A series of descriptive analyses were performed to depict the characteristics of the sample. In order to evaluate the primary endpoint, the mean score and its SD were calculated for each MARS subscale (engagement experienced while using the app; functionality; aesthetics; quality of the information received; subjective perception of the app quality; expected impact on knowledge, attitudes, and probability to change user behaviors) at 3 months; in addition, the total number of times the PainRELife mobile app was accessed by each participant was determined.

Further, a new variable named total app quality was created using the mean values of the engagement, functionality, aesthetics, and information quality MARS subscales. The final measurement of app quality was the average value of the 4 means [43]. Pearson correlational analysis was performed among all self-reports used (NRS, PSEQ, SDM-Q-9, MARS) and the total number of times the app was accessed during the 3 months of the study. A repeated measures ANOVA was performed to detect variation in pain intensity (NRS) from T0 (baseline) to T2 (3 months). Further, a new dichotomous variable named “frequency of use” was created considering the entire number of times the app was accessed (mean 22.92, SD 15.60; range 2-73) and the lowest number of times the PainRELife mobile app needed to be accessed (21 times) by participants to finalize the study’s tasks. The “frequency of use” variable permitted dividing the participants into groups based on higher and lower frequencies of access. Further, a Student t test was run to evaluate the difference between the “frequency of use” and PSEQ, SDM-Q-9, and MARS scores. Data were analyzed using SPSS version 26.0 (IBM Corp). 

This study received ethical approval in December 2021 (R1597/21-IEO 1701) from the Ethical Committee of the IEO and respects the Declaration of Helsinki and Good Clinical Practice Guidelines. All participants read and signed the informed consent form, which encompassed a comprehensive and exhaustive explanation of the primary and secondary endpoints, study procedures, and possible risks and benefits. Participants were not compensated and were able to withdraw their participation at any time during the study. Concerning privacy and confidentiality protection, all data collected were deidentified and anonymized, complied with national data protection legislation, and will be stored in the IEO databases for 10 years.

The authors affirm that human research participants provided informed consent for publication of their data.

The sociodemographic, cancer, and treatment characteristics are listed in Tables 1-3.

The total MARS score (ranging from 1 to 5) provided overall medium-to-high mean values for each subscale (range 3.31-4.18; see Table 4) and a mean total app quality score of 3.90 (SD 0.506), suggesting generally good usability as evaluated by the participants. This was also confirmed by the total number of times the participants accessed the app during the entire study (mean 22.92, SD 15.60; range 2-73). In particular, 3 of 5 subscales had the highest scores: functionality (mean 4.14, SD 0.630), information (mean 4.18, SD 0.608), and behavioral change (mean 4.05, SD 0.666).

On the functionality subscale, 57% (15/23) of the participants judged that the mobile app is straightforward to use. Moreover, 91% (21/23) of the participants affirmed that the interactions are reliable and intuitive (ease of use: 8/23, 35% agree; 13/23, 57% strongly agree), positively evaluated the design (gestural design: 8/23, 35% agree; 12/23, 52% strongly agree), and evaluated the navigation properties as good (navigation: 12/23, 52% agree; 8/23, 35% strongly agree). However, some slight uncertainties were observed regarding the general performance of the mobile app, specifically moving between pages and sections (performance: 8/23, 35% undecided; see Table 5).

Concerning the distribution of responses in the information subscale, 78% (18/23) of the participants reported that the information in the mobile app is evidence-based (information: 9/23, 39% agree; 9/23, 39% strongly agree), relevant, focused on chronic pain in breast cancer and its management during the disease clinical pathway (quality of information: 9/23, 39% agree; 11/23, 48% strongly agree), and trustworthy (credibility: 22/23, 96% strongly agree). Further, the amount of information (quantity of information: 7/23, 30% agree; 9/23, 39% strongly agree) and how the information is reported using different setups (visual information: 11/23, 48% agree; 9/23, 39% strongly agree) were considered positive by the participants (see Table 6). Finally, most participants stated that the mobile app's goals are achievable (goals: 11/23, 48% agree; 3/23, 13% strongly agree), even if 30% (7/23) reported some concerns.

Last, the distribution of responses in the behavioral change subscale revealed that 83% (19/23) of the participants strongly agreed that the mobile app had improved awareness about the issue of chronic pain in the cancer disease pathway, and 70% (16/23) agreed that the app increased chronic pain–related knowledge. Likewise, 70% (16/23) of the participants reported that the mobile app might support attitudes toward chronic pain (attitudes: 9/23, 39% agree; 7/23, 30% strongly agree; see Table 7).

In addition, most of the participants reported that the mobile app would potentially be helpful to bolster help-seeking behaviors (help seeking: 5/23, 22% agree; 9/23, 39% strongly agree) as well as support an intention to change (intention to change: 5/23, 22% agree; 9/23, 39% strongly agree). Still, 39% (9/23) showed concerns about the capacity to transform intention into a fundamental behavioral change (see Table 7). Overall, participants judged the app to be well-targeted (engagement subscale: mean 3.31, SD 0.617) and the app’s layout to be adequate (aesthetics subscale: mean 3.98, SD 0.849); likewise, the overall subjective quality was adequate (subjective quality subscale: mean 3.50, SD 0.494).

According to the Student t test, younger participants used the mobile app less (mean 44.15, SD 7.11) than older participants (mean 50.33, SD 8.80; t₂₃=–1.937, P=.03; d=0.77). A difference in pain self-efficacy was observed between participants with higher versus lower frequency use (t₂₃=1.644, P=.057; d=0.65). The latter data indicate that, at T2, participants with a lower pain self-efficacy (mean 40.83, SD 14.58) used the mobile app more than participants with a higher pain self-efficacy (mean 48.46, SD 7.90).

The repeated measures ANOVA revealed that participants reported a reduction in pain intensity from T0 (mean 5, SD 1.68) to T2 (mean 3.72, SD 2.59; F_2,48=3.407, P=.04). A positive correlation was found between the total number of times the mobile app was accessed and pain intensity at T2 (r=0.458, P=.03).

No correlations were detected between the MARS subscales and PSEQ or NRS. A negative correlation was observed between the subjective quality subscale and the number of times the mobile app was accessed (r=–0.498, P=.02). Further, the engagement (r=0.445, P=.03), information (r=0.427, P=.04), and subjective quality (r=0.548, P≤.007) subscales were positively correlated with shared decision-making.

Considering the primary endpoint of this pilot study, patients with breast cancer provided a generally positive rating for the usability of the PainRELife mobile app. Patients with breast cancer appreciated both the quality and quantity of the health information provided on chronic pain and how chronic pain should be managed during the survivorship trajectory. Specifically, the information in the different sections and modules were perceived as informative and comprehensible (20/23, 87%) and from credible sources of information (18/23, 78%). Most of the patients with breast cancer reported that habitual use of the mobile app helped increase help-seeking behaviors for chronic pain (14/23, 61%), their general attitudes toward chronic pain, and their willingness to identify preeminent strategies for managing chronic pain. These results are particularly noteworthy considering that many studies have suggested that chronic pain syndrome in patients with breast cancer is commonly undiagnosed and not often considered by oncologists [8]. In addition, many cancer patients report poor knowledge about cancer-related chronic pain, available interventions, and possible health system resources [44].

Furthermore, the positive correlation between the total number of times the mobile app was accessed and pain intensity (P=.03) might indicate that patients with breast cancer who had a higher pain level might have utilized the mobile app to find evidence-based information and strategies to self-manage their pain. This possible interpretation might be linked to the difference in pain self-efficacy observed between the participants with higher versus lower frequency mobile app use. Specifically, participants with lower pain self-efficacy used the mobile app more than participants with higher pain self-efficacy. Perhaps the participants with lower pain self-efficacy used the mobile app to find a strategy or way to manage their pain. Self-efficacy has a crucial role for patients with cancer, and studies have reported that it improves psychological well-being, reduces fear of cancer recurrence, enhances self-care, and improves management of symptoms such as pain [45,46]. Vinnikova et al [47] observed that individuals might use mobile apps to learn more about their health problems and monitor their physical and psychological status. Furthermore, considering that self-efficacy is an attribute of cancer pain self-management, the prevalent use of the mobile app by participants with a lower self-efficacy could explain the percentage of participants who reported concerns about intention to change and the capacity to transform intention into a fundamental behavioral change [48].

A second series of results are linked to the secondary endpoints. Participants reported a reduction in pain intensity at 3 months (P=.04). We argue that the use of the mobile app might have relieved the pain experience, disease, and treatment-related symptomatology as observed in other previous studies [16,19]. One noteworthy finding is related to the association between specific features of the mobile app, evaluated with the MARS, and shared decision-making. Indeed, participants who provided higher positive evaluations about engagement (P=.03), information (P=.04), and subjective quality (P=.007) also reported higher perceptions of having shared decisions along their care pathway. We argue that patients with breast cancer who feel involved in their treatment decisions are more engaged with the mobile app. For this type of patient, information retrieved in the mobile app might be used to reinforce and reiterate the ability to achieve a shared decision throughout their care pathway.

Despite the interesting and challenging results, this pilot usability study had some limitations that must be considered. The primary limitation is the relatively small sample size of patients with breast cancer and the decision to use a single group to test usability. This decision might have caused a loss of pertinent information about the patients’ perceptions of the usefulness of this digital health technology. However, our sampling strategy is consistent with the pilot study design and methodological guidelines [49-51]. Related to this point, we must also mention that the statistical significance reported for pain self-efficacy might be considered borderline (P=.057). However, the effect size is medium-to-large (d=0.65), which supports the statistical difference between the groups. We argue that the P value might be due to the small sample size [52]. We also argue that this could be a significant result that has to be further investigated in successive studies, considering the positive effect of cancer pain self-management on QoL, when compared with pharmacological treatments such as opioid consumption [48].

In addition, the “frequency of use” variable had a moderately high SD (15.60). However, the distribution of the participants between the low frequency (n=13) and high frequency (n=12) groups was homogeneous and balanced. Furthermore, we hypothesized the presence of selection bias resulting from the inclusion criteria, which required participants to have internet access and a personal smartphone. This criterion may have limited the inclusion of certain vulnerable groups among patients with breast cancer, such as older adults or individuals with lower health literacy and socioeconomic challenges who could be at higher risk of undiagnosed chronic pain. Most of our patients with breast cancer also reported medium-to-low pain during the entire study and were in the acute and extended stages of the survivorship trajectory, which might have affected the frequency of mobile app use. Indeed, even if the total number of times the mobile app was accessed was relatively high and satisfactory, in the last month of the study, some participants decreased their total usage; 2 of 25 participants used the mobile app only at enrollment. The last limitation is related to the previous one and concerns the lack of assessment of the timing of mobile app use. Indeed, only the total number of times the mobile app was accessed was collected and evaluated. These limitations have been intensely discussed in the full protocol published elsewhere, and we plan to address them in future studies [7].

The data retrieved from this pilot study evaluating patients’ experiences using a novel and integrated health ecosystem for the management of chronic pain for breast cancer survivors are consistent with other studies highlighting that digital health technologies, when developed using a patient-driven approach, might be considered valuable tools for increasing the participation of patients with breast cancer in clinical care. In addition, these tools permit the achievement of critical clinical outcomes and improvement in QoL [4,8,22]. Moreover, health-integrated ecosystems permit secondary key outcomes such as reducing the burden on health care professionals and optimizing health system resources. Finally, we argue that digital integrated health ecosystems might be important devices for improving the ongoing monitoring of physical status, psychological burden, and socioeconomic issues during the cancer survivorship trajectory.