Correlation and Level of Agreement between the Ocular Surface Disease Index and the Symptom Assessment in Dry Eye Questionnaires: A Survey-Based Study

Rodríguez-García, Alejandro; Ruiz-Lozano, Raúl E.; Bustamante-Arias, Andrés; Pantaleon-Garcia, J.; Hernández-Quíntela, Everardo; Navas, Alejandro

doi:10.1080/02713683.2023.2211249

Cited by 4 publications

(2 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, watery eyes were reported to occur by 78.4% of the patients, but only in 1.0% as the sole DEQ-5 symptom, thus making this question largely redundant. The correlation between the OSDI and SANDE was found to be similar to that reported during the SANDE questionnaire development (r = 0.64) [31] and subsequent evaluations (r = 0.53 [32]; r = 0.59 [33]). However, this translated to only 28 to 41% of the variance in the OSDI being accounted for, challenging its ability "to provide clinicians with a short, quick, and reliable measure for dry eye disease symptoms" [31], although it has been shown to have a high sensitivity and specificity to a TFOS DEWS II diagnosis of dry eye disease [25].…”

Section: Discussionsupporting

confidence: 79%

“…Normalising the questionnaire ranges allowed comparison of the reliability of the questionnaires in evaluating change. The correlations between questionnaires were again only moderate as has been found in a similar analysis between the OSDI and SANDE (r = 0.47 [31]; r = 0.63 [33]). Using data from a treatment study with clear responders and non-responders [28], the discord between the questionnaires was again evident.…”

Section: Discussionsupporting

confidence: 73%

See 1 more Smart Citation

Appropriateness of Questionnaires for the Diagnosis and Monitoring Treatment of Dry Eye Disease

Wolffsohn,

Travé-Huarte,

Craig

et al. 2024

JCM

View full text Add to dashboard Cite

Objectives: If questionnaires contributing to the diagnosis of dry eye disease are to be recommended as alternatives to existing questionnaires, they must be comparable, with similar repeatability and treatment sensitivity. Comparability was thus examined for three common dry eye questionnaires along with identifying the individual questions that most strongly predicted overall scores. Methods: Anonymised data (n = 329) collected via the Ocular Surface Disease Index (OSDI), 5-item Dry Eye Questionnaire (DEQ-5) and Symptom Assessment in Dry Eye (SANDE) questionnaires (including responses to individual questions) from consenting patients were drawn from real-world dry eye clinics/registries in the United Kingdom, Australia and New Zealand; at follow-up, normalised changes were evaluated in 54 of these patients. Treatment data were also analysed from a 6-month, randomised controlled trial assessing artificial tear supplement treatments with 43 responders and 13 non-responders to treatment identified. The questions extracted from the OSDI which form the abbreviated 6-item OSDI were also analysed. Results: The agreement between the questionnaires ranged from r = 0.577 to 0.754 (all p < 0.001). For the OSDI, three questions accounted for 89.1% of the variability in the total score. The correlation between the OSDI and OSDI-6 was r = 0.939, p < 0.001. For the DEQ-5, two questions accounted for 88.5% of the variance in the total score. Normalised treatment changes were also only moderately correlated between the questionnaires (r = 0.441 to 0.595, p < 0.01). For non-responders, variability was 7.4% with both OSDI and OSDI-6, 9.7% with DEQ-5, 12.1% with SANDE-frequency and 11.9% with SANDE-severity scale. For responders, improvement with drops was detected with a 19.1% change in OSDI, 20.2% in OSDI-6, 20.9% in DEQ-5, and 27.5%/23.6% in SANDE-frequency/severity scales. Conclusions: Existing commonly used dry eye questionnaire scores do not show high levels of correlation. The OSDI was the least variable of the questionnaires and while displaying a slightly lower treatment effect than either the DEQ or SANDE, it was more sensitive to detection of a treatment effect. The quicker-to-complete OSDI-6 exhibited essentially the same outcome as the OSDI, with similar variability and treatment sensitivity.

show abstract