Families were eligible to participate if (1) the adolescent’s primary caregiver self-identified as Hispanic; (2) the adolescent was between the age of 12 and 15 years; (3) the adolescent lived with an adult primary caregiver (ie, biological parent or legal guardian; herein referred to as “parent”) willing to participate; (4) both parent and adolescent owned a smartphone with internet access; (5) the adolescent did not meet recommendations for fruit and vegetable intake as determined by the fruit and vegetable brief measure [33]; (6) the adolescent did not meet physical activity guidelines as determined by the 60-minute MVPA screening measure [34]; and (7) the adolescent self-reported engaging in more than 2 hours of daily screen time, based on three items from the Youth Risk Behavior Surveillance Survey [35].

Families were excluded if (1) the adolescent’s BMI-for-age-and-gender was <5% (underweight) or ≥ 95% (obesity); (2) the adolescent had a chronic medical condition (eg, type 2 diabetes) that required intensive lifestyle modification; (3) the adolescent had a diagnosed developmental delay that would interfere with understanding program materials; (4) the parent or adolescent had a diagnosed medical or psychiatric condition and was taking medications that would interfere with changes to physical activity or diet (eg, adolescent was diagnosed with Attention Deficit Hyperactivity Disorder and was on stimulant medication); and (5) the family was planning to move out of South Florida during the study period.

We used several recruitment strategies, including partnering with local pediatric primary care clinics in South Florida, drawing from consent to contact lists, offering monetary incentives for referrals from enrolled study participants, advertising the study on social media, and distributing flyers at community events.

Our research team contacted a total of 540 potential families. Of those, 45% were ineligible, 31% did not respond to our calls, 12% refused to participate, and 2% were enrolled in a competing study (see Multimedia Appendix 1 for participant flow diagram). Of the 55 eligible families, 82% (n=45 families) provided written informed consent and assent and agreed to participate in the study.

We prescreened potential participants in person at pediatric primary care clinics or by phone. To confirm the adolescent’s eligibility, study staff measured their height and weight and calculated their BMI percentile. Due to the COVID-19 pandemic and the university policy to conduct research remotely at the time, we conducted recruitment for our field trial (IDEAS phase 7) exclusively by phone and determined adolescent eligibility using parent-reported height and weight to estimate adolescents’ BMI percentile.

We audio- or video-recorded all qualitative data collection sessions, and they were transcribed verbatim by a professional transcription agency (National Captioning Institute). Transcription was performed singly, and a member of our research team conducted quality checks of the transcripts against the original audio recordings to ensure completeness and accuracy. If participants did not consent to being recorded, extensive notes were taken instead.

Different sets of Hispanic parent-adolescent dyads participated across 3 phases of data collection (conceptual prototypes, n=20 dyads; paper and minimally functional prototypes, n=15 dyads; and fully functional prototype, n=10 dyads). Procedures for each of these phases are detailed below.

Parents and their adolescents (n=20 dyads) participated in separate focus groups or individual interviews. We initially set out to conduct focus groups; however, due to scheduling challenges, we used a combination of group-based and individual interview approaches. The purpose of these initial focus groups or interviews was to understand the needs and behaviors of our target users. Prior to qualitative data collection, each participant completed a brief demographic survey. Parents self-reported their age, sex, country of birth, years living in the United States, marital status, education, employment, household income, and potential history of diabetes. Adolescents self-reported their age, sex, country of birth, and years of living in the United States. Parents and adolescents also completed the Mobile Device Proficiency Questionnaire (MDPQ-16) [36], a 16-item validated instrument that assesses proficiency related to mobile device use. Sample items include “I can find health information online,” “I can listen to music,” and “I can send emails” (all using a mobile device). Participants rated their level of proficiency on a 5-point response scale ranging from never tried at all to very easily. The measure yields 8 subscales (eg, communication, entertainment) and one total score, with higher scores reflecting greater mobile device proficiency.

We used a semistructured guide (Multimedia Appendix 2) with open-ended questions designed to understand participants’ current lifestyle behaviors (eg, “How [if at all] are you healthy together with your parent/adolescent?”) and use of technology for health (eg, “How [if at all] have you used technology for your health/health care?”). We additionally gathered participants’ feedback on initial intervention concepts and prototypes (“conceptual prototypes”). Bilingual study staff conducted focus groups and interviews in participants’ homes or at the University of Miami in either English (one adolescent focus group [n=3], n=8 individual adolescent interviews) or Spanish (one adolescent focus group [n=2], n=7 individual adolescent interviews; all parent focus groups and individual interviews). Focus group sessions were ~55 minutes, and individual interviews were ~36 minutes. Participating parents and their adolescents received US $50 and 2 movie tickets, respectively.

Based on feedback from our initial focus groups and interviews, our team developed a series of paper and minimally functional prototypes, or those that meet some, but not all, specifications for a product. Our paper prototypes consisted of physical printouts of illustrations depicting intervention features. Our minimally functional prototypes consisted of a webpage accessed via smartphone with learning activities that aligned with didactic content.

A different set of parents and adolescents (n=15 dyads) completed the aforementioned demographic survey and participated in separate individual interviews. We used a semistructured guide (Multimedia Appendix 3) with open-ended questions designed to solicit participants’ feedback on potential logos (eg, “What images come to mind when you hear ‘Healthy Juntos’ [Healthy Together]”) and to encourage participants to “think aloud” while interacting with the paper and minimally functioning prototypes to understand their navigation of the intervention, identify potential problems, and assess their interpretation of the prototype’s design elements. We also collected feedback on participants’ favorite and least favorite components, suggested improvements for the functionality of game prototypes, and impressions of the overall design of the program. Bilingual study staff conducted interviews in participants’ homes or a public setting (eg, library) in either English (n=13 adolescents, n=3 parents) or Spanish (n=2 adolescents; n=12 parents). The parent and adolescent interviews lasted ~42 minutes and ~30 minutes, respectively.

After the interviews, participants completed the Usefulness, Satisfaction, and Ease of use (USE) questionnaire [37]. This 30-item measure assesses four dimensions of usability: usefulness (8 items), ease of use (11 items), ease of learning (4 items), and satisfaction (7 items). Response options ranged from 1 (strongly disagree) to 7 (strongly agree). The 4 mean scores were averaged to calculate a total usability score, with higher scores indicating greater usability and satisfaction. Participating parents and their adolescents received US $60 and 3 movie tickets, respectively.

Based on feedback from the paper and minimally functional prototypes, our team developed a fully functional prototype of a web-based application, available in English and Spanish, for an 8-week family-based intervention. Although ≥26 contact hours over 2‐12 months are recommended for pediatric obesity treatment [38], no equivalent recommendation exists for preventive interventions, with prior meta-analyses (including our own) finding no clear relationship between intended dose and weight outcomes [13,39]. The initial intervention duration was informed by the principal investigator’s (PI) prior work developing and evaluating an 8-week family-based (face-to-face) intervention that demonstrated postintervention improvements in the lifestyle behaviors of non-Hispanic Black adolescents and parents [40,41].

We evaluated our fully functional prototype in a field trial (n=10 dyads). Field trials involve the collection of data in users’ natural environments and evaluate technology for flaws that will only emerge through use over time [42]. We gave parents and adolescents a physical activity monitor (Fitbit) to track their steps and held a one-hour videoconference call to set up their usernames, sync their Fitbit devices to the app, and introduce them to our intervention. At the end of each week, participants rated the week’s content and provided feedback on what they liked, disliked, and would change about each session, as well as any technical issues they encountered. Participating families also received a follow-up phone call after Week 4 to check on their progress. At the end of the eight weeks, study staff invited families to participate in exit interviews via videoconference using a semistructured guide with open-ended questions (Multimedia Appendix 4). The interview guide was designed to capture participants’ experience using the program (eg, “How [if at all] did the program help you meet your goals?” and “How [if at all] did the program impact your relationship with your adolescent/parent?”). Bilingual study staff conducted interviews in either English (n=7 adolescents, n=1 parent) or Spanish (n=1 adolescent; n=7 parents). Interviews lasted an hour on average. Parents received a US $5 digital gift card for every weekly survey and a US $40 digital gift card for participating in the exit interview. Adolescents received a US $30 digital gift card for participating in the exit interview. Parents and adolescents were allowed to keep their Fitbit devices once the study was completed.

We summarized quantitative data using descriptive statistics and did not perform any inferential statistical tests. Analyses were conducted in SAS (version 9.4; SAS Institute Inc).

We used a rapid qualitative analysis to analyze qualitative data [43-45]. Rapid qualitative analysis leverages a team-based approach to summarize key points from qualitative data into a matrix, allowing for efficient and systematic exploration of relevant themes. This method can facilitate the development of technology-based interventions [46]. At each phase of qualitative data collection, we first reviewed transcripts for accuracy against the original audio or video recordings. We then developed a summary template to condense data from each transcript across a series of domains generated from the questions in the interview guides. Study staff created matrices that displayed all participant responses corresponding to each domain. In the matrices, each focus group or interview was represented by a unique row, and each domain was represented by a unique column. We distributed our completed matrices to study staff for revision and used a team-based discussion approach to draw key conclusions that informed our intervention development.

Twenty adolescents (mean 13.5, SD 1 year old, 13/20, 65% female, BMI percentile: mean 59.1, SD 26.1) and their parents (mean 43.2, SD 6.2 years old, 18/20, 90% female, BMI: mean 30.1, SD 12.2) participated in this phase. Participants’ responses on the MDPQ-16 showed high mobile device proficiency for both parents and adolescents across all eight dimensions (Multimedia Appendix 5).

In focus groups and individual interviews, parents and adolescents defined a healthy lifestyle as one that includes exercise, healthy eating (eg, fruits and vegetables, limited junk food and sugar), adequate hydration, and sufficient sleep. They rated their health as “average” or “in between,” acknowledging both positive habits and their need for improvement. They described limited use of technology for health purposes, primarily for internet searches related to medical symptoms, recipes, and workouts. Most reported not having any health-related apps installed on their smartphones. When asked about features they liked about either health apps or their favorite apps in general, parents and adolescents emphasized the importance of interactive, social features, such as the ability to communicate with others, and regularly updated content or information. Participants, particularly adolescents, expressed a desire for minimal reading requirements.

Participants viewed conceptual prototypes in their preferred language (English or Spanish) developed by the PI using PowerPoint (Microsoft Corp) to illustrate 3 possible components or features of a family-based digital lifestyle intervention based on the conceptual framework: (1) didactic information on healthy lifestyle behaviors, (2) a family behavior change tool, and (3) a positive parenting tool (Figure 2).

In response to these prototypes, participants suggested that the intervention provide culturally relevant healthy meal recipes, information about nutrition labels, and appropriate portion sizes. They also recommended that we describe the benefits of exercise for adolescents and adults and offer workout routine ideas. Regarding behavior change, participants suggested that we include individualized goal setting with visuals to track their progress, a way for family members to share goals and progress, and reminders and motivational messages. Finally, regarding parenting, parents suggested that we provide strategies to improve family communication, tips on how to deal with difficult situations, and ways to motivate and reward adolescents. They also recommended that we include affordable options for family activities.

We developed a series of 10 paper prototypes using PowerPoint, and our technology partners (led by author RS) developed minimally functional prototypes to preliminarily assess product usability (Figure 3).

The paper prototypes included representations of different icons and features of the intervention (ie, a sample log-in page, a home page with key icons, and separate pages that corresponded to each icon). Illustrations showed educational tools (ie, didactic health videos, serious games, culturally tailored recipes, and exercise tutorials), behavior skills training tools (ie, step counting, monitoring diet, and physical activity), and parenting and family motivational tools (ie, a parenting podcast, family goal comparison, and family activities). We envisioned that participants would receive weekly didactic content, set and monitor physical activity and healthy eating goals, and engage in weekly family discussions and health challenges. Parents would also have access to a parenting “podcast” to learn skills for positive communication and parenting (eg, active listening). Both parents and adolescents could compare progress, exchange preselected encouraging messages, maintain a food photo journal, participate in weekly games (eg, create your own healthy plate), and learn about the cultures and traditions of different Latin American countries each week. The minimally functional prototypes were of learning activities that aligned with didactic content. These activities included tasks like dragging and dropping healthier food options onto a plate, reading and answering questions about a nutrition label, and timing themselves while doing various physical activities.

Fifteen adolescents (9/15, 60% female; mean 14.2, SD 1.3 years) and their parents (14/15, 93% female; mean 46.3, SD 4.7 years) participated in this phase. Participants suggested improvements related to aesthetics (eg, icon design, image quality) comprehension (eg, simple directions, better feature descriptions), content (eg, removal of certain features), and functionality (eg, different buttons; Table 3). Overall, participants provided mostly positive feedback on the prototypes, highlighting ease of use, functionality, the ability to compare parent and adolescent progress, and enthusiasm for recipes and exercise components. Adolescents were content with the organization of the pages and the color scheme. The intervention prototypes received positive ratings from both parents and adolescents, with USE questionnaire average scores higher than 5 (out of 7) across all four usability dimensions (Multimedia Appendix 5).

We made significant changes to our prototypes and worked with our software development partners to build out a fully functional prototype (Figure 4) that included eight modules of content and corresponding activities.

We made changes to aesthetics (eg, updated icons and logo), comprehension (eg, enhanced clarity through clearly labeled icons and an instructional welcome video), and content (eg, workout videos to emphasize proper form and incorporated healthy alternatives to typical Hispanic foods). To address issues with the functionality of games, we added clear instructions to each page, replaced the “drag and drop” feature with a simpler “click to select,” and incorporated congratulatory messages upon game completion. We simplified food tracking by requiring participants to simply enter consumed foods, reducing the burden on users. In addition, our team created animated video clips to deliver didactic content, identified publicly available video clips of culturally relevant recipes, produced short exercise suggestion and safety videos, recorded audio clips to deliver positive parenting content, and worked on programming behavior change features (eg, goal setting, self-monitoring, and family support).

As illustrated in Figure 4, the home page gave participants access to content and features through five icons: (1) Semana Healthy (healthy week), (2) My Progresso (my progress), (3) My Comidas (my foods), (4) Bien Fit Workouts (very fit workouts), and (5) Muy Tasty Recetas (very tasty recipes). Although all content was available in English and Spanish (based on participants’ preferred language), the icons on the home page were labeled in Spanglish (ie, combination of English and Spanish) to appeal to adolescents’ biculturalism [47]. The Semana Healthy icon contained all weekly modules (for a total of 8 modules; see Table 4). Within Semana Healthy, participants could complete self-guided weekly tasks that included setting weekly goals for daily steps and healthy meals, watching two to three 2-to-5-minute animated didactic videos, playing a game that reinforced weekly didactic content, and learning about physical activities (eg, salsa dancing) and foods from a featured Spanish-speaking country. Parents additionally had access to one weekly 10‐15 minute “Parenting Radio” episode hosted by the PI that discussed family communication and positive parenting skills. Except for the “Parenting Radio” feature, all content was the same for parents and adolescents. Finally, within Semana Healthy, parent-adolescent dyads were prompted to increase healthy interactions and communication through health and discussion activities. For family health activities, participants were prompted to select at least one of three healthy activities (eg, go for a walk together). For family discussion activities, participants discuss 2‐5 open-ended questions. Participants could document their completion of these activities through typing a summary in a text box. The app also included self-monitoring features (“My Progreso,” “My Comidas”) that allowed participants to view their own and their dyad member’s average daily steps (read in from Fitbit) and type in the foods they ate. They could also access a library of 40 workout videos with cues for form and safety created by our team (“Bien Fit Workouts”) as well as links to 46 culturally relevant video recipes from Kiwilimón.com (“Muy Tasty Recetas”).

Ten adolescents (mean 13.9, SD 1.3 years old; 5/10, 50% female, BMI percentile: mean 65.3, SD 25.5) and their parents (mean 43.7, SD 5.6 years old; 10/10, 100% female, BMI: mean 25.8, SD 2.5) agreed to participate in this phase. Eight of ten families enrolled in our field trial completed all weekly sessions and postintervention interviews. Results from the USE questionnaire showed scores above 6 (out of 7) for both parents and adolescents across all 4 usability dimensions, indicating high satisfaction with Healthy Juntos and perceiving the intervention as useful and easy to learn and use (Multimedia Appendix 5).

In exit interviews, participants reported a positive experience, with many indicating improved communication and a sense of “feeling closer.” In addition, they reported increased physical activity levels and healthier dietary choices, such as consuming more vegetables and reducing unhealthy food intake. In terms of intervention usage, most parents reported completing the required weekly activities in a single session, while adolescents often fell behind their parents’ pace. The most positively received intervention components were the didactic videos, parenting radio episodes, and the healthy games feature. However, the messaging feature was underused, with participants preferring verbal motivation. In addition, many participants found the progress tracking feature confusing and commented that logging their meals was tedious, as was keying in summaries of their family health and discussion activities. Most did not use the recipes and suggested including easier, more diverse, and “fun” recipes.

Notably, all but one parent expressed interest in having weekly or biweekly human contact to help them overcome barriers and provide feedback. Our experience managing the field trial also indicated the need for more systematic human support. To encourage families to complete sessions, we supplemented automated text reminders with phone calls to parents. Parents used these calls to solicit technical and problem-solving support.

Based on participants’ feedback, we made a final round of modifications to enhance aesthetics, comprehension, content, and functionality in preparation for a pilot randomized controlled trial (Table 5 and Figure 5). We also developed a health coaching protocol for brief weekly video conference consultations with parents.