Agreement and clinical alignment of the Ocular Surface Disease Index and Standard Patient Evaluation of Eye Dryness in a large dry eye clinic cohort.

OBJECTIVE: To compare correlation, agreement, and symptom-sign alignment between the Ocular Surface Disease Index (OSDI) and Standard Patient Evaluation of Eye Dryness (SPEED). DESIGN: Retrospective, cross-sectional chart review. PARTICIPANTS: A total of 3337 adults at a tertiary dry eye clinic (baseline encounters, 2019-2025). METHODS: Paired OSDI and SPEED totals (N = 3268) were assessed using Spearman correlation, Bland-Altman agreement after scaling SPEED to 0-100, and Lin's concordance correlation coefficient. Internal consistency and exported-total versus item-derived discrepancies were assessed in the item-level subset. Symptom-sign associations were evaluated using false discovery rate-adjusted Spearman correlations. A tear film homeostasis-loss composite (≥2 abnormal among osmolarity, noninvasive tear breakup time, fluorescein staining) was modelled using modified Poisson regression. RESULTS: OSDI and SPEED were moderately correlated (ρ = 0.53; P < .001) but showed poor agreement after rescaling, with a mean bias of 19.6 points, wide 95% limits of agreement (-21.3 to 60.5), and low concordance (concordance correlation coefficient = 0.38). Internal consistency was high (OSDI α = 0.89; SPEED α = 0.87), and large scoring discrepancies were uncommon (OSDI >5 points, 1.1%; SPEED >5 points, 0.4%). Objective sign correlations were weak (|ρ| ≤ 0.18). Greater OSDI was associated with homeostasis loss (adjusted prevalence ratio 1.03 per 10 points; P = .001), whereas SPEED was not (adjusted prevalence ratio 1.00 per 5 points; P = .86). CONCLUSIONS: Moderate correlation between OSDI and SPEED masked clinically relevant individual-level discordance after rescaling, limiting cross-instrument substitution. Both questionnaires showed weak correlations with objective signs, supporting multidomain assessment in dry eye disease.