Our overall qualitative approach was deductive, informed by a preexisting Resilience in Illness Model that described factors associated with resilience in illness [24,25] in AYA with cancer.

Those AYA aged 13 to 25 years, previously diagnosed with sarcoma and either undergoing active treatment or within 5 years of transitioning to survivorship care, were eligible to participate. To capture a range of perspectives, we purposefully sought a representative sample of AYA in terms of age, gender, and treatment history (eg, on and off treatment, limb-salvage procedure, amputation). Inclusion criteria for parents were having an adolescent (aged 13-17 years) son or daughter enrolled in the study and being fluent in English. Given that there was likely to be variability in the living situation of young adults (eg, living independently or at college), we limited our parent interviews to parents of adolescents aged 13 to 17 years. For these interviews, we sought parents’ perspectives surrounding their adolescent’s day-to-day challenges. Health care providers currently providing clinical services to AYA with sarcoma were eligible to participate. Health care providers were purposefully recruited to represent the range of roles in a treatment team (eg, oncologists, nurses, and social workers).

The AYA and adolescents’ parents were recruited from a pediatric sarcoma clinic in Los Angeles, California. Research team personnel distributed research flyers to families waiting for clinic appointments and screened interested patients or parents over the phone. Eligible AYA and parent participants provided online assent and consent prior to beginning study procedures. Potential health care provider participants were identified from author NF’s network of providers. Author NF emailed study invitations to providers, and interested providers were contacted by research personnel. Providers gave oral consent over the phone prior to beginning study procedures.

We developed semistructured interview guides to explore ways in which a mobile-based mindfulness and social support program can address the psychosocial needs of AYAs with cancer during and posttreatment. Three separate interview guides were created for AYA, parents, and health care providers. The first section of the interview guides was informed by the Resilience in Illness Model (RIM) [24,25], which identified psychosocial factors associated with resilience in AYA with cancer (ie, spiritual perspective, social integration, family environment, coping skills, and hope-derived meaning). This model was chosen to help identify potential resilience-enhancing program components. The second section of the interview guides focused on the use of mobile technology and preferences surrounding potential program content (ie, mindfulness and Facebook social support component). First drafts of the semistructured interview guides were created by author ED and were revised with significant input from the research team. To facilitate the comparison of multiple perspectives on the same topics, the parent and provider interview guides mirrored the patient interview guide and included queries about their beliefs about AYAs’ needs for a mobile-based psychosocial intervention. Interview guides are available on request.

Two female research team members with qualitative research experience conducted the interviews. No interviewer and interviewee had a prior relationship. Patient and parent interviews were conducted in-person (at the research study’s hospital-based offices) or by phone. Health care provider interviews were conducted by phone. All interviews lasted between 45 and 90 minutes. Interviewees were offered Amazon gift cards of US$25 (AYA and parents) or US$100 (providers) as compensation for participating in the study. The research team continued to collect interview data until 10 AYA, five parents of adolescents, and six providers had been interviewed. This coincided with when the team reached a consensus regarding thematic saturation. All interviews were audio recorded, transcribed verbatim, and deidentified for analysis. Interviews were conducted from December 2016 to March 2017. The study was approved by the University of California, Los Angeles Institutional Review Board (IRB).

The coding scheme was created by the research team. We used deductive coding, informed by the RIM model [24,25]. As such, author ED created an initial coding structure based on the interview guide, with each factor in the RIM model represented. In addition, codes were created to capture beliefs about the proposed program delivery method (mobile device) and program content (mindfulness and social support). Three research team members trained in qualitative analysis coded the interview transcripts in duplicate and discrepancies were resolved with discussion. Coded interview transcripts were entered into NVivo software (version 10) [26]. Two members of the team independently reviewed data associated with each code and met weekly to discuss emerging themes, ending the discussion when three to five themes were agreed upon. Themes, supporting quotes, and exceptions to the themes were entered into an Excel sheet.

The study consisted of a single-arm, pre-post evaluation of the pilot program.

The inclusion criteria for AYAs and providers used for the formative interview phase were also used for the pilot test. Parents were not included in the pilot phase of the study. Participants included 17 AYA, seven of whom had participated in the qualitative interviews, and six health care providers.

The AYA were recruited from the same sarcoma clinic from aim 1. The same recruitment method used during the formative interview stage was used to recruit AYA and provider participants for the pilot study.

The 4-week Mindfulness for Resilience in Illness program consisted of a mindfulness mobile app, a private Facebook group, and a provider guide. The mindfulness mobile app was delivered although a customized version of an existing mindfulness mobile app (Whil Concepts, Inc) available through the Apple iOS App Store and Android Google Play Store. A total of 28 mindfulness audio exercises were selected from the Whil library for the 4-week program. The Learning to BREATHE framework [27] and themes from the qualitative interviews informed the selection of mindfulness meditations for the four weekly program themes: (1) breathe and listen to your body, (2) dealing with difficult emotions, (3) dealing with negative thoughts, and (4) being kind to yourself through challenging times (Textbox 1). Program host videos featuring two sarcoma survivors were also included in the app. The videos were used to familiarize AYA with mindfulness terms and introduce and conclude the weekly program exercises. Four weekly blog posts designed to reinforce the weekly mindfulness concepts were also created for the app and posted on the Facebook group.

The goals of the private Facebook group were to (1) solicit feedback about the app and (2) promote social support among group members. A sarcoma survivor was recruited to moderate the Facebook group, which included posting daily content and facilitating conversation among participants by responding to posts. A total of 38 messages were created for the moderator to post daily (one or two per day).

The provider guide (e-book) was developed to help providers support patients using the Mindfulness for Resilience in Illness program. The 12-page guide provided information about the rationale for teaching mindfulness skills to AYA with cancer and encouraging peer-to-peer social support, with specific strategies for supporting patients through each of the 4 weeks of the program.

Both AYA participants and parents (if the AYA was younger than 18 years) completed online assent or consent forms via a secure online survey platform. Following assent or consent, participants also completed demographic questions and psychosocial measures (described subsequently) via the online survey platform. Before beginning the program, AYA participants were invited to attend a 20-minute webinar that provided (1) an overview of how to use the mobile app and the Facebook group and (2) an outline of the 4-week program components (eg, meditations, Facebook posts) and expectations (ie, one meditation per day for 28 days and responding to daily Facebook posts). All AYA participants began the 4-week study on the Monday after the week that the webinars were held. Psychosocial measures and an acceptance test were completed immediately following the completion of the program. The AYA were compensated with a US$75 Amazon gift card.

Three people were involved in the daily administration of the Facebook group: the moderator and two monitors. The moderator was paid US$25 per day for a total of US$750. In addition, two members of the hospital-based research team (authors SRM and LCS) served as monitors for the Facebook group. They were trained by authors LKZ and LAP, a pediatrician and clinical psychologist, to review posts on the Facebook group and to follow the emergency protocol in the case of code of conduct violations (eg, bullying) or signs of self-harm. No such incidents occurred during the study period.

After indicating interest in participating, health care providers were emailed the IRB-approved information sheet serving as informed consent. Providers were emailed the Mindfulness for Resilience in Illness provider guide e-book, instructions for downloading the app, and a link to the online acceptance test. Providers were compensated with a US$100 Amazon gift card. This pilot study was approved by the University of California, Los Angeles IRB and registered on ClinicalTrials.gov (NCT03130751).

Five adolescents and their parents and six young adults (>18 years) were approached and screened. Of the five adolescents and their parents screened, all were eligible and enrolled. Of the six young adults screened, six were eligible and five enrolled. One patient indicated interest but was lost to follow-up. Seven providers were approached and six were screened. One provider declined indicating they were too busy. In total, 21 participants were interviewed. The AYA (n=10) ranged in age from 14 to 23 years (mean 19.3, SD 3.4 years; 50%, 5/10 female). All AYA reported a current or past diagnosis of sarcoma, except for one patient who had another cancer diagnosis but was receiving treatment through the sarcoma clinic. Four of the five parent participants and three of the six providers were female. Participant demographic data are presented in Table 1.

Our exploration of the ways in which a mobile-based mindfulness and social support program can address the psychosocial needs of AYAs with sarcoma yielded three main themes: anxiety about reoccurrence, openness to mindfulness, and desire to connect with other patients.

Examination of the RIM-based [24,25] interview themes guided the development of the Mindfulness for Resilience in Illness program content. Specifically, a persistent theme of anxiety related to uncertainty surrounding reoccurrence and treatment adverse effects confirmed the utilization of a mindfulness-based program. Qualitative data revealed that imaging scans are a source of anxiety, and one participant expressed concern about using the term “body scan” to describe a mindfulness activity. As such, we did not use this specific exercise and reviewed program content for other potential procedure-related terminology. In addition, discussion of self-image concerns informed our decision to end the program with an emphasis on self-compassion.

The strong interest of the AYA in connecting with and learning from other AYA who have experienced cancer guided the content development for both the mindfulness app and the Facebook-based social support group. Given that AYA were interested in—but mostly unfamiliar with—mindfulness, we decided to create videos featuring two sarcoma survivors as program hosts. Facebook group refinement included tailoring activities to promote group interaction and ensuring that group participation would be private.

Building on the interview results, the Learning to BREATHE program [27] informed the structure of the mindfulness app portion of the program. The Learning to BREATHE [27] program was chosen because it is a sequenced mindfulness program appropriate for introducing AYA to mindfulness. Learning to BREATHE has been shown to affect positive outcomes in high school [27,34] and college-aged students [35].

Provider interviews indicated that providers wanted to support their patients’ use of a mindfulness-based program. Providers also indicated that providers focused on addressing psychosocial needs of AYA (eg, psychologists, social workers, and child life specialists) would be most likely to implement the program. Collectively, these findings guided the development of a provider e-book guide.

All program content was reviewed by the study team to ensure that the content would be developmentally and cognitively appropriate for the intended participant sample. Intervention components are described in the aim 2 section; Textbox 1 and Table 3 present the app program content and a description of the Facebook group content, respectively.

Of the 26 patients approached, 22 patients were screened. Of those not screened, two were lost to follow-up, one was not interested, and one was not available during the intervention period. Of the 22 AYAs screened, 22 were eligible and 20 enrolled in the study. One patient indicated interest and was invited to enroll but did not complete online consent forms. The other patient declined because of disinterest. Six providers were screened, found to be eligible, and enrolled. In total, 26 people participated in the pilot study. Twenty AYA participants consented to be in the study. Seventeen of these participants (mean age 19.1, SD 3.7 years; 41%, 7/17 female) downloaded the app and joined the Facebook group (seven of these participants participated in the aim 1 interviews). Of the remaining three participants, one formally withdrew, and two did not participate in the study after completing the online consent and prestudy questionnaires. Data presented throughout this report exclude these three individuals who withdrew from the study. Six additional health care providers consented to participate in the pilot study stage; all six were female. Based on information gathered in the earlier interviews about who would be most likely to offer a psychosocial intervention to patients, we purposefully recruited three child life specialists: two nurses and one psychologist. See Table 4 for details about participant characteristics.

On average, participants completed a mean 16.9 (SD 11.9, range 0-28) of the 28 unique sessions, used the app for a mean 10.2 (SD 8.2, range 1-23) days during the evaluation period, and engaged with the app for a mean 112.5 (SD 79.4, range 1-208) minutes.

All but one person (16/17) in the Facebook group posted at least one reply to the moderator’s prompts. The mean number of responses was 15.2 of 31 (49.0%, range 0%-96.8%); therefore, participants responded to approximately half the moderator’s posts. The daily response rate ranged from 23.5% to 75.0% (mean 49.9%); which means on any given day approximately half the participants responded to the moderator’s post. See Table 3 for a sample of the daily moderator prompts, example responses, and the mean response rates for each category of post.

Overall, participants responded that they enjoyed using the app (mean 5.69, SD 1.12; 1=not at all to 7=very much so), would be likely to continue using the app (mean 5.13, SD 1.54; 1=definitely no to 7=definitely yes), and would be very likely to recommend it to others (mean 6.19, SD 1.05; 1=definitely no to 7=definitely yes). See Table 5 for the ratings of specific features.

Open-ended responses about the experience of AYA with the mindfulness app revealed that some participants found the audio meditations to be relaxing, and in some cases detailed how they were helpful. As one male explained, “I loved that it helped remind me to be mindful of my actions that contribute to battling my anxiety” (patient 205). Others explained that the audios helped with sleep; one female said, “I sometimes have [a difficult time] time sleeping or getting a full night rest but after listening to the voices I slept through the whole night” (patient 206). Participants also offered a range of suggestions for improving the mindfulness app, which centered around improving navigation, more opportunities for active participation, and more references to the cancer experience. Recommendations are listed in Textbox 2.

Participants responded that they found the Facebook group somewhat helpful (mean 4.56, SD=1.75; 1=not at all helpful to 7=very helpful). All 16 participants who completed the acceptance test replied “yes” when asked, “Would you recommend that this type of online social group be part of a future support program for youth with cancer?”

Participants described their experience with the Facebook group in open-ended responses. Participants felt strongly that online social support was a crucial and absent component of their cancer experience. As one young female described, “A social group for young people affected by cancer is something major that is missing for the emotional side of treatment” (patient 215). Some participants stated appreciation for the group; for example, “I liked that it was a safe place to expose your own thoughts/anxieties” (patient 205). In addition, they appreciated that the questions posed by the moderator were not only direct questions about the app program; for example, “I liked that there were not just questions referring to the meditations” (patient 213). Some responded that being expected to respond daily was too much; for example, “That there were questions every day, it was a little hard to keep up with responding” (patient 209).

Participants also offered several useful suggestions for future iterations of the online social support group. To enrich the experience, a few participants suggested facilitating more meaningful relationship-building conversations. For example, a young female said:

Another female suggested that content could be more cancer-specific:

Suggestions for future enhancements are summarized in Textbox 3.

Results of paired samples t tests indicated there were no significant differences from pretest to post test on measures of perceived social support, mindfulness, body image, or quality of life. Means are presented in Table 6.

Providers were enthusiastic about the program. Asked if the mindfulness app would be very helpful for AYA with cancer, four of six responded 7, and two of six responded 6 (where 1=not at all helpful to 7=very helpful). Five of six responded 7, and one of six responded 6 when asked if they would be very likely to recommend it to a family with an adolescent or young adult managing cancer. Please see Table 5 for other quantitative responses.

Providers’ open-ended responses about the mindfulness app and the Facebook social support group also revealed enthusiasm for the approach. One provider reported that the peer-to-peer resources were important (eg, “Peer videos were great! I think you get the most buy in from teenagers when they can hear directly from others who have similar experiences”). Others described the importance of helping AYA to identify and regulate emotions, for example, “I also specifically liked the information about the temporary state of emotions” (provider 02) and “Love the ride the waves quote. This is so true. We cannot take away our emotions, but we can learn to deal with them” (provider 03).

Providers also reported that the provider guide would be a helpful resource (four of six chose 7, one of six chose 6, one of six chose 5; 1=not at all helpful to 7=very helpful), and that the program overall was something that they would be very likely to recommend to another health care provider (five of six chose 7, one of six chose 5; 1=not at all likely to 7=very likely).

To aid with future efforts, providers were asked what information they would find most useful for introducing the mindfulness app to patients. Most (four of six) responded that they would like to receive the provider guide paired with video training on relaxation and mindfulness basic practice. Three providers responded that they would like to receive the provider guide paired with a live webinar run by an expert in mindfulness; two responded that they would like to receive the provider guide paired with a list of practitioners who could use it to help support patients. Suggestions for improvement are summarized in Textbox 4.