Guided by Bowen and colleagues’ [19] recommendations for designing feasibility studies, we aimed to enroll 100 participants in the study [19]. Because these early trials are used, in part, to estimate effect size and power for future trials, we did not expect to be fully powered to detect changes in primary study outcomes. Participants were recruited nationally via internet-based strategies, including social media (ie, Facebook, Twitter), social networking sites, and online and email listservs. All recruitment methods were approved by the Arizona State University Institutional Review Board. The study was advertised as a mobile app health education intervention. We recruited self-reported patients with hematological cancer (ie, patients with MPN) because we have a progressive line of work involving patients with MPN using mindfulness approaches to reduce symptom burden (eg, yoga, meditation). However, the app was developed to be applied in app-based meditation interventions across various populations. Interested participants completed an eligibility link on REDCap (Vanderbilt University). Inclusion criteria were as follows: (1) had a diagnosis of MPN (ie, polycythemia vera, myelofibrosis, essential thrombocythemia) identified by a treating physician, (2) had access to a smartphone on a regular basis, (3) had access to reliable home internet, (4) could read and understand English, and (5) were 18 years or older. The exclusion criteria were as follows: (1) planned change in pharmacologic intervention (ie, new drug, bone marrow transplant) during the study interval (ie, 12 weeks) and (2) resided outside of the United States. If eligible, the patients were emailed a link to a video explaining the informed consent and study procedures. If interested, patients were asked to respond to the email indicating that they reviewed the video and had the opportunity to ask questions. Patients then completed an electronic informed consent delivered via REDCap prior to participation. If ineligible, patients were sent an email notification thanking them for their interest in the study and to respond if they were interested in being notified about future studies.

The study was a single-group, beta test study of a new control intervention to be used as a comparator in future studies involving mindfulness and meditation mobile apps. A total 96 participants were enrolled in the study and assigned to a health education podcast group.

All analyses were conducted using SPSS 26.0 (IBM Corp). Descriptive statistics were used to characterize participants’ app-use patterns over time, and frequency data from the satisfaction survey were used to describe participants’ perceptions of the podcasts and the prescription schedule. Changes in PROMIS outcomes and MPN-SAF TSS were assessed using multivariate analyses of variance.

Initially, 96 patients with MPN were enrolled in the study; however, 19 (19.8%) never began the intervention without disclosing their reasoning and without responding to contact attempts (see Figure 2). Of the 80 patients who participated in the intervention, 39 (48.8%) completed the entire study (ie, sustained participation through the follow-up period). There were 28 participants lost to follow-up that never responded to the three contact attempts after initial enrollment and 16 that dropped out and provided the research team with a specific reason for discontinuing the study. The reasons for dropout included content not being MPN-specific (n=10), lack of time (n=3), internet connectivity issues (n=1), or illness or hospitalization (n=2).

Enrollment by week is presented in Figure 3. Analyses included data from all participants who were enrolled in the study through the point that they stopped using the app, defined as the point at which participants did not engage with the app for any subsequent weeks during the intervention period. For example, if a participant used the app during weeks 1, 2, 3, and 8, they were included in adherence analyses for weeks 1-8 (with use during weeks 4-7 calculated as 0 minutes) but were not included in analyses after the point when they stopped using the app (weeks 9-12), at which point they were considered to have discontinued study participation.

As shown in Table 1, the sample was predominately White, non-Hispanic, and female. The average age was 56.1 (SD 10.9) years. Most participants had earned a higher education degree, were married, and had an annual household income more than US $61,000. Study noncompleters were demographically similar to those who completed the study.

The purpose of this study was to beta test a health education podcast control group delivered via a smartphone app to inform the development of a comparator that could be used in future app-based meditation intervention studies. Feasibility benchmarks were met for demand (ie, for 10 of the 12 weeks at least 70% of participants listened to at least 42 min/week of podcasts); although, the attrition should be noted as an important qualifier when considering engagement more broadly. Acceptability benchmarks were not met, as less than half of participants enjoyed the podcasts and were satisfied with the intervention. As expected, there were no changes in global health, cancer-related symptoms, or total symptom burden (ie, MPN-SAF TSS) over the 12 weeks.

Use benchmarks were set at ≥70% of participants listening to ≥70% (ie, 42 min/week) of the prescribed total weekly podcasts. In nearly all weeks (ie, 10 out of the 12 weeks) this benchmark was met. This is encouraging as a control group in which participants adhere to the study prescriptions allows researchers to control for nonintervention treatment effects such as time and attention [4]. Furthermore, participants averaged 103 min/week of podcast listening or viewing. Other studies that have tested mobile-app control groups have not reported weekly participation data [13,26]. In our research studies in which we used a commercially available meditation app in college students for 8 weeks [15] and to patients with MPN cancer for 4 weeks [16], average weekly participation in meditation on the app was ~38 min/week and ~71 min/week, respectively. Our findings are promising because the use of the podcast control app was comparable to the use levels in our intervention studies, supporting its feasibility for a control group with the same time and attention as a consumer-based mobile app meditation intervention.

Satisfaction benchmarks were not met, indicating that participants’ overall satisfaction with the app was lacking. Less than half of participants indicated that they enjoyed the health education podcasts and less than half indicated that they were satisfied with the intervention. This is likely due to the podcast topics and technical issues and difficulties that came up during the intervention related to the functionality and usability of the app. The majority of participants that responded to an open-ended question to provide additional feedback about the study reported that they thought the app content should be modified to include more MPN-specific education or higher quality podcasts. The podcasts selected for the control app were intended to provide more generic cancer-related health education and not MPN-specific education. A better understanding of the potential users’ content preferences with a more user-centered approach is necessary for future iterations of the control app, especially because the app was designed to be able to change health-education podcasts per target population being studied. A user-centered approach when developing products or software may help the user feel more at ease and make engaging with the content more intuitive, potentially improving adherence and enjoyment [27].

Over half of participants reported experiencing trouble accessing podcasts or content, or general problems with the functionality of the app. The app was not developed to be of commercial quality but was rather developed to be a “shell” of a design to be improved upon in subsequent developments after beta testing. Therefore, it is not surprising that such technical difficulties related to app functionality and accessibility were experienced. Specifically, there was a decrease in participation during week five, coinciding with participants reporting technical difficulties to the research staff. It is well known that user experience is a critical component to the success of mHealth apps and that apps must appeal to the motivations of the user [28]. For example, the unified theory of acceptance and use of technology (UTAUT) suggests that users’ expectations of how an app will perform and how much effort it takes to use the app will influence their intentions and behaviors [29-31]. Future iterations of this app will use the current findings and a model-driven approach (eg, the UTAUT) to inform the development of the next version of the app to be used as a comparator in a randomized control trial.

Importantly, there were no significant changes in health or cancer-related symptoms from baseline to postintervention. This is despite more than half of the participants reporting that they made changes to their normal daily activities because of something they learned from the podcasts. This is encouraging and indicates that this control group design is an appropriate time- and attention-matched condition that is not associated with any meaningful change in study outcomes. Education-based control groups have been used successfully as comparators in a range of smartphone-based interventions across different populations, typically without having significant effects on psychological or physical outcomes [13,16,26]. This indicates that an education-based control group could be appropriate for use in efficacy trials without producing changes in primary outcomes that are physical or psychological in nature. Future iterations of the app will be evaluated as a comparator in studies with other populations.

This study is not without its limitations. First, attrition during the study was high, such that only 51% (41/80) of participants completed the full 12-week intervention. Given that the podcast app use (minutes listened per week) was high even when including use of participants who did not complete the study, future research should collect more nuanced data on participant satisfaction to improve podcast app user experiences and reduce attrition rates. For example, the lack of a postintervention interview did not allow for deeper qualitative analyses into what participants liked and did not like related to the app. This qualitative data could have been a useful addition for gathering deeper insights into user satisfaction and informing future development of the app. Second, the sample was predominantly White, non-Hispanic, and female. This is not representative of the general population of patients with MPN and, more importantly for the purpose of this beta test, does not capture feedback that reflects the experiences of different genders, races, and ethnicities. Future feasibility research should aim to include more diverse samples to gather more representative feedback. Third, there was no comparative group, and this must be considered when analyzing the results. Fourth, offering a US $25 digital gift card for completion of all questionnaires could contribute to higher engagement rates. Researchers may need to account for an impact on their participant engagement if they do not provide similar incentivization. Finally, the use of cancer-specific educational content would limit the app as currently developed to use in cancer studies only. However, the content of the app can easily be adjusted to fit the needs of different populations while maintaining the integrity and functionality of the app to match the Calm meditation app.

Although this app was originally designed to be a comparative control app for app-based mindfulness meditation interventions involving Calm, the app is not reliant on the use of Calm and may be used as a control app in future app-based interventions. Currently, this app is not available to other researchers because it was only developed as a template to then create an improved comparator app for future mindfulness meditation app-based interventions based on the data collected. Once the app can be determined feasible, future interventions will be developed to assess app-based mindfulness meditation interventions as compared to the control app to assess many aspects of health and well-being across various populations.

In summary, a 12-week mobile app health education podcast met the demand benchmark but not the acceptability benchmark for feasibility in a sample of patients with hematological cancer. Using the mobile app health education podcast was not associated with significant changes in cancer-related symptoms. Participants reported dissatisfaction with content and technical or functionality difficulties, which will be addressed in the future development of the health education podcast app. Based on findings from this study, a mobile app health education podcast may be a feasible option as a time- and attention-matched comparator condition for efficacy trials with more extensive formative research for the content of the podcasts and its acceptability by the specific population.