JMIR Publications

ABSTRACT

Background:

Immersive extended reality (XR) promises to transform healthcare, but public acceptance and user-side determinants of acceptance remain largely unknown.

Objective:

To estimate national XR acceptance and identify sociodemographic, psychosocial, health, and digital determinants among adults in China.

Methods:

We conducted a national cross-sectional survey (June–September 2024) of 35,861 adults across 800 communities in China. We measured XR acceptance (0–100) and 141 potential predictors spanning demographic, adversity, personality, literacy, lifestyle, physical, and psychosocial domains. Survey weights were applied to match the national age–sex distribution. Key predictors were identified using hierarchical block-wise linear regression with false discovery rate (FDR) control, 10-fold cross-validated elastic net, and a weighted classification tree.

Results:

Mean acceptance was 63.1 (95% CI 62.8-63.5), varying by province (47.9-72.3). It peaked young (18-24 women, 30-34 men) then declined; sex differences were minimal (<3 points). Acceptance varied across 15 chronic conditions (lowest: rare diseases, 54.2; highest: hyperlipidaemia, 62.3). Of 141 predictors, 76 were FDR-significant, 70 of which were retained by elastic-net. Acceptance was strongly associated with socioeconomic factors (higher social status, standardized β=0.17; higher youth socioeconomic status, 0.14; better youth economic environment, 0.06), digital capital (prior digital-health use, 0.15; eHealth literacy, 0.09), and key traits (self-efficacy, 0.10; personal meaning, 0.06). Other positive predictors included having two types of medical insurance (0.07), stable sleep duration (0.07), and childhood psychological abuse (0.07). Strong negative predictors included older age (-0.08), couple-only household (-0.08), childhood sexual abuse (-0.08), attention-deficit/hyperactivity disorder (-0.07), more siblings (-0.07), and childhood physical abuse/violence exposure (both -0.05) (all adjusted-p<0.05). A six-node classification tree had modest discrimination (test AUC=0.61, accuracy=0.681) but high specificity (0.903) and low sensitivity (0.242), effectively identifying likely non-acceptors.

Conclusions:

XR acceptance in China is moderate but heterogeneous, varying by region, age, and health status. Predictors cluster into four key domains: structural-environmental, trait-resilience, social-digital capital, and behavioral-health. While overall prediction is difficult, our model reliably identifies likely non-acceptors, enabling targeted outreach and digital equity efforts.