A qualitative study was used. Study results are presented according to the COREQ (Consolidated Criteria for Reporting Qualitative Research) checklist [12] (see Checklist 1).

Ethics approval for this study was granted by the Auckland Health Research Ethics Committee (AH21899) on February 23, 2021. All participant data were deidentified after the interviews and focus groups.

This study aimed to recruit 10 participants from each of the following groups: adolescents with T1D, adolescents with T2D, parents, and health care professionals. A total of 15 adolescents with T1D and 4 adolescents with T2D aged 12-16 years were recruited from 2 specialist pediatric diabetes clinics in Auckland, New Zealand, between April and August 2021. In addition, 3 adolescents with T1D were recruited via a Facebook advertisement. A total of 12 parents of adolescents with diabetes were recruited in clinic and 3 were recruited via a Facebook advertisement. A total of 12 health care professionals from 2 Auckland-based specialist diabetes teams were recruited via emails sent out by the first author.

Inclusion criteria for adolescents were being aged 12-16 and having been diagnosed with T1D or T2D more than 6 months ago. Exclusion criteria for adolescents included not being an English speaker or having a serious developmental or psychiatric disorder (eg, psychosis). Parents were eligible for inclusion if they were a parent of an adolescent diagnosed with T1D or T2D more than 6 months ago. Health care professionals were included if they provided care for adolescents with diabetes. During recruitment, 41 adolescents and their parents were approached to take part in the study. Of those, 38 agreed to take part in the study (24 adolescents and 15 parents) and 3 (parents) declined. Of the 20 health care professionals invited to participate via email, 11 agreed to take part in the study and 9 declined due to the lack of interest or time. Overall, 11 of the original 38 adolescents and parents were lost to follow-up, and 4 dropped out after signing informed consent. A total of 34 individuals participated in the focus groups (13 adolescents, 10 parents, and 11 health care professionals). Informed consent or assent (for adolescents aged <16 y) was obtained at recruitment, and participants were asked to fill out a baseline questionnaire regarding basic demographic information, as well as the length of diagnosis and insulin regimen for adolescents. Parents and adolescents completed consent or assent and baseline questionnaires in person at recruitment or on the web if recruited over Facebook. Health care professionals were sent a digital copy of the informed consent and baseline questionnaire.

A total of 9 focus groups, each with 2-4 participants and lasting 30-90 minutes, were conducted over the web using Zoom (Zoom Video Communications) videoconferencing or in person using a semistructured interview schedule. Interviews over Zoom or over the phone were conducted if participants were not able to take part in the focus groups due to scheduling constraints. Adolescents with diabetes, their parents, and health care professionals participated in separate focus groups. Adolescents and their parents did not participate together. One week before the focus groups or interviews, participants received a link to download the Whitu app with instructions to familiarize themselves with the app before attending the focus groups or interviews. In the focus groups, the facilitators showed screenshots of each of the modules to remind the participants of the module content. Each focus group was facilitated by 2 of the following researchers: KG (female European health psychology student), Anna Boggiss (female European health psychology PhD candidate), DL (male Asian health psychology student), and Kalolaine Finaulahi (female Tongan psychology student). Six individual interviews were conducted by KG. There was no established relationship with the participants prior to study commencement. During the focus groups and interviews, participants were also asked about previous use of digital interventions. Whitu modules were shown to participants, and feedback on the current content, look, and feel of the app and possible diabetes-related improvements to the app was obtained. Upon completion, participants were thanked for their time and provided a gift voucher worth NZ $40 (~US $25). Audiotaped recordings were then transcribed by the first author.

Transcripts were analyzed by 2 members of the research team (KG and HT) in NVivo (Lumivero) using directed content analysis, an approach where data exploration is guided by an existing framework or theory [13]. The qualitative analysis was informed by user engagement frameworks, including the Engagement, Functionality, Aesthetics, and Information domains from the user version of the Mobile Application Rating Scale [14], and used a largely deductive approach. The researchers began by independently coding the data, then forming categories, and lastly identifying descriptive themes. Any coding discrepancies were resolved by discussion with a third member of the research team (AS).

Characteristics of the study participants can be seen in Tables 1-3 below.

The four main descriptive themes identified from parents were (1) support for the use of a well-being app, (2) general acceptability of Whitu, (3) similar desire for more diabetes-related content, and (4) recommendations for parental involvement and support.

Three main themes were identified from health care professionals: (1) limited awareness of well-being–related digital interventions, (2) variable responses to Whitu, and (3) some desire for clinician involvement or control.

Qualitative data from adolescents, parents, and health care professionals revealed important insights into perceptions of Whitu, and overlapping themes were identified between the 3 groups. Overall, there was sufficient acceptability of an app such as Whitu for supporting the well-being of adolescents with T1D. However, some modifications to the current version are required to ensure it is fit for purpose for this audience. Despite efforts to recruit and interview adolescents with T2D, insufficient data were obtained from adolescents with T2D to understand their perspectives. Further work needs to be done to improve the recruitment of this cohort.

Our finding that adolescents in this study had limited prior use of well-being apps is in line with previous research, which found that despite a quarter of adolescents being open to the idea of using mobile health (mHealth) apps, only 7.3% had ever actually used one [15]. It is likely that non–health-related technologies such as games and music are more commonly used to manage mood and stress [16]. However, this does not detract from the appeal of digital technology for digital natives, and therefore, it is encouraging that most participants were supportive of Whitu being adapted to be more useful for adolescents with diabetes.

The acceptability of Whitu by adolescents with T1D and their parents is in line with its acceptability by participants in previous studies of Whitu among young people without T1D [9,10]. It also aligns with previous research findings that clean, symmetrical, and creative apps are associated with better perceptions of app quality [17]. Adolescents and parents also enjoyed the content of Whitu and found the modules and exercises helpful and relevant. These findings echo previous qualitative studies where adolescents expressed their desire for mHealth apps to include resources relevant to their needs [16]. Although general well-being information was appreciated by adolescents, more diabetes-related content was also requested by all of them. This combination of views is consistent with previous findings that adolescents with diabetes prefer face-to-face and digital psychosocial interventions to include tailored, diabetes-specific content [18-20] and that adolescents without diabetes want well-being apps to be relevant and based on their specific needs [16]. It was reassuring to see that Whitu was perceived as culturally acceptable by adolescents in this study. They expressed their enjoyment of the karanga and the fact that the app was congruent with Māori culture and Māori models of mental health and well-being. These results highlight the importance of co-design to build tools that align with the culture and values of the app users.

Adolescents and parents expressed a preference for Whitu to help connect adolescents with peers who also live with diabetes. Several qualitative studies have found similar findings, where adolescents have reported that connecting with their peers was an integral part of psychosocial interventions for youth with diabetes [18,19]. This is supported by other studies in digital health where adolescents with diabetes prefer to have some form of social connection in mHealth apps [21]. Although the current version of Whitu has not been designed to be externally supported by parents or clinicians, it is interesting to note the findings of recent studies that peer and professional support may increase user engagement with health and well-being apps [22]. It may be that this is of greater relevance for adolescents with T1D who are more used to receiving health support than those without diabetes.

Health care professionals displayed limited familiarity with digital mental health interventions and a greater focus on the identification and management of mental health problems than the preservation or improvement of patient well-being. Given that technology including diabetes pumps and glucose-tracking apps are commonly used these days [23], this may be more a reflection of their limited familiarity with contemporary well-being interventions [24] and anxiety regarding the management of mental health problems during busy medical clinics, rather than their status as “digital immigrants” or the lack of interest in holistic patient care. Given the importance of health care professionals in the implementation of digital interventions and that digital mental health tools in diabetes should support and augment existing psychosocial support, it is essential that further iterations of Whitu are closely developed with this group to ensure their acceptability and feasibility within clinical settings. Formal implementation research may also need to be undertaken to maximize the impact of a diabetes-specific version of Whitu.

The main strength of this study was the inclusion of multiple perspectives from adolescents living with T1D, parents, and health care professionals, enabling us to gain a rich understanding of how Whitu might be adapted to meet the needs of this patient population.

In contrast, a limitation of this study was that we were unable to recruit any endocrinologists to explore their views of the Whitu app. This is somewhat mitigated by the fact that the research team includes 2 pediatric endocrinologists; however, endocrinologists represent an important stakeholder group, and their input and support of any future apps is integral to the success and implementation of well-being apps as part of diabetes care. Another limitation was our difficulty in recruiting adolescents with T2D (n=2) and their parents (n=1), which is also an issue faced by previous researchers [25]. As a result, current findings cannot be generalized to those with T2D. Given the high rate of T2D among Māori and Pacific adolescents and the lack of interventions developed for this group [8], this is an urgent gap to address in the literature. Future research in New Zealand should aim to use Kaupapa Māori research approaches (Māori research paradigm) that are led or coled by Māori researchers, which have shown to be highly effective for building relationships, engaging and retaining participants, and prioritizing Māori worldviews [26,27]. Other strategies to increase participant recruitment include using more culturally appropriate recruitment strategies, for example, recruiting through schools with a high percentage of Māori and Pacific students or via Māori and Pacific community groups. Cultural and language support for focus groups and greater use of face-to-face interviews may also yield more useful results. Lastly, future research should also offer adolescents with diabetes the opportunity to have individual interviews, as they may prefer to discuss potentially sensitive topics privately, without their peers present.

Based on the findings from this qualitative study, we have recently adapted Whitu to be a diabetes-specific well-being app called LIFT: Thriving with Diabetes. We are currently conducting a feasibility study of LIFT among youth with T1D and their parents. The adapted version contains the original 7 modules; however, they have all been tailored to be specific to diabetes. We have also added additional content to allow parents to better support adolescents and to improve parental well-being. Importantly, we are also planning a subsequent qualitative study to explore whether the newly adapted LIFT well-being app is engaging and useful for youth with T2D. This would allow us to explore whether LIFT needs to be further tailored for T2D.

With diabetes-specific modifications, LIFT (the newly adapted version of Whitu) may be an acceptable digital intervention for improving the well-being of adolescents with T1D. Continued collaborative development of the app with adolescents with T2D, parents, and clinicians will be important to ensure its utility, uptake, and use in clinical care.