To tackle the rising prevalence and underreporting of the “3-highs” and obesity rates in Singapore [10-12,14], we integrated a suite of point-of-care assessments to enable early detection and monitoring. These included artificial intelligence–assisted eye screening; cordless upper-arm blood pressure and heart rate monitors; smart scales to measure weight, BMI, and body fat composition; and capillary blood glucose measurement.

To create a safe and private environment for our telemonitoring program, we repurposed the biosafety booths, which were originally designed for mass swabbing exercises [35] (Multimedia Appendix 1). The adaptable modular design of the kiosks allowed us to customize the telemonitoring program according to specific requirements. In addition, the kiosks were equipped with an internet connection that enabled the uploading of health data, printing of health reports, and conducting teleconsultation when necessary.

At each implementation site, trained health ambassadors were stationed to provide real-time assistance and support to users, particularly those with limited digital or health literacy. These ambassadors helped users navigate the telehealth kiosks, conduct health assessments, and facilitate understanding of their results.

The integration of a referral management system was a pivotal and essential component of our program. Recognizing the importance of seamless coordination and collaboration with health care providers, we established a streamlined procedure to ensure that users in need could be identified and referred to polyclinics and community nurse posts to receive timely, targeted, and subsidized diagnosis and follow-up care.

The user journey in our telemonitoring prototype is illustrated in Figure 1. Health ambassadors would first verify users’ identities using a secure registration system on smart tablets, then guide the users through the profile surveys with health history, various measurements, and communication of the results, as well as any remarks from the users. Users’ eye reports were sent to the Singapore National Eye Center for manual verification, while the other devices had been approved by the Health Sciences Authority and were able to provide immediate results under the operation of trained health ambassadors. In the event of abnormal readings, users would be directed to the referral system for professional diagnosis and follow-up care.

Facilitators and barriers to telehealth implementation were identified through our scoping review [37] and an earlier local qualitative study involving 15 older adults, 12 health service providers, and 2 telehealth providers. These determinants were organized according to the social ecological model (SEM) [36] across 5 levels, which are individual, interpersonal, technological, system, and policy levels. Key barriers included low digital literacy, difficulty building rapport with telehealth deliverers, reliance on internet connectivity, insufficient physical examination, insufficient manpower to deliver telehealth services, as well as ethical and data security concerns.

Based on the determinants identified, we selected a set of implementation strategies from the ERIC taxonomy [38] and operationalized them across the SEM levels to address the multilevel determinants.

At the individual level, the trained health ambassadors provided users with on-site assistance, friendly and multilingual materials, as well as personalized health feedback upon receiving the printed health reports to enhance engagement. At the interpersonal level, ambassadors from different age groups were recruited and trained using a train-the-trainer model that covered not only basic health knowledge and health assessment skills but also empathy and communication techniques. All training materials were reviewed by professional medical advisors to ensure clinical accuracy and alignment with local guidelines. The health ambassadors were paired to comanage each kiosk—typically matching a younger ambassador who was more adept with technical operations with an older ambassador who brought stronger interpersonal communication skills (Multimedia Appendix 2). This intergenerational pairing was intentional to enhance both technical proficiency and user engagement, fostering a warm and age-friendly environment that resonated with older adults. Technological strategies included centralized technical support, simplified digital workflows, and iterative user testing to improve system usability and reliability. At the system level, our telehealth kiosks were deployed in accessible community locations through phased implementation, in collaboration with community, health care, and corporate partners. Safety protocols and feedback loops were established to guide iterative improvements. At the policy level, funding was secured through university, community sources, and corporate sponsorships, while academic partnerships were formed to support planning, training, and sustainability.

Over the 8 weeks in phase 1, a total of 627 screenings for 277 males and 350 females (mean age 57.0, SD 10.2 y) were conducted and 66 health ambassadors were trained, while another 45 registered volunteers were waiting for training. Phase 2 completed 103 screenings (30 males, 68 females, mean age 65.0, SD 9.5 y, of whom 5 were returning users. Phase 3 reached 79 users (38 males, 37 females, mean age 68.0, SD 9.2 y), including a total of 4 returning users.

Across all 3 phases, similar recruitment strategies were used, although media release proved most effective in phase 1. In phase 2 (n=103, 61 responses were received with a response rate of 59.2%), most users learned about CTS through publicity materials (n=22, 36.1%) and door-to-door publicity (n=15, 24.6%), while others cited passing-by (n=10, 16.4%), social media (n=8, 13.1%), and referrals from their social network (n=6, 9.8%). Over half (n=34, 55.7%) registered independently, and those who received assistance were mainly supported by health ambassadors (n=15, 55.6%), followed by family members or helpers (n=9, 33.3%) and community center staff (n=3, 11.1%). In phase 3 (n=79, 12 responses were received with a response rate of 15.2%), awareness was primarily through passing-by (n=5, 41.7%) and door-to-door publicity (n=4, 33.3%), with fewer citing publicity materials (n=3, 25%). Fewer registered independently (n=5, 41.7%), and all assisted registrations were supported by health ambassadors.

During phase 1, partnerships were established with 4 community partners, 2 medical partners, and 6 corporate partners. Funding was received from 5 sources, including research grants, university CEP funding, as well as corporate sponsorships. As the service progressed into phases 2 and 3, partnerships expanded to 6 grantors, 10 government and community partners, 2 medical partners, and 8 corporate sponsors. In addition, an additional academic partnership was established.

CTS operations in phase 1 spanned 236 hours, during which health screenings and ambassador training were conducted concurrently. The cancellation or no-show rate was approximately 26.8%. In phase 2, the program operated for a total of 290 hours (243 h for screening and 47 h for training), with a no-show rate of 15.6%. Phase 3 recorded 126 total operating hours (90 h for screening and 36 h for training), with a cancellation or no-show rate of 11.2%.

The implementation costs for all 3 phases were estimated based on the sponsorship with actual claims made to research funding and university CEP funding. The implementation costs for phase 1 and phases 2 and 3 were S $19,884.73 (US $14,913.55) and S $3806.24 (US $2809.01), respectively.

The project remained free of adverse events throughout all 3 phases, with no occurrences of COVID-19 infections, injuries, or data breaches reported. This absence of incidents underscores the effectiveness of the safety protocols and the efforts of all stakeholders in maintaining a secure and safe environment.

We referred 163 (26%) users to the community nurse post and Singapore National Eye Center for follow-up care in phase I. We failed to consistently track referral rates in phases 2 and 3, as the medical partners involved operated under distinct referral workflows and documentation systems. In addition, a resurgence of COVID-19 cases during these 2 phases caused interruptions in referral processes, particularly affecting coordination between community sites and health care providers. These variations made it difficult to obtain a standardized and complete record of referral outcomes across both locations.

In phase 1, we received a total of 389 user feedback responses (response rate=62%), and the responses were predominantly positive (Table 1). In addition, 66.3% (258/389) of users were able to book an appointment independently and the rest required some assistance from either their social networks or our health ambassadors. Furthermore, 83.3% (324/389) of the users were satisfied with the waiting time. Moreover, 94.1% (366/389) of the users were happy with the waiting time, 99.7% (388/389) of the users found the verbal instructions given by the health ambassadors satisfactory, and 99.7% (388/389) expressed overall satisfaction with the experience. In general, 98.5% (383/389) of the users expressed interest in continuing to use the CTS as a community resource for health monitoring.

In phases 2 and 3, a total of 73 user feedback responses were collected (61 from phase 2 and 12 from phase 3). Overall, users expressed high satisfaction and perceived CTS as a usable service. However, response rates were lower compared with phase 1, which several users attributed to the length and complexity of the questionnaire, particularly challenging for older users. Table 2 presents the survey items along with a summary of the subscale scores.

User feedback from phase 1 reflected strongly positive perceptions of CTS as a health initiative tailored for older adults. Many users appreciated the convenience of receiving multiple health screenings in one location:

Users also described the service as smooth and efficient and praised the professionalism and warmth of the health ambassadors, for example:

Several users appreciated the benefits of such community-based telehealth kiosks for the older population and expressed hope that CTS would be expanded to other regions, calling it:

While CTS received a high level of acceptability across all 3 phases, user feedback also highlighted potential areas for improvement, for instance, simplifying the registration process, enhancing the environment inside the kiosk and at the waiting areas, and improving the clarity of health reports and publicity materials. A summary of these suggestions and representative quotes is provided in Table 5. Building upon the insights collected from phase 1, enhancements were implemented in phases 2 and 3, such as (1) assigning health ambassadors to timely answer users’ inquiries via WhatsApp (Meta) messages; (2) streamlining the registration process, enhancing the user-friendliness of the registration platform, and sending booking confirmations and reminders through SMS text messages; (3) translating the publicity materials into not only Chinese but also Malay and Tamil; (4) involving medical students and health professionals to enhance the abnormality detection algorithms by taking possible age-specific reference ranges into consideration; and (5) improving the physical environment by providing fans both inside and outside the telehealth kiosks, as well as increasing the seating capacity in the waiting area.

In addition to these readily achievable improvements, three more popular suggestions were shared via the health ambassadors: (1) extending the current spectrum of health assessments, (2) providing additional educational health information, and (3) upgrading the current referral management system.

Technology advancements have enabled individuals to remotely monitor their physiological parameters outside the clinical settings and seek actionable insights for improving health outcomes [65,66]. Besides income, education, employment, and environment, digital technology has emerged as a new key determinant of health [67,68]. It is noteworthy that while digital health interventions are intended to reduce health inequalities by extending access to health services and addressing unmet health needs through personalized care, there is a concern that they might unintentionally exacerbate health inequities to some extent [69-71]. Therefore, digital inclusion–informed strategies are needed in designing community-based telehealth programs.

In response, we adopted a “phygital” approach, combining both digital and physical elements. While we acknowledge that the implementation of the health ambassador scheme in CTS may limit the generalizability of our findings to other community-based telehealth programs that lack or have minimal human support, as discussed in a study by Zhang et al [30], this approach has the potential to bring significant value to telehealth program delivery. For example, a study by Courtney et al [29] raised several concerns, such as older adults overreacting to minor abnormalities in the health data, the potential risk of using the kiosk as a substitute for regular or urgent health care services, and the false reassurance of blood pressure alone being a comprehensive indicator of overall health status.

In CTS, our well-trained health ambassadors could play a crucial role in mitigating these concerns. Besides conducting health measurements and generating health reports, they assisted users in explaining their measurements based on local clinical guidelines and taking notes of some remarks (eg, their perceived health and self-observed individualized reference ranges) in the reports. This combination of subjective perceptions and objective measurements enhances the delivery of more person-centered care [72]. In addition, health ambassadors could offer actionable advice for lifestyle modifications to positively impact health outcomes; they can also actively encourage all users to undergo regular health screenings at medical institutions and refer them to appropriate medical partners for follow-up care when necessary. Such a comprehensive approach serves to not only minimize the potential misinterpretation of health data but also foster a proactive mindset toward maintaining overall health and well-being.

Community-based telehealth programs allow users to regularly monitor health at community-based facilities, and thus generate a comprehensive record of their health data. The integration of these diverse objective health data points, subjective self-perceived health status, potential individualized reference range, and lifestyles, together with users’ participation in community events and health promotion activities, could support primary care providers in forming a more complete and personalized understanding of an older individual’s health. This, in turn, enables the delivery of more effective, customized, and equitable health care services.

However, challenges arise as these data often remain disconnected from primary care records, possibly due to data inconsistency, lack of standardized data handling and ownership protocols, concerns of breaching data privacy and confidentiality, as well as the risk of generating inaccurate or incomplete data [73,74]. Addressing these gaps and establishing mechanisms to bridge them could be a dedicated effort moving forward.