Internet Health Care Service Use Behavioral Pattern Among Older Adults and the Role of the Technology Acceptance and Social Ecological Theory Model: Cross-Sectional Survey.

Abstract

The rapid growth of internet health care (IH) offers older adults convenient medical services like remote consultations and health monitoring. However, its adoption among this group remains low, highlighting a significant digital divide. Understanding the behavioral patterns and determinants of IH use in the older population is crucial for optimizing digital health design and improving service accessibility. This study aimed to analyze the multidimensional influencing factors of Chinese older adults' use of IH services based on the integrated framework of the technology acceptance model and social ecological model, and explore their behavioral patterns and key driving factors. A cross-sectional study design was adopted to conduct a multistage stratified cluster random sampling survey in 3 cities in Shandong Province from May 2024 to July 2024, with a total of 1828 older adults aged 60 to 75 years included. The study uses latent category analysis to classify the use of IH service behaviors and employs multiple logistic regression, decision tree models, and structural equation modeling to analyze influencing factors and mediating pathways. Five distinct user groups were identified: nonusers (n=911), registration-dominant users (n=286), low-activity users (n=320), moderate comprehensive users (n=288), and full-service users (n=23). Multinomial logistic regression with nonusers as the reference group identified key determinants: individuals with below primary education had 96% lower odds of membership (odds ratios [OR] 0.039, 95% CI 0.012-0.084) compared to the reference group with junior college education or above in moderate comprehensive users, while male participants had higher odds of being full-service (OR 1.980, 95% CI 1.126-3.514) or moderate comprehensive (OR 1.310, 95% CI 1.012-1.705) users. Older age was consistently associated with lower adoption across all classes. Full-service users exhibited exceptionally high social support (OR 4.502, 95% CI 3.601-5.627), while moderate comprehensive users showed the highest technology acceptance (OR 2.803, 95% CI 2.355-3.342). The decision tree model (area under the curve of 0.94) found the optimal path: sufficient social support (≥2), good health status (>5), and high technical acceptance (≥30) yield the highest use probability (92%→96%). Mediation analysis indicated that social support influences usage willingness through both direct and indirect pathways. The direct effect was 0.712 (95% CI 0.552-0.972; P<.001). Among indirect pathways, technology availability and practicality accounted for the largest proportion of mediation (19.7%, 95% CI 16.8%-22.6%), followed by technology acceptance (13.7%, 95% CI 11.1%-16.3%) and social influence (8.9%, 95% CI 6.9%-10.9%). Optimizing age-friendly design, strengthening social support networks, and improving technological usability are keys to increasing the adoption of IH services among the older population. Future policies should develop targeted intervention strategies for different user groups to narrow the digital health divide.