Patient Perspectives on AI- and XR-Enabled Telemedicine: A Cross-Sectional Survey in Romania

Abstract

<b>Background and Objectives</b>: As artificial intelligence (AI) and extended reality (XR) enter routine care, understanding patient acceptance is essential. We assessed attitudes toward AI/XR-enabled telemedicine among Romanian patients and examined correlates of acceptance. <b>Methods</b>: We analyzed 198 survey responses to a 20-item questionnaire. Ordinal items were encoded 1-4. The Acceptance Index measured trust in AI, perceived improvement in care, and willingness to choose AI-assisted visits (on a 1-4 scale). <b>Results</b>: Respondents were predominantly 31-50 years old (62.6%) and university educated (76.2%); 27.3% reported prior experience with AI/XR. Acceptance averaged 3.27 ± 0.59 (α = 0.780) and did not differ by age (<i>p</i> = 0.922). Prior users showed higher acceptance than non-users (3.47 ± 0.47 vs. 3.19 ± 0.59; <i>p</i> = 0.0011). Knowledge (ρ = 0.189, <i>p</i> = 0.048) and perceived accessibility (ρ = 0.229, <i>p</i> = 0.016) correlated with acceptance; privacy concern did not (ρ = 0.072, <i>p</i> = 0.455). Subgroups: Prior use was associated with higher acceptance across education levels, with a significant effect in secondary education (Holm-adjusted <i>p</i> = 0.029; Cliff's δ = 0.56). Ordinal logistic model: higher willingness to pay (OR 6.81, 95% CI 3.39-13.66, <i>p</i> < 0.001) and greater perceived accessibility (OR 1.83, 95% CI 1.03-3.24, <i>p</i> = 0.040) independently predicted choosing AI-assisted visits. <b>Conclusions</b>: Patient acceptance of AI/XR-enabled telemedicine was moderate to high, strongest among prior users, and increased when access felt easy. Knowledge modestly supported acceptance; privacy concerns did not diminish it. Clear value propositions, streamlined access, and optional exposure pathways may enhance informed uptake.

Ähnliche Arbeiten