Reliability of telemedicine for real-time paediatric ophthalmology consultations

Abstract

Background/aims To assess the accuracy of real-time telemedicine to diagnose and manage paediatric eye conditions. Methods Design: Prospective, non-inferiority study analysing agreement in diagnoses and management plans between telemedicine and in-person examinations. Setting: Paediatric ophthalmology clinic. Population: Children 0-17 years, English-speaking or Spanish-speaking, able to participate in age-appropriate manner, either previously seen by the optometrist and required ophthalmology referral or newly referred from outside source. Procedures: Paediatric optometrist conducted examinations using digital equipment and streamed live to a paediatric ophthalmologist who recorded diagnoses and management plans, then re-examined patients in-person. Subjects were masked to the fact they would see the ophthalmologist in-person, same-day. Main outcome measures: Discrepancy in management plan or diagnosis between telemedicine and in-person examinations. Non-inferiority threshold was <1.5% for management plan or <15% for diagnosis discrepancies. Results 210 patients participated in 348 examinations. 131 (62.4%) had strabismus as primary diagnosis. In these patients, excellent and almost perfect agreement was observed for angle measurements (intraclass correlation coefficients=0.98-1.00) and disease categorisation (kappa=0.94-1.00) (p<0.0001 in all cases). No primary diagnoses changed, and no management plans changed following in-person examination. 54/55 patients who consented for surgery at the initial visit did so while masked to receiving an in-person examination. Families felt comfortable with the quality of the telemedicine examination (98.5%) and would participate in another in the future (97.1%). Conclusion Paediatric ophthalmic conditions can be reliably diagnosed and managed via telemedicine. Access for underserved populations may be improved by collaboration between ophthalmologists and optometrists using this technology.