Abstract ID: 26-391
Ask the patient: Do Reported Clinical Improvements Reflect Meaningful Patient Benefit in Thyroid Eye Disease? A Systematic Review of Outcome Heterogeneity in RCTs
Author: Yegee Han Base Hospital / Institution: Imperial College London
Presentation Type: ePoster Presentation
Purpose
To characterise reporting of patient-reported outcomes (PROs), clinician-reported outcomes (CROs), and surrogate outcomes (SOs) in thyroid eye disease (TED) randomised controlled trials (RCTs); to evaluate whether changes in CROs and SOs translate to patient-reported benefit.
Methods
4 databases were searched (November 2025) for RCTs of any intervention in adult TED reporting ≥1 PRO. Screening, extraction, and RoB 2 assessment followed standard methods. Outcomes were mapped by instrument, scale, unit of analysis, responder threshold, population, and timepoint.
Results
63 RCTs reported 48 distinct outcomes (24 CROs, 8 PROs, 14 SOs, 2 composite), median 8 per study. Most common: proptosis (94%), disease activity (78%), diplopia (76%), overall response (70%), palpebral aperture (67%), TRAb (51%), and quality of life (49%). Nominally identical outcomes were frequently operationalised differently, varying in instruments, definitions, thresholds, reporting formats, populations, and timepoints. 34 distinct post-baseline timepoints were used: proptosis alone at 29, with no single timepoint shared by >30 RCTs. Selective reporting by arm, subgroup, visit, or responder status was common. Where data were comparable, treatment effects tended to move in the same direction across PRO and CRO domains, but few RCTs were sufficiently similar in intervention, outcome, and timepoint to permit formal cross-domain statistical testing.
Conclusion
TED RCTs ask whether treatment worked without first agreeing what ‘work’ means or who decides so. They report apples and oranges: shared outcome labels can conceal meaningful differences in construct, measurement, threshold, population, and timepoint. The evidence base is too heterogeneous to reliably determine whether clinical/surrogate improvements translate into meaningful patient benefit. International consensus on a TED core outcome set is needed, incorporating routine GO-QoL assessment, standardised diplopia grading, prespecified timepoints, transparent responder definitions, and interpretation against minimal clinically important differences.
Additional Authors
| First name | Last name | Base Hospital / Institution |
|---|---|---|
| Mohammud | Musleh | Imperial College Healthcare NHS Trust |
| Kleopatra | Alexiadou | Imperial College Healthcare NHS Trust |
| Vickie | Lee | Imperial College Healthcare NHS Trust |