This study used a descriptive, qualitative design using in-person, semistructured interviews. Through this qualitative research, researchers aimed to understand the specific experiences of older adults and their caregivers with digital health services, identify their usage intentions, influencing factors, and service needs, and provide recommendations to advance the development of digital health services. Furthermore, we adhered to the Standards for Reporting Qualitative Research guidelines [26] to present the findings of this research (see Multimedia Appendix 1).

The study population included older adults with experience using digital health services and their caregivers. The researchers adopted a purposive sampling method to select research sites in Hefei City, Anhui Province, China, specifically including (1) seven community health service centers: Yaohai District (eastern Hefei), Shushan District (southwestern Hefei), and Baohe District (southern Hefei), which cover the geographical distribution of both old urban areas and emerging development zones and (2) the Geriatric Inpatient Ward of the First Affiliated Hospital of Anhui Medical University. This hospital is the largest public Grade A tertiary general teaching hospital in Anhui Province. The inclusion and exclusion criteria for older adults and caregivers were shown in Table 1. Eligible older adults and caregivers were sought, and participants were briefed on the study objectives and procedures. If they showed interest, they were invited to participate in the study interview. Data saturation was considered to have been achieved when semistructured interviews with successive participants failed to generate any new open-ended codes, categories, or thematic insights on 3 consecutive occasions. The research team then collectively confirmed that no additional meaningful information would be derived from further interviews.

This study has obtained ethical approval from the Biomedical Ethics Committee of Anhui Medical University (Ethical approval number: 83242378). In accordance with the Declaration of Helsinki, the study adheres to the principles of informed consent and privacy protection, and all investigations were conducted with informed consent. All data are strictly confidential and shall be used exclusively for this study. Participants took part in this study voluntarily and did not receive any financial compensation.

Based on the research objectives, a semistructured interview guide was developed through an extensive review of relevant literature and discussions among research team members. Before the formal interview, we conducted a preinterview with 2 older adults and their caregivers and adjusted the interview questions based on the results. The final interview outline was shown in Multimedia Appendix 1. Prior to the formal interviews, we screened the research participants to ensure that they had experience in using digital health services. At the very beginning of the interview, the research participants were provided with descriptions of the 4 types of digital health services, and then they chose the services they had used in their lives, including health monitoring (Mo), health management (Ma), intelligent life assistance (I), and online health care services (O). Detailed descriptions of the digital health services refer to the outline of the interview in Multimedia Appendix 1. Interviews were conducted in Chinese by 2 researchers (JYR and CXG), who had no prior contact with the participants. Interview data were collected between December 2023 and February 2024 via in-person, semistructured, in-depth interviews, supplemented by field notes. Interviews were arranged at locations convenient to participants and in quiet, undisturbed settings to prevent interruptions. Before each interview, participants’ permission to record audio was obtained.

Conventional content analysis guided our analytic approach [27]. Audio recordings were transcribed verbatim immediately after each interview. The textual data were then imported into NVivo (12.0 software; QSR International Pty Ltd). To protect privacy, participants’ real names were replaced with letters. Data analysis, coding, and theme extraction were performed using NVivo 12.0 software. Two researchers (co-first authors JYR and CXG), both of whom had training in qualitative methods, repeatedly read the interview materials (including audio transcripts and field notes) to capture the exact meanings participants intended. JYR repeatedly read the first 4 transcripts, marked all meaning units relevant to the phenomenon, and generated an initial list of codes. Subsequently, JYR and CXG collaborated to develop the codebook. JYR drafted names, definitions, and typical excerpt examples for each initial code, forming the preliminary codebook. The 2 researchers then jointly reviewed the first 4 transcripts, checking the fit between each code and its data source line by line, merging similar codes, refining definitions, and resolving any inconsistencies through discussion. Thereafter, the codebook was applied to subsequent transcripts. When new meanings emerged that could not be captured by existing codes, new codes were added and discussed again for revision. This iterative process continued until no new codes emerged and the codebook was considered saturated and stable.

The final determination of themes was completed by all authors (JYR, CZG, QQC, TY, YL, YXZ, and LFX). After all transcripts had been coded, JYR exported the final list of codes with their definitions and preliminarily clustered them into subcategories and candidate themes based on their shared latent meanings. The research team then held multiple discussions, engaging in repeated debate and revision of the naming, definitions, and boundaries of the themes until consensus was reached. Finally, each theme was assigned a formal name and clear definition, and the most representative participant quotations were extracted as examples.

Before initiating the study, researchers received specialized training in interview techniques and conducted pilot interviews, refining the interview guide based on their outcomes. Before formal interviews, recording equipment was tested to ensure optimal functionality and audio quality, and written informed consent was obtained from all participants. Interviews with older adults and their caregivers were independently administered by 2 postgraduate nursing students, with no cross-discussion between the interviewers during data collection. All textual data were anonymized and coded; researchers maintained a strictly neutral stance throughout the analytical process to ensure objective interpretation, and study findings were subjected to rigorous review by the research team to uphold methodological objectivity. In cases of discrepancies in extracted themes, the research team engaged in collective deliberation to reach consensus on final thematic classifications, thereby ensuring the reliability and consistency of the study’s results.

The study included 18 older adults with a mean age of 70.17 (SD 6.21) years and 17 caregivers with a mean age of 50.65 (SD 15.37) years. Ten of these participants formed matched older adult and caregiver dyads. Additional demographic characteristics are presented in Multimedia Appendix 1.

The qualitative analysis began with 630 initial codes. Through an iterative process of comparison and consolidation of similar codes, these were refined into 470 distinct codes. These codes were then organized into a hierarchical structure, resulting in 12 subthemes and 4 main themes: (1) usage and preference of services, (2) specific service experiences, (3) influencing factors, and (4) suggestions and expectations. An example of the coding process and the complete coding framework are provided in Multimedia Appendix 1. The similarities and differences in perceptions of digital health services between older adults and caregivers were summarized in Table 2.

This study identified four key findings regarding digital health service experiences among home-dwelling older adults and their family caregivers. First, both groups primarily used health monitoring and basic health management functions but showed distinct preferences for service types. Second, they shared positive experiences of improved convenience and health support, while older adults uniquely reported dignity-related distress and physical discomfort, and caregivers emphasized privacy and safety concerns. Third, usage was driven by health needs and social influences but blocked by low digital literacy, economic constraints, and time shortages. Fourth, both groups prioritized age-friendly design, authoritative platforms, on-site guidance, and family or peer support to optimize future services.

In general, the types of services frequently used by older adults include health monitoring, health management, artificial intelligence–powered living assistance, and Internet-based medical services, which echo the categories of customer-system interaction, decision support, and service delivery in the World Health Organization (WHO) classification of digital health interventions [28]. In addition, regarding health monitoring and management, older adults showed high acceptance and demand. Older adults tend to use functions that support their autonomy in daily life and show a strong preference for low-cost or free services, such as using an electronic blood pressure monitor and an electronic glucose meter, and accessing online health information. Although digital technology has shown great potential for improving health services for older adults [29], the penetration rate of digital technology in transitional and low-income countries is low, older adults’ purchasing power is limited, and their health literacy is low. Moreover, older adults in rural and remote areas often face challenges, including inadequate infrastructure and limited access to technology [30]. Therefore, it is reluctant to try services with high technical access requirements, complex operational processes, and high use costs, such as hip protective airbags, smart homes, and body sensors.

The continuous innovation and widespread adoption of digital health technologies have led both older adults and their caregivers to recognize the advantages of digital health services. Research indicated that the fulfillment of health needs and positive user experiences facilitate the willingness of older adults and caregivers to adopt digital health services, which is consistent with the Technology Acceptance Model [41], in which perceived usefulness and perceived ease of use are key influencing factors. In addition, subjective norms, such as encouragement from family members, peers, and health experts and the sharing of positive usage experiences, serve as important references for older adults. This finding aligns with the Technology Acceptance and Use Unified Theory (UTAUT), which emphasizes social influence as a critical factor in technology adoption [42]. Furthermore, older adults demonstrated positive attitudes toward government-supported digital health services [43], which is consistent with the concept of facilitating conditions in UTAUT, reflecting that policy-level promotion can enhance older adults’ awareness and acceptance of digital health services.

Nevertheless, they continue to face multiple barriers in actual use. Older adults generally have low digital health literacy, making it difficult for them to adapt to rapidly evolving internet-based health technologies [22]. The study found that they exhibited avoidance and rejection attitudes toward some digital health services and were more inclined to trust the diagnoses, treatment opinions, and effects of offline medical services. This preference reflects a tendency to adhere to familiar practices even when new alternatives are available. In addition, the financial burden associated with purchasing digital devices or accessing paid digital health services remains a salient challenge for both low-income older adults and caregivers [34].

Caregivers also face barriers related to their roles. Lack of time is a major constraint, particularly for younger adult caregivers, who struggle to balance work, caregiving responsibilities, and personal life, leaving them little time to assist older adults with digital health resources [44]. Although these caregivers recognize the potential benefits of such services, they find it difficult to actively use them, creating a gap between intention and behavior. Older spousal caregivers encounter technical difficulties similar to those faced by older adults. In addition, caregivers expressed heightened concerns regarding safety and privacy, particularly regarding the security of online platforms and the risk of misuse of personal information. They generally perceive older adults as being in a passive position in the digital society and worry about the safety of older adults using digital health services independently, tending to restrict their use [45].

The Technology Acceptance Model and UTAUT models emphasized rational cognitive factors and ignored the influence of additional factors on users’ behavioral intentions, such as time, cost, safety, emotions, and whether they possess the necessary skills to use digital services. Therefore, these classic models were insufficient to explain older adults’ adoption behaviors of digital health services. Some studies [46,47] had attempted to extend these models by adding new variables or integrating with other theoretical frameworks to make them more applicable to older populations in digital health. Future research should extend these classic models by integrating insights from qualitative findings, constructing a theoretical framework that is more tailored to older adults for digital health services.

Based on feedback from older adults and their caregivers, this study proposes strategies to optimize the adoption of digital health services from 3 dimensions, namely government, service providers, and social support. The government should rely on grassroots institutions such as communities to establish an integrated digital health literacy education, training, and promotion system that covers the continuum from home to community and from offline to online [48], enhancing older adults’ comprehension of and experience with digital health technologies, thereby facilitating the widespread adoption of digital health services among this population. Additionally, it is recommended that economic incentives be implemented to reduce usage costs—for example, providing subsidies for digital devices and including relevant expenses in medical insurance reimbursement—to alleviate financial barriers for older adults and low-income groups. For instance, in Nanjing, China, the government provides digital devices, such as smart bracelets, to older adults as a welfare benefit [49,50]. Older adults can access services either through automatic feedback from the devices or through active calls, improving service accessibility and acceptability. Simultaneously, the government should strengthen platform supervision, improve mechanisms for digital privacy protection, and establish authoritative information communication channels jointly led by multiple formal medical institutions to enhance public trust in and adoption of such services. Digital health service providers emphasize the need for age-appropriate, personalized design centered on older adults’ needs, with a focus on simplifying operational processes, optimizing user interface interactions, and accounting for age-related physical and cognitive characteristics [51]. Design should be grounded in older users’ actual usage habits to lower technological barriers. Previous research indicates that interface adaptability and user-centered design can significantly enhance older adults’ engagement [52]. For example, older adults have suggested that platforms should offer multiple input and output methods (eg, text and audio) to meet the needs of users with varying educational levels and processing speeds. In terms of after-sales support, it is recommended to establish user profiles, conduct regular follow-ups on usage experiences, and set up after-sales service stations in communities, pharmacies, and other locations to provide continuous support. Additionally, service providers should formulate reasonable pricing in line with government guidance, offer discounts or medical insurance support for specific older adult groups, and balance social responsibility with economic benefits. Social support plays a crucial role in helping older adults improve their digital health literacy. “Digital reciprocity”—where adult child caregivers support older adults in learning and using digital health services—is recognized as the most direct and practical approach. Face-to-face guidance from adult child caregivers can encourage older adults to adopt an open attitude toward digital technology, gradually guiding them to accept and actively learn health knowledge, thereby enabling older adults to receive both technical support and positive emotional incentives [53]. If adult child caregivers are unable to assist due to time constraints or if older adult spousal caregivers themselves face barriers to using digital technology, social forces such as peer groups, community organizations, or young volunteers can be leveraged to establish a “community school” model, forming an extra-familial reciprocity structure [41]. This mechanism can help older adults overcome technology anxiety, better perceive the value of digital health services, and enhance their intention to use and self-efficacy.

This study had several limitations. First, regarding the sample selection, the study was confined to Hefei, China, which may limit the generalizability of the findings to regions with different health care and digital infrastructure contexts. Second, to respect participants’ willingness and to capture diverse attitudes and experiences, older adults and their caregivers were not fully paired: only 10 dyads were matched, while the remaining participants were recruited independently. This partial pairing reduced the strength of dyadic-level interpretations, although separate analyses still provided valuable insights into each group’s experiences and needs. Third, all participants had prior experience with digital health services, which may have introduced a positive bias toward perceived benefits and facilitators. Future research should adopt large-scale, multicenter sampling strategies, prioritize the recruitment of fully paired older adult and caregiver dyads, and include participants with varying levels of digital health experience to enhance representativeness.

Beyond sample-related improvements, future studies should also investigate the heterogeneity of digital health services themselves. Specifically, within the same population, older adults’ experiences, acceptance, and barriers may differ significantly across various types of digital health services. Exploring these intraindividual differences would help identify which service characteristics are most conducive to adoption and sustained use among older adults. Such insights are critical for tailoring service design and implementation strategies to diverse user needs and contexts.

This study analyzed the usage of digital health services from the dual perspectives of older adults and their caregivers. Although all participants could benefit from digital health services, their willingness to use digital health technologies remained low due to various barriers, including low digital health literacy, weak economic foundations, and inadequate service security. Therefore, it is suggested that future promotion and optimization of digital health services be advanced collaboratively across three dimensions: government guidance, service provider optimization, and the construction of social support systems. This integrated approach aims to improve the user experience for older adults and caregivers, facilitate sustained adoption of these services, and ultimately enable older adults to benefit from digital health technologies.