Despite global declines in HIV incidence and mortality, this progress remains uneven across many geographic regions and key affected populations, such as men who have sex with men (MSM), especially in Malaysia, where HIV incidence among MSM continues to increase [1-3]. Recent estimates indicate that HIV prevalence among MSM has reached an all-time high of 21.6% [4], with especially alarming trends in Kuala Lumpur, where HIV prevalence remains high and appears to be increasing [3]. Young men who have sex with men (YMSM; younger than 29 years old) in Malaysia, in particular, account for 44% of new HIV infections in Malaysia [5], driven by high rates of condomless sex, low uptake of HIV testing, and limited access to pre-exposure prophylaxis (PrEP) [6-9]. In addition to heightened risk for HIV infection, Malaysian YMSM also experience high levels of comorbid medical and psychiatric conditions, including high rates of tobacco use [10], alcohol consumption, and methamphetamine use [6,11,12], mental disorders [13,14], and sexually transmitted infections (STIs) [15-18]. Furthermore, their misperceptions of invulnerability and related impulsive acts, and lack of accurate knowledge on the association between these comorbidities and their HIV risk, are associated with lower perceived well-being [19,20]. These overlapping comorbid conditions not only contribute to elevated risk-taking [21,22], but also undermine engagement in HIV prevention services, such as HIV testing and PrEP. A more holistic approach that addresses these intersecting comorbid conditions may significantly reduce HIV risk and improve overall health outcomes among Malaysian YMSM.

In Malaysia, where same-sex sexual activity is criminalized under both secular and Shariah law, YMSM face significant legal and social barriers to accessing HIV prevention and related care [23]. These punitive legal frameworks contribute to pervasive stigma and discrimination, often reinforced in health care and clinical settings, which can deter YMSM from seeking in-person services [2]. Mobile apps offer a promising digital solution by providing safe and user-controlled access to information, screening tools, and linkage to services. By bypassing face-to-face interactions with potentially stigmatizing providers, mobile apps can reduce opportunities for enacted stigma, thus creating safer and more affirming pathways to care. In this context, digital platforms may not only improve engagement in HIV prevention and mental health services but also help restore autonomy, confidentiality, and dignity among YMSM who would otherwise avoid traditional care settings due to fear of exposure, judgment, or legal repercussions [24-26]. As such, mobile apps represent a scalable and context-sensitive approach to reducing stigma-related barriers and supporting the health and well-being of this highly marginalized population. Moreover, these apps have the potential to serve as a “one-stop-shop” model to address these complex and overlapping conditions faced by Malaysian YMSM. Fortunately, smartphone ownership is nearly universal in this population (97%) [27,28] with the majority (63%) frequently using apps for health information seeking [28], highlighting the feasibility and acceptability of mHealth apps for HIV prevention and care in this vulnerable population.

Many existing HIV prevention apps, however, endure low user interest and poor retention, especially among YMSM. Studies of mobile health interventions indicate that less than 20%‐40% of users remain active after a few months, and many lose engagement shortly after initial use (eg, only approximately 19% completed activities in the “Stick To It” HIV/STI screening app) [29]. HIV prevention fatigue, resulting from repetitive exposure to traditional HIV prevention interventions, such as persistent recommendations for condom use, can also undermine gains in this group due to boredom and diminishing interest in existing HIV prevention methods, which can lead to increased condomless sex [30-32]. Gamification, the integration of game design elements such as rewards, social incentives (eg, competition and collaboration), goal-setting, storytelling, and elements of chance and surprise into nongame contexts [33-35], can address the issues of low engagement and HIV prevention fatigue. More broadly, gamification strategies have been associated with higher long-term engagement in digital health tools, often boosting motivation and adherence to healthier behaviors [36,37], and better health outcomes [34] in areas such as depression, diabetes, cancer, nutrition, and sexual health [38].

In the context of HIV, gamification has shown promise in maintaining user interest and improving retention in HIV-related mHealth tools [39-43]. A recent meta-analysis of gamified interventions among MSM found significant short-term reductions in condomless anal sex and improved PrEP adherence [44]. Furthermore, qualitative evaluations of gamified platforms reveal that point-based rewards and social connectivity increased acceptability and ongoing use among MSM users [45]. By leveraging these game mechanics, HIV prevention apps can overcome attrition, sustain behavior change, and thereby enhance their public health impact. Gamification can also be particularly useful in addressing comorbidities among YMSM, including substance use, smoking, depression, and stigma, which require sustained efforts in self-regulation and adherence. Transferring these efforts into a fun and engaging digital environment may influence users’ health attitudes, self-efficacy in behavioral change, and intrinsic motivation toward health-promoting behaviors, all of which are associated with greater adherence to health-promoting activities [46,47]. Expanding app functions to track additional health domains (eg, smoking, physical activity, mood, and alcohol and drug use) may also diversify content, reduce HIV prevention fatigue, and position the app as a comprehensive “one-stop” platform that integrates HIV and non-HIV health needs while helping to destigmatize prevention [40].

Despite evidence that gamified apps enhance adherence to both HIV treatment [39-41] and PrEP [42] and promote HIV testing [43] in the US context, gamified HIV prevention apps have not yet been adapted for Malaysian YMSM. This presents a gap given the digital gaming culture among Malaysian YMSM, with 61.5% reporting gaming engagement [48] and 80% primarily using smartphones for digital gaming [49], suggesting high feasibility and scalability of gamified HIV prevention apps. In a context marked by rising HIV incidence, persistent structural and stigma-related barriers to prevention and care, limited culturally tailored programs, and growing fatigue toward traditional HIV prevention methods, culturally adapted gamified strategies offer an innovative and engaging approach to sustain prevention behaviors and reduce HIV incidence in this key affected population.

The alpha version of the My-Hero was adapted from an existing mHealth platform, HealthMPowerment (HMP), which was developed based on the integrated behavioral model (IBM) [50,51]. The IBM extends the theory of planned behavior (TPB) by including 5 additional components known to impact HIV prevention behaviors, including affective motivations, behavioral skills and capacity, perceived and actual environmental constraints, prior habits, and salience of these behaviors. This model recognizes that decision-making is often affectively, rather than analytically, motivated and acknowledges that YMSM’s lives may be impacted by intersectional stigma, psychological distress, and substance use, which hinder adoption and maintenance of HIV prevention behaviors. In the context of sexual health, the IBM has been used to guide mHealth strategies in HIV and STI testing and PrEP uptake. Specifically, educational content, motivational messaging, and app features were designed to incorporate the IBM components, aiming to increase motivational beliefs, perceived norms and perceived behavioral control, attitudes toward testing, and self-efficacy, which then led to improvements in routine HIV and STI testing and PrEP uptake in both clinical trials and observational studies [52,53]. Building on this foundation, the IBM guided the design and development of the key features of HMP, including gamification elements, to ensure its breadth and interactivity by allowing MSM to (1) learn credible HIV prevention content through HMP’s curated content library; (2) customize HMP’s content by discussing relevant topics with clinicians and through conversations with peers; (3) engage in gamified activities that elicit reflection on environments, habits, behaviors, and motivations as a means to identify new risk reduction strategies; and (4) practice skills through interactive activities and games to reinforce self-efficacy and behavioral skills. Accordingly, MY-Hero includes multiple features that address these important domains, including information (education modules), motivation (gamification incentives and normative encouragement), and behavioral skills (through interactive tools and practice challenges). Multimedia Appendix 1A presents the key features and functionalities of HMP. Multimedia Appendix 1B presents their alignment with the IBM. Building on this foundation, this study aims to adapt HMP to the Malaysian context by identifying app features, design principles, and gamification elements that are most relevant and culturally appropriate for Malaysian YMSM. To achieve this, we employed an interdisciplinary expert panel approach to systematically elicit and refine recommendations that would enhance the app’s cultural relevance, feasibility, and potential public health impact.

mHealth research is interdisciplinary by nature. The successful design and implementation of mHealth tools require expertise spanning applied computing, medicine, health care information systems, human-computer interaction, collaborative technologies, and social sciences [54]. Integrating knowledge from these domains is essential to understand how mHealth tools can be effectively used for disease prevention [55]. As technology continues to evolve rapidly, an interdisciplinary approach to mHealth tool development is particularly important to keep pace with technological advancements and address complex health care challenges [56,57]. Collaborating with interdisciplinary expert groups from different fields brings broader perspectives and enhances the quality, relevance, and impact of mHealth research [56,58], ensuring that they are not only effective in improving targeted health outcomes but also user-friendly, culturally appropriate, and responsive to the diverse needs of the target population, particularly in stigmatized cultural contexts such as Malaysia, where stigma negatively impacts health behaviors and overall well-being.

In the context of health, the expert panel approach uses structured discussions that integrate domain-specific expertise to guide the design, refinement, and implementation of health interventions, which is particularly useful for complex and interdisciplinary studies.

Expert panels are especially valuable when empirical evidence is limited or when multiple disciplinary perspectives are needed to inform intervention development and decision-making [59,60]. Expert panels have been used in mHealth studies, including physical activity [61], breastfeeding [62], obesity management [63], anemia and epilepsy prevention [64,65]. Unlike traditional qualitative methods such as focus groups or interviews, an interdisciplinary expert panel convenes experts from multiple fields who possess extensive experience and contextual knowledge of the relevant sociocultural and technological landscape. This approach enables the generation of practice-oriented insights that extend beyond theoretical frameworks, thereby helping bridge the gap between academic knowledge and real-world practice [66]. Expert panels are particularly valuable when a study involves hard-to-reach populations, such as YMSM in stigmatized contexts, where direct engagement may be more sensitive, complex, or limited [67]. Even when the target population is accessible, consulting experts at an early stage helps ensure that the app is appropriately designed before end-user involvement. This is especially important because poorly designed apps may not only fail to meet users’ needs but may also inadvertently cause harm or discourage user engagement among vulnerable populations [68]. Additionally, in the context of mHealth for HIV prevention among MSM, the interdisciplinary expert panel approach has not been widely used. This presents a valuable opportunity to align app design with end-user needs while prioritizing usability and accessibility, thus strengthening the development, relevance, and effectiveness of such interventions.

Given the potential of gamified mHealth tools to address HIV prevention and comorbidity management needs among Malaysian YMSM, this study aimed to use an interdisciplinary expert panel approach to identify key app features, design principles, and gamification elements of the MY-Hero app with the goal of making it culturally appropriate and tailored to HIV prevention and comorbidity management needs of this population.

We collected data using an interdisciplinary expert panel approach that incorporated structured panel discussions to gather and synthesize expert insights to (1) identify essential app features for HIV prevention and (2) explore potential improvements (eg, design principles) to maximize the app’s impact among Malaysian MSM. The expert panel was composed of researchers with extensive experience in health sciences, technology, and design sciences, as well as health care professionals and local LGBTQ+ community leaders with expertise in the sociocultural factors that are unique to the Malaysian context and the Malaysian YMSM population.

Experts were recruited using a convenience sampling strategy to ensure representation across 3 key domains relevant to the development of the MY-Hero app. Convenience sampling is commonly used in studies employing expert panels, often through established professional networks or other readily accessible experts [69-72]. In this study, experts were identified through the research team’s professional networks and referral recommendations. We assembled three expert panels representing (1) health sciences, including infectious diseases, HIV, and STIs; (2) technology and design sciences, including eHealth, mHealth development, and UX design; and (3) Malaysian lesbian, gay, bisexual, transgender, queer, and other (LGBTQ+) community leadership and advocacy.

Experts were selected based on their demonstrated expertise in their respective fields, including strong scientific credentials, a high publication record, or substantial professional experience. For the technology and design sciences panel, inclusion criteria required each expert to have either a doctoral degree and at least 1500 citations or an h-index greater than 10, or more than 10 years of experience designing gamified apps, or more than 10 years of experience designing and implementing mHealth tools for care delivery. For the health sciences and Malaysian context panels, inclusion criteria required at least 10 years of clinical experience in HIV prevention and care or at least 10 years of community service experience in the Malaysian context.

We identified 12 experts (5 leaders from the LGBTQ+ community, 4 from technology and design sciences, and 3 from health sciences) and invited each of them via email to participate in the study. Of these 12 experts, 9 accepted the invitation, with 3 experts participating in each panel. Table 1 presents the characteristics of the experts in each panel. To minimize potential bias, all invited experts were asked to disclose any potential conflicts of interest related to the development or evaluation of digital health technologies or HIV prevention interventions. None reported financial or professional conflicts of interest that would have influenced their contributions to the panel discussions. In addition, the experts were not involved in the direct development of the MY-Hero app and were invited specifically to provide independent perspectives on its development.

Selected experts were initially contacted by email and received an invitation describing the purpose of the study, the goals of the expert panel, and a brief overview of the MY-Hero app. Those who expressed interest and agreed to participate (N=9) subsequently received a follow-up email with several proposed dates and were asked to select a convenient time for participation. Each panel comprised experts from the relevant disciplines, and all panels were conducted between May and August 2023. All the expert panel sessions were conducted online via Zoom (Zoom Video Communications, Inc) to facilitate participation across geographic locations. Each session followed a standardized discussion protocol and a core set of questions, with slight variations tailored to each panel’s area of expertise and experience. Each panel lasted approximately 90 minutes.

At the beginning of each session, participants were introduced to the HMP mobile platform through a live demonstration of its key features and functionalities (see Multimedia Appendix 1). This demonstration was followed by a series of discussion questions developed based on the IBM model. This approach ensured that the discussion addressed the 5 core components of the IBM that affect HIV prevention behaviors (ie, affective motivations, behavioral skills, capacity, perceived and actual environmental constraints, prior habits, and salience of the behavior). Multimedia Appendix 2 presents the questions and how each question aligns with the IBM components. For each expert panel, we tailored the questions accordingly depending on their area of expertise. For example, technology and design science experts received questions emphasizing data privacy and security, as mHealth apps are regulated by the Health Insurance Portability and Accountability Act (HIPAA), and privacy and security, as well as usability considerations relevant to mHealth apps. Health science experts received questions tailored to HIV prevention strategies, implementation science, and HIV-related comorbidities. Finally, local community experts received questions related to sociocultural factors, stigma, and community-specific barriers and facilitators relevant to HIV prevention among YMSM in Malaysia.

This study was approved by the Yale University Human Research Protection Program and Institutional Review Board (protocol number 2000032375). All experts were required to provide informed consent before participating in the study. After the study details, including its purpose, protocol, expectations, and potential risks and benefits were explained, verbal informed consent was obtained. After each panel concluded, the recorded video and automated transcript were downloaded from Zoom. All data were then anonymized, deidentified, and stored in Yale Secure Box, a HIPAA-compliant cloud platform accessible only to authorized staff through Yale Central Authentication Service. The platform is protected by strong firewall controls and monitored by the Yale Information Security Office. Participants did not receive monetary compensation for their participation.

Each expert panel discussion was automatically transcribed verbatim using Zoom. The transcripts were reviewed against the recordings for accuracy and cleaned prior to analysis. Data were analyzed using a hybrid inductive-deductive thematic analysis approach. Two authors (JAW and XZ) independently read all transcripts multiple times to familiarize themselves with the data. Initial open coding was conducted using an inductive open-coding process in NVivo (Lumivero), where codes were generated directly from the data without imposing predefined categories. The 2 analysts then met to compare codes, resolve discrepancies through discussion, and consolidate them into a preliminary coding framework.

Following this inductive phase, the coding framework was examined through the lens of the IBM model. Specifically, codes and emerging themes were reviewed and organized in relation to the IBM constructs that guided the expert panel discussion questions (Multimedia Appendix 2). Themes related to knowledge and awareness of HIV prevention behaviors were mapped to information-related components (eg, salience of behavior), themes related to emotional engagement and motivation toward prevention behaviors were mapped to motivation-related constructs (eg, affective motivation), and themes concerning user capability and practical use of HIV prevention tools were mapped to behavioral skills and capacity and environmental constraints. This process allowed the analysis to remain grounded in participants’ insights while ensuring that the resulting themes were theoretically aligned with the IBM framework.

Codes were subsequently grouped into broader themes through an iterative process of comparison across all 3 panels. A codebook, including theme definitions and exemplary quotations, was developed and refined throughout the analytic process. The final themes represented areas of convergence across panel discussions as well as unique contributions from specific expert groups (ie, health sciences, technology, and design sciences, and LGBTQ+ community experts). Thematic saturation was assessed across panels by examining the extent to which new concepts continued to emerge. After analysis of the third panel, no substantively new themes were identified, and discussions largely reinforced and elaborated on previously identified themes. Therefore, we determined that thematic saturation was reached, and no additional panels were convened. This decision was based on the observed thematic redundancy and the adequacy of thematic depth in addressing the study aims [73].

Thematic analysis yielded 3 key domains. The first domain, implementation science, highlights factors influencing the adoption and sustained engagement with the MY-Hero app among Malaysian YMSM. The second domain, design science, addresses UI, gamification, and technological enhancements to optimize UX. The third domain, the Malaysian context, focuses on culturally specific considerations, such as privacy and security concerns, that are critical to the app’s relevance and acceptability. Together, these domains capture the essential components needed to ensure the effectiveness of the MY-Hero app. Table 2 presents the 3 major domains and subthemes under each domain.

Our expert panels emphasized several strategies that can increase broader adoption of the MY-Hero app among Malaysian YMSM, as well as features that can increase its effectiveness in improving HIV prevention behaviors.

The aim of this study was to use an interdisciplinary expert panel strategy to identify key features in a gamified mHealth app, MY-Hero, for HIV prevention and comorbidity management among Malaysian YMSM. The expert panel identified several themes, including (1) establishing clinical affiliation and linkage to HIV testing, PrEP, and related services, (2) enhancing UI and UX by optimizing usability and interactivity, (3) incorporating customizable health content, tailoring interventions based on individual profiles and preferences; and (4) leveraging certain gamification elements carefully, such as contextualized reward systems and progression tracking, to reach high adoption and sustained engagement. The panel also highlighted (5) the importance of privacy and data security in the Malaysian context. Given that the Malaysian YMSM often experience substantial stigma and discrimination that pose barriers to their access to HIV prevention care and treatment [1-3], along with being disproportionately affected by medical and psychiatric comorbidities, like cigarette use [10], alcohol use, and methamphetamine use [6,11,12], depression [13], and STIs, including syphilis [15-17], an innovative, accessible, and culturally responsive gamified mHealth app, such as the MY-Hero app, offers a promising opportunity by providing a safe, confidential, and engaging platform where YMSM can address their diverse health needs.

The MY-Hero app was developed from the HMP platform with a range of core, established features, including HIV testing information and the ability to order home kits, a medication tracker, and other health-related tools. Following the results, we made several improvements in the MY-Hero app. Multimedia Appendix 3 includes implementation details for each gamification element, including how each gamified feature is operationalized in the app, the mechanism through which it is triggered, the type of reward or feedback provided (eg, virtual points, badges, or content unlocking), and the types of user behavioral data tracked to support these features (eg, module completion, quiz responses, and engagement metrics). Multimedia Appendix 4 presents screenshots of the MY-Hero app illustrating its main features and UI, including how users interact with the app, app navigation structure, gamified features, and other educational features.

These improvements were implemented and prioritized based on four guiding principles, including (1) feasibility with the technical infrastructure and resource constraints, (2) cultural and contextual relevance for Malaysian YMSM, (3) impact on HIV prevention and comorbidity management. Identified subthemes that directly influence user privacy and security, HIV prevention and comorbidity efforts were prioritized for immediate implementation (ie, clinical affiliation and enhanced password protection), and (4) within the scope of the work, which was to design an independent end-user mHealth app to support Malaysian YMSM’s self-management on HIV prevention and comorbidities. Suggestions that could enhance user engagement were integrated while balancing concerns about unintended competitive effects, such as leaderboard modifications and interactivity between users. Design simplification, including reducing unnecessary features and improving navigation, was also prioritized to improve usability.

One of the major themes was user privacy and security. The MY-Hero app was hosted on the HMP server, and the security for the HMP platform operates at both the app and server levels to ensure robust protection of user data. At the app level, users authenticate through the OpenID Connect (OIDC) framework built on OAuth 2.0 (Open Authorization 2.0), with secure login mechanisms including passwords and mobile biometric authentication (FaceID/TouchID), and access is managed through role-based permissions. At the server level, the platform uses a multilayer architecture (proxy, application, and database), with each study hosted on a dedicated server and database instance within Florida State University’s secure data center, where encrypted storage protects data at rest. Data are also encrypted during transmission through enforced HTTPS protocols, and additional protections include restricted database access, routine security patching, and nightly encrypted backups. Following the experts’ recommendations, we further strengthened privacy protections by allowing users to register with pseudonyms, encrypting email addresses, and not requiring real names when ordering HIV testing kits or redeeming swags. To support user interactivity while maintaining anonymity, the platform only displays limited profile information, such as avatars and aggregate in-app activity indicators (eg, number of likes, comments, or posts). No personally identifiable information is visible to other users. These measures were implemented to minimize risks to privacy and confidentiality, which are particularly important given the sensitive legal and social context surrounding same-sex behavior in Malaysia. By strengthening privacy protections, these design choices may also increase users’ trust in the platform, encourage greater self-disclosure and information exchange, and ultimately enhance user engagement and the app’s overall effectiveness [85].

Two recommendations were not implemented, including linkage to care providers outside of the study team (suggested by expert AY) and creation of separate leaderboards for individual in-app activities (suggested by expert AB). These were not feasible because they were outside the scope of the study and constrained by budget and time limitations. Additionally, activity-specific leaderboards, particularly those related to sensitive activities such as sexual behavioral tracking, were not adopted because they could introduce privacy risk by potentially exposing users’ confidential information and jeopardizing their safety, especially in a context where same-sex is criminalized in Malaysia. Table 3 summarizes the enhancements in MY-Hero based on study results, the design domains, and recommendations that were not implemented, along with the rationale for those decisions.

A central theme was the establishment of clinical affiliation and provision of inclusion of harm reduction services to improve access to HIV care and treatment within the app. This aligns with prior research indicating that mHealth apps are more effective when linked to offline care pathways [86,87]. In the context of HIV, mHealth interventions linked to offline services, such as clinic scheduling, telehealth PrEP models, and provider messaging systems, produce greater improvements in HIV testing, retention in care, and medication adherence compared to standalone informational apps [88-90]. Conversely, studies of gamified or informational apps without integrated care pathways report more modest or short-lived behavioral effects, underscoring the importance of structural linkage in sustaining HIV prevention behaviors. By enabling appointment scheduling and direct communication with Malaysian health care providers and the ability to order harm reduction supplies within the app, this platform can help mitigate many barriers faced by Malaysian YMSM, including stigma and discrimination in health care settings [91]. Aligning with multiple components of the IBM model, clinical affiliation can provide many benefits. Specifically, it reduces perceived and actual environmental constraints by minimizing stigma-related barriers in traditional face-to-face settings. It also enhances behavioral skills and capacity by simplifying the process of accessing care and increases the salience of the behavior by embedding HIV-related health actions within a trusted digital platform. Furthermore, by creating positive experiences in health care engagement in a gamified app environment, it may help disrupt prior negative habits and build affective motivation toward sustained HIV prevention behaviors. Despite being unable to establish clinical affiliations outside of the study team due to the project’s scope, time, and budget constraints, we enhanced the app by incorporating the appointment feature that allows users to connect with clinical providers within our study team for counseling, appointment scheduling, and direct messaging for health concerns. In future work, we plan to explore how integrating direct communication with care providers in Kuala Lumpur may further support and improve health behaviors among Malaysian YMSM.

Another finding was the need to collaborate with local nongovernmental organizations (NGOs). This process will require not only software developers but also partnerships with health care providers for service integration and possibly local stores for reward sponsorship. Prior research among MSM populations indicates that trust in HIV-related mHealth interventions depends not only on app functionality but also on who develops, endorses, and governs the platform. For example, studies among Malaysian MSM have shown that users value transparency regarding app ownership and management and perceive endorsement by reputable NGOs or health authorities as enhancing credibility and quality assurance [92]. At the same time, these studies also highlight concerns about government access to personal data and potential surveillance in criminalized settings, suggesting that institutional involvement must be paired with robust privacy protections and transparent governance structures to maintain user trust. In the Malaysian context, one solution could involve early engagement with the Ministry of Health in Malaysia and local NGOs to establish clinic linkages and content accuracy. The Malaysian Ministry of Health endorsement or involvement could lend credibility and help in scaling the app to reach the intended audience. By positioning the MY-Hero app as a credible and trusted resource, especially if managed by a reputable NGO or research institution in partnership with the Ministry of Health, it can strengthen Malaysian YMSM’s affective motivations and trust, which are critical in a context where fear of surveillance or discrimination is prevalent.

Our findings also emphasized the importance of UI and UX enhancements in driving engagement and sustained use. Experts from design and health sciences emphasized that an intuitive interface is essential for user satisfaction, highlighting the importance of simplicity, streamlined navigation, and enhanced customization. Customizable avatars, interactive feedback (haptic or animated responses), and progression systems were also suggested. Gamification elements were also recognized as a pivotal feature of the app. Experts advised on the careful design of competition and reward systems to reinforce positive health behaviors rather than incentivizing risk behaviors. This reflects broader findings in gamification literature, where properly designed game mechanics can enhance motivation and behavior change, particularly in public health interventions [93-102]. This also provides a theoretical insight, in that interventions must temper generic behavioral techniques with context-specific psychosocial factors. Research on gamified health interventions indicates that competitive elements can produce adverse effects on users’ motivation and engagement when poorly matched to the context or user needs [103,104]. For example, studies have documented potential unintended side effects (ie, reduced health motivation) of gamification in health behavior change systems [105]. Likewise, research on competitive gamified learning found that simple competition was often disengaging for or rejected by many users [104]. Finally, analyses of leaderboard effects in health tracking show that competitive mechanics may benefit some user subgroups while harming others, underscoring the importance of tailoring such strategies to the specific audience and setting [103]. Particularly in stigmatized contexts, the incorporation of competition needs to be tailored and designed carefully.

Additionally, the concept of surprise elements, such as unlocking new features or rewards, can maintain long-term engagement, a strategy that has been successfully implemented in commercial gaming apps [106]. The inclusion of unexpected rewards positively influences user motivation and engagement by disrupting routine interaction patterns and generating new interest [107]. Such disruption can stimulate anticipation cycles, thereby increasing engagement intensity and encouraging continued participation over time [108]. Moreover, instructional design literature highlights that controlled unpredictability activates curiosity, one of the core intrinsic motivational drivers, prompting users to remain engaged in order to discover what comes next [109]. Our experts highlighted that such surprise can be particularly important for YMSM, who respond positively to new and surprising content. In HMP, the rewards were only virtual badges. Building on our experts’ suggestion, we expanded this gamification element by providing tangible, redeemable rewards such as T-shirts, water bottles, power banks, and a surprise party pack that includes harm reduction items, such as lube, condoms, fun sexual activities, or games. All these aforementioned suggestions on UI and UX align with the behavioral skills construct in the IBM model in several ways. First, interactive and personalized features, improved navigability, and more customization and simplicity can enhance affective motivations by making the app more enjoyable, user-friendly, and emotionally resonant. Progression systems and intuitive design support users’ behavioral skills and capacity by making tasks easier to understand and accomplish. Features such as surprise elements and visual progress tracking increase the salience of HIV prevention behaviors by reinforcing users’ attention and engagement. Furthermore, by building new health habits through repeated in-app interactions, these elements can help replace prior disengagement with the proactive health behavior. To maximize the effectiveness of these improvements, our findings suggest that they must be implemented with careful, context-specific design considerations.

Concerns about data privacy and security were identified as a critical theme, reflecting their heightened relevance within the Malaysian context. Community experts highlighted that fear of data breaches and privacy violations could deter YMSM from engaging with the MY-Hero app [26]. This concern is consistent with prior research among Malaysian MSM, where cross-sectional findings indicate that privacy, confidentiality, and data security concerns are significant barriers to willingness to use HIV-related mHealth apps, directly influencing trust and uptake [26]. Similarly, qualitative focus group studies with Malaysian MSM have identified fear of third-party data access, uncertainties surrounding data storage and management, and broader ethical concerns about confidentiality as prominent factors shaping app acceptability [26,92]. Beyond the Malaysian context, reviews of digital health technologies have consistently identified privacy and confidentiality as persistent ethical challenges in eHealth and mHealth interventions, underscoring the importance of robust safeguards to support sustained participation [110,111]. Moreover, digital privacy risks are particularly consequential for marginalized populations in settings where same-sex behavior is criminalized, as data breaches may exacerbate stigma, discrimination, or legal vulnerability [26,112]. Thus, strong security measures, including end-to-end encryption, anonymized user data, and transparent data governance policies, are essential for fostering trust and encouraging app adoption. See Table 2 above that illustrates the linkage between each theme and each of the 5 components of the IBM model that are important to enhancing HIV prevention behaviors.

The findings highlighted several design principles that are important and unique to the Malaysian context. These included collaborating with local NGOs to enhance the reach of the MY-Hero app, adapting UI to reflect Malaysian cultural identities through customizing avatars and simplified UI, carefully designing leaderboard and interactivity features to avoid unintended consequences, and enhancing data privacy and security. These improvements were made in direct response to the Malaysian sociocultural and legal context, particularly the criminalization and stigma that this vulnerable population experiences.

This contextualization has both theoretical and practical implications. First, it suggests that the IBM model may benefit from the incorporation of a sixth construct, that is, contextual consideration. While each existing 5 construct addresses key determinants of health motivation and behaviors, our findings demonstrate that sociocultural, structural, and legal contexts fundamentally shape how those constructs operate in practice. Integrating contextual considerations can enhance the scalability, cultural appropriateness, and feasibility of interventions tailored to specific populations. The contextualization we made in the MY-Hero app also distinguishes the app from previous Western mHealth interventions by adding cultural specificity that is important to countries in Southeast Asia.

Practically, the findings underscore the importance of culturally balanced design in multicultural settings. Malaysia’s diverse population, including Malay, Chinese, and Indian communities, needed careful consideration of features and UI to ensure its acceptability, reach, and scalability. In this way, MY-Hero can serve as a skeleton for adapting mHealth interventions to other Southeast Asian countries facing similar structural and cultural dynamics. More broadly, these findings can inform the design and implementation of HIV-related mHealth interventions in other multicultural countries with concentrated epidemics, such as South Africa [113], Thailand [114], and Brazil [115,116], where cultural diversity and stigma similarly shape intervention uptake. Finally, several recommendations were not fully implemented due to budget and timeline constraints, as well as privacy considerations. By balancing expert recommendations with what is doable, this study shows how mHealth interventions should be adapted to align with real-world implementation constraints and the needs of vulnerable populations.

A key strength of this study was the use of an interdisciplinary expert panel leveraging both scientific evidence and community insight. This approach allowed us to generate a rich, contextually grounded set of recommendations relatively quickly, without the logistical and ethical complexities of conducting large-scale end-user workshops in a stigmatized population. The inclusion of community experts alongside data and health scientists helped ensure that the features were not only theoretically sound but also pragmatically relevant and respectful of Malaysian YMSM’s lived realities.

Additionally, our approach deliberately linked each identified theme to the IBM model. This theoretical grounding can help in later evaluating the app’s mechanism of action. We can hypothesize, for example, that the avatar and reward system will enhance affective motivation, leading to greater intention to test for HIV, which in turn (with easy service linkage) results in higher testing uptake. Additionally, grounding the discussion in the IBM gave the analysis a structured lens, which is a novel aspect of gamification design, as many design processes might not explicitly reference theory during feature brainstorming.

Another strength is the novelty of our focus. To our knowledge, this is one of the earliest attempts to outline a comprehensive gamified mHealth solution specifically for HIV prevention and comorbidities in Malaysian YMSM. While several mHealth interventions have been developed in Malaysia to support HIV prevention, such as the JomPrEP (University of Connecticut) app, MY-Hero differs in several important ways. For example, JomPrEP primarily focuses on facilitating PrEP awareness, access, and linkage to services through informational resources and service navigation [77,78]. In contrast, MY-Hero is a gamified “one-stop shop” for Malaysian YMSM to address not only HIV prevention efforts but also comorbid health needs, such as mental health and substance use. Additionally, by incorporating a diverse set of gamification elements, MY-Hero aims to move beyond information delivery toward a more fun, engaging, and sustained behavior change approach within the Malaysian HIV prevention landscape. It addresses a clear gap in the literature and local practice, as current HIV programs in Malaysia have yet to fully harness gamified mHealth, despite the substantial needs [68].

There are several limitations to acknowledge. Our expert panel did not include many actual end users (ie, Malaysian YMSM themselves), apart from 3 community representatives who had insight into the community’s needs. This approach was intended to minimize potential risks to YMSM participants in a study setting given the social context, but it means some user preferences might not have been fully captured. For instance, while experts assumed avatars would be popular, it is possible some YMSM might find them trivial or childish; only direct user testing can validate these assumptions. We mitigated this by relying on community experts’ knowledge and literature on user preferences. Future research should focus on pilot-testing MY-Hero with Malaysian YMSM to evaluate usability and its impact on HIV prevention behaviors and comorbidities. Second, experts were recruited using convenience sampling, which may introduce selection bias and limit the diversity of perspectives represented in the panels. Although we attempted to reduce this risk by applying strict eligibility criteria, including extensive professional experience, demonstrated expertise in relevant fields, and familiarity with HIV prevention or gamification, our panel may still reflect the viewpoints of a specific network of experts. Future research could incorporate a broader range of stakeholders, including additional community representatives, implementation partners, and end users, to further refine the intervention design and ensure broader representativeness.

By integrating expert insights from health care, technology, and community perspectives, this study highlights several key principles for the design of gamified mHealth interventions for HIV prevention among young and marginalized populations. First, interdisciplinary collaboration is essential for developing interventions that are clinically and theoretically grounded and responsive to the needs and preferences of the target users. Second, carefully designed gamification elements, including rewards, customization, and surprise, can enhance user engagement and health motivation, while some elements like competition require thoughtful and careful implementation to avoid unintended negative effects. Third, cultural adaptation and tailoring to contextual factors, including trust, privacy, and partnerships with local community organizations, are critical for ensuring acceptability and sustained engagement of mHealth interventions.

The findings demonstrate how gamification elements can be strategically aligned with behavioral targets from the IBM to support HIV prevention efforts. Beyond informing the development of the MY-Hero app, this study highlights transferable design principles and a replicable interdisciplinary development process that can guide the creation of scalable, culturally responsive gamified mHealth interventions across diverse settings. These insights also create opportunities for future interventions that integrate emerging digital technologies, such as AI, video game-based learning, and metaverse-enabled social and immersive experiences to deliver more personalized, adaptive, and engaging health support.