Patient‐Centred Communication About AI‐Related Information in Perioperative Care: An Attention‐Controlled Randomised Trial of Anxiety, Perceived Control and Interpretive Preparedness

BACKGROUND OR CONTEXT: Patients encountering artificial intelligence (AI) in health care may be uncertain about whether AI supports or replaces clinicians, who remains accountable and how general AI-related information applies to their own care. These are questions about how clinicians talk about AI rather than about how AI performs, and they may be particularly salient before surgery. This trial evaluated communication about AI-related information only: no AI system, clinical algorithm, device or decision-support tool was used, tested or applied to any participant's care. OBJECTIVE: The aim of this study was to evaluate whether a brief, public-informed, clinician-delivered protocol of structured AI communication improves immediate interpretive preparedness before surgery compared with time-matched attention-control supportive communication. DESIGN: Single-centre, two-arm, parallel-group, attention-controlled superiority randomised trial with 1:1 allocation. SETTING AND PARTICIPANTS: Adult patients scheduled for elective surgery at Ağrı Training and Research Hospital, Ağrı, Türkiye, were randomised 1:1 to structured AI communication (n = 60) or attention-control supportive communication (n = 60). INTERVENTION: Structured AI communication was a 5-7-min nurse-delivered session explaining that AI may support but does not replace health professionals, that responsibility for clinical decisions remains with the care team, that AI-related information may be general or probabilistic, and that such information should be interpreted within the patient's own clinical context. Attention-control supportive communication was time-matched nurse-patient contact about routine perioperative preparation without AI-specific content. MAIN OUTCOME MEASURES: The primary outcome was pre-surgery anxiety measured using the Amsterdam Preoperative Anxiety and Information Scale (APAIS) anxiety subscale. The key secondary outcome was perceived control. AI-related uncertainty, information clarity and trust in the clinical team were measured immediately after the session as study-specific communication-uptake indicators rather than as validated independent clinical endpoints. RESULTS: All 120 randomised participants completed the pre-surgery endpoint assessment. Pre-surgery anxiety was lower after structured AI communication than after attention-control supportive communication (9.28 ± 2.55 vs. 12.48 ± 3.01; baseline-adjusted mean difference [MD], -2.56; 95% CI, -2.87 to -2.26; p < 0.001). Structured AI communication was also associated with higher perceived control (25.83 ± 4.04 vs. 22.62 ± 3.90; adjusted MD, 3.73; 95% CI, 3.34 to 4.12; p < 0.001), greater information clarity and trust, and lower AI-related uncertainty. Clinically meaningful anxiety reduction (≥ 2 APAIS points) was more frequent after structured AI communication (48/60, 80.0%) than after attention-control supportive communication (1/60, 1.7%). The unblinded design and the immediate self-reported assessment may have amplified these differences through expectancy, novelty, social desirability and measurement reactivity. The responder result should therefore be interpreted as a short-term self-reported difference rather than evidence of durable clinical benefit. DISCUSSION AND CONCLUSIONS: In a single centre, a brief public-informed protocol was associated with better immediate self-reported psychological and interpretive preparedness before surgery. Structured AI communication about AI-related uncertainty and human accountability is a plausible patient-centred component of technology-enabled care. These findings do not show that AI technology itself improves patient outcomes, and they do not establish durable anxiety reduction, behaviour change or clinical benefit beyond the immediate preoperative window. LIVED EXPERIENCE OR PUBLIC CONTRIBUTION: Three lived-experience advisors reviewed patient-facing materials during protocol refinement. After analysis, a six-member public advisory group reviewed plain-language findings and advised on interpretation, practitioner-facing wording and calibrated trust, accountability and human oversight. Their feedback informed the Discussion and Practitioner Points. PATIENT AND PUBLIC CONTRIBUTION: Three lived-experience advisors with previous sur cal experience reviewed the patient-facing communication materials during protocol refinement. After analysis, a six-member public advisory group reviewed a plain-language summary of the findings and informed interpretation, practitioner-facing wording and the framing of calibrated trust, accountability and human oversight. TRIAL REGISTRATION: This randomised controlled trial was prospectively registered at ClinicalTrials.gov (Identifier: NCT07393945).