Patient Satisfaction With Autonomous Artificial Intelligence–Assisted Telemedicine In Cataract Care: A Cross-Sectional Study Of Preoperative Assessment And Postoperative Follow-Up

To evaluate patient satisfaction with autonomous artificial intelligence (AI)–assisted telemedicine used for preoperative assessment and postoperative follow-up in routine cataract surgery.

Ophthalmology Department, Ashford and St Peter’s Hospitals NHS Foundation Trust, London, United Kingdom.

 Cross-sectional observational study evaluated patient satisfaction following the use of an autonomous AI-assisted telemedicine system (Dora) (Ufonia limited, Oxford, England) for routine cataract surgery pathways. The system conducted preoperative assessment calls 1–2 weeks before face-to-face assessment and postoperative follow-up calls 3 weeks after surgery. Patient satisfaction was anonymously measured using a 5-point Customer Satisfaction CSAT questionnaire at the end of each AI-led call, with both quantitative scores and qualitative feedback obtained from transcribed audio data.

A total of 2,473 patients were included, with 1,538 receiving postoperative and 935 preoperative AI-assisted telephone calls between August 2024 and January 2026. The median age was 71 years (range 40–100), with 60% female. In the postoperative cohort, 76.0% rated satisfaction as CSAT 5, with a mean CSAT of 4.63 and an overall satisfaction rate of 91.35%. In the preoperative cohort, 65.9% rated CSAT 5, with a mean CSAT of 4.53 and a satisfaction rate of 88.88%. Notably, 27% of patients assigning low CSAT scores (1–2) provided positive qualitative feedback, indicating discordance between numerical scores and narrative responses.

Autonomous AI-assisted telemedicine achieved high patient satisfaction for both preoperative assessment and postoperative follow-up in routine cataract surgery. These findings support its feasibility and acceptability as part of cataract care pathways, while highlighting the value of combining quantitative satisfaction scores with qualitative patient feedback.