data protection regulations. No personally identifiable information was collected, and confidentiality of participants was strictly maintained. No financial compensation was provided to participants for their participation in this study.

Dolodoc is a mobile app developed to assist patients in self-managing chronic pain. It provides self-management strategies and encourages users to maintain a digital journal to dynamically monitor their pain and related quality-of-life indicators. The development and implementation of this digital tool have been described elsewhere [30], and they follow UX design principles for optimization. The core content of the app consists of a library of 84 evidence-based self-management strategies that users can test in real life, tracking their effects within the app on various pain severity indicators, including quality-of-life measures. More details about the app, including illustrative examples, are available in Multimedia Appendix 1 and in previous publications [23,30]. The strategies were developed through consultations with a focus group of patients with chronic pain and professionals specializing in chronic pain care. Participants first discussed the impact of pain on their daily life in various contexts. From these discussions, 7 quality-of-life dimensions emerged: mood, social support, sleep, work, intimacy, relaxation, and daily activities. Drawing on these dimensions and scientific evidence, a team of psychologists developed motivational pain-management strategies: each quality-of-life dimension corresponds to 4 to 20 strategies, totaling 84 self-management strategies. The self-management strategies were reviewed at the time by a pain specialist and a communication expert.

Descriptive and inferential statistics were applied depending on the nature of the data. Microsoft Excel, Jupyter Notebook [31], and Orange Data Mining (University of Ljubljana) [32] were used for data analysis and visualization.

Over a 6-month period, from January to June 2024, a total of 3589 unique sessions were recorded, with 59% (2116/3589) on iOS and 41% (1473/3589) on Android. The mean duration of each session was 3 minutes and 21 seconds (201 s; Figure 3A and Multimedia Appendix 4 for percentile distribution). On average, 10.39 page views occurred per session. We observed 0.07% (25/37,255) missing values in the page titles data because these events were not counted as page views. Unique sessions containing only missing values were excluded from analysis (n=7) The bounce rate (ie, the percentage of sessions where only 1 page view occurred) was 23.34% (836/3582).

Among Android users, 1473 sessions were observed among 802 unique visitors. A total of 60.6% (486/802) of these visitors visited the app only once, and the mean number of sessions per visitor was 1.84, although a few outliers displayed more than 10 sessions during this period (Figure 3B and Multimedia Appendix 4 for percentile distribution). As mentioned in the methods section, iOS data were not available for this analysis.

Table 1 shows the top 10 most visited pages for all users. When navigating the app, the most visited page was the “first launch page,” showing that a substantial proportion of users used Dolodoc only once. Next, visitors most often viewed the home page and the page where they could monitor quality-of-life dimensions through a 4-point visual scale. Interestingly, the page where the strategies are compiled was the first action users took in the mobile app, because the 3 previous pages are automatically displayed upon connection. These results are consistent with the architecture and purpose of Dolodoc, guiding users toward self-observation and exploration of the included strategies for self-managing chronic pain.

Interestingly, the time spent on each page of the app does not follow the same architectural logic as the most visited pages (Table 2). Users spent the most time on less accessible pages (ie, pages that required more clicks to be accessed) of the app, such as a page where they could describe and add characteristics to their pain named “add your pain,” a page containing useful links, and a page where they could add their current medication named “add medication.” These results showed what users actively sought as they navigated the app, providing valuable feedback and insight into their expectations and needs.

This study provides a detailed evaluation of specific dimensions of UX, focusing on content acceptability and real-world use patterns, rather than on interaction design or interface usability. The population that evaluated the app’s content was unevenly affected by pain when assessing its impact on 7 quality-of-life dimensions, highlighting the variability in the experience of chronic pain and the need for personalized approaches to pain management. Self-management strategies included in Dolodoc were perceived as engaging by users, based on 5 acceptability criteria: understandability, motivational aspect, feasibility, relevance, and alignment with quality-of-life dimensions. This result was not straightforward, because although the app was developed using a user-centered design, end users were not involved in developing this content [23]. Although participants in the acceptability study perceived the content as acceptable and engaging, real-world use data showed that 60% (486/802) of users discontinued the app after a single visit. This suggests a potential discrepancy between perceived acceptability in an experimental context and actual acceptability in real-life conditions. However, users were actively engaged with less accessible features that required deeper navigation and/or active data entry, which may explain the longer visit durations. Personal pain and medication descriptions underscore the importance of being active with the app, and the need for personalization to enhance engagement and motivation [27]. Pages offering educational and social support were also among the most visited in the app, underscoring their relevance. Overall, this mixed methods pilot study provides preliminary insights into UX parameters of a mobile app for the self-management of chronic pain, highlighting directions for future research rather than serving as a comprehensive review of the topic.

The main limitation of the acceptability study is the relatively small number of patients (N=33), which limits generalizability; therefore, the results must be interpreted as preliminary. Additionally, patients were recruited primarily by health care professionals, which may lead to selection bias, with the most motivated patients participating. Further exploration of the difference between active and passive approaches would have brought valuable information to the study. However, our sample was diverse for demographics and pain-related characteristics.

A major limitation of this study is the lack of qualitative feedback from users who discontinued the app early. While the use analytics indicate that approximately 60% (486/802) of users accessed the app only once, the reasons underlying this early dropout remain unknown. Without direct user feedback, it is not possible to determine whether the discontinuation was due to usability issues, a lack of perceived relevance, insufficient onboarding, or unmet expectations. This limits our ability to fully interpret acceptability in real-world conditions, particularly for users who did not engage beyond an initial session. Taken together, these findings indicate that good content acceptability alone is not sufficient to ensure sustained use or even initial adoption, highlighting the multifaceted nature of UX in mHealth interventions.

Another limitation of the acceptability study is the assessment of tailored acceptability criteria, instead of a standard measurement instrument, such as the user version of the Mobile Application Rating Scale (uMARS) [33]. However, our study focuses on a French-speaking population, and at the time of the study’s conception, the uMARS had not been translated into French. As mentioned in the methods section, the main limitation of the use study was excluding iOS users from the number of sessions per visitor calculation. Furthermore, interpreting bounce rate and session data requires caution, as these metrics do not directly reflect user satisfaction or use.

This study highlights the challenges of adopting a pain self-management app and identifies an important gap between acceptability and long-term engagement. This rare combination of real-world and experimental UX observations in the context of chronic pain provides useful, patient-centered insights that could help improve Dolodoc and inspire similar mHealth innovations. Indeed, postproduction UX studies in the context of mHealth solutions for chronic pain are sparse and usually based on small user samples, lacking real-world use data [24,34-36]. Using both ecological and survey-based data helps mitigate the “trial bias,” which has been previously described as an important confounder in user engagement studies of unguided eHealth interventions [34].

Users were satisfied with the content of this mHealth solution, suggesting that the user-centered design that guided its development [23] led to good acceptability of Dolodoc, similar to other mHealth interventions for chronic pain. In comparison, other therapeutic modalities, such as cognitive behavioral therapy [37], mindfulness-based interventions [38], physical activity [39], and multimodal group therapy [40], are also generally well accepted by patients experiencing chronic pain. However, real-world use habits in our study demonstrate the need to enhance long-term engagement further. This study does not identify specific barriers to adoption and sustained use; however, poor long-term engagement has already been described for other mHealth interventions in chronic disorders [16] and chronic pain [41], highlighting important barriers to use, such as the lack of reminders [26,42].

The diversity of pain profiles among our sample of users in the acceptability study, and the fact that global users actively spent most of their time on pages that were dedicated to personalization features, reveal the importance of personalizing pain-management mHealth solutions, in accordance with previous reports [25,26]. The personalization of mHealth apps positively influences users’ trust, thereby increasing the likelihood of their adoption [43].

Despite the limitations mentioned, the study provides a strong initial assessment of the acceptability and use of Dolodoc. Although these UX results are specific to this mHealth solution, they align with reports on similar interventions regarding use [44,45] and acceptability [46].

On the basis of the results of this study, it seems relevant to add features to improve engagement over time, such as a reward system and additional features [26]. Additionally, it would be useful to revise the app navigation logic to facilitate access to the most frequently used pages. Future improvements could also help us collect UX data longitudinally by letting users rate content within the app using like or dislike buttons.

Future research should prioritize qualitative investigations to understand the reasons for early discontinuation. Semistructured interviews or focus groups involving users who abandoned the app after initial use could provide valuable insights into barriers related to usability, perceived value, emotional experience, or contextual factors. Such qualitative feedback would complement use analytics and content acceptability measures, enabling a more comprehensive understanding of UX and informing more targeted design improvements.

However, optimizing digital therapeutic alliance [47] remains a key challenge for the future of mHealth. In this regard, digital navigation follow-ups have emerged as a potential solution, grounded in human interactions, to maximize patient engagement with mHealth solutions [48] and increase their clinical impact [49].