Development of a new telemedicine tool for the screening of thyroid eye disease

Introduction Graves' disease (GD) is the most common cause of hyperthyroidism, and its primary extrathyroidal manifestation is Thyroid Eye Disease (TED), an autoimmune orbitopathy that significantly impairs quality of life. Early detection of TED is crucial for timely intervention, but access to specialist ophthalmological assessment can be limited. Telemedicine presents a potential solution for remote screening and monitoring, yet validated tools for TED are lacking. This study aimed to develop and validate a telemedicine-based pathway, incorporating a novel study-specific questionnaire (SSQ) and patient-provided photographs, to accurately screen for TED in patients with GD. Materials and Methods In this paired diagnostic accuracy study, 218 consecutive patients with GD were recruited from a tertiary care center. Participants completed a remote telemedicine assessment, which included the validated Graves' Ophthalmopathy Quality of Life (GO-QoL) questionnaire and an SSQ designed to remotely assess the Clinical Activity Score (CAS) and Gorman score for diplopia. Patients also uploaded facial photographs. All participants subsequently underwent a comprehensive, masked in-person ophthalmological examination, which served as the reference standard for diagnosing TED according to EUGOGO guidelines. The diagnostic accuracy of the telemedicine tools was evaluated using receiver operating characteristic (ROC) curve analysis, and agreement between remote and in-person scores was assessed using Bland-Altman analysis, intraclass correlation coefficients (ICC), and weighted kappa. Results Of the 218 participants, 65 (29.82%) were diagnosed with TED (TED+) and 153 (70.18%) were not (TED−). Patients in the TED+ group reported a significantly lower quality of life (mean GO-QoL score: 51.54 ± 13.51) compared to the TED− group (94.68 ± 7.66; p < 0.001). The remote assessments of diplopia (Tele-Gorman) and clinical activity (Tele-CAS) showed good-to-substantial agreement with in-person findings (ICC = 0.724 and 0.771, respectively), though a small but significant underestimation was noted at higher severity levels. The SSQ alone demonstrated excellent diagnostic performance, with an Area Under the Curve (AUC) of 0.93. At an optimal cut-off of ≥4, the SSQ achieved a sensitivity of 90.77% and a specificity of 95.42%. Incorporating photographic analysis (TeleTED score) yielded a marginally higher AUC of 0.94, with a sensitivity of 89.23% and a specificity of 98.69% at a cut-off of ≥5. Conclusions A streamlined telemedicine pathway, centered on a purpose-designed questionnaire, can triage patients with Graves' disease for TED with high discriminative accuracy. The tool demonstrates good agreement with in-person clinical grading, making it a reliable instrument for remote screening. This approach is immediately translatable into a high-sensitivity pre-screening protocol that can help preserve healthcare resources and accelerate access to specialist care for patients at genuine risk of developing thyroid eye disease.