ONE UNIFORM RASCH SCORING FOR ALL LANGUAGE VERSIONS OF CATQUEST-9SF TO BE USED BY THE EUROPEAN REGISTRY OF QUALITY OUTCOMES FOR CATARACT AND REFRACTIVE SURGERY (EUREQUO)

Published 2022 - 26th ESCRS Winter Meeting

Reference: FP70 | Type: Abstracts | DOI: 10.82333/1twz-dh65

Authors: Mats Lundström* ¹ , Konrad Pesudovs ²

¹Department of Clinical Sciences, Ophthalmology,Lund University,Karlskrona,Sweden, ²School of Optometry and vision sciences,University of New South Wales,Adeleide,Australia;Department of Optometry and Vision Sciences,University of New South Wales,Adeleide,Australia

Purpose

To calculate one uniform Rasch scoring based on aggregated raw scores from six language versions of Catquest-9SF and to include the uniform Rasch scoring into the EUREQUO system.

Setting

This study uses data collected previously in six clinics located in six different countries: Sweden, the Netherlands, France, Germany, Italy, and Denmark.

Methods

Patient-reported data was collected by use of Catquest-9SF before cataract surgery. The same procedure was performed with six different language versions of Catquest-9SF in six different countries. Each one of these studies has been published in peer review journals. In the present study we included the raw scores from the six studies into one single file. A Rasch analysis was performed on the aggregated data. The Rasch scoring obtained from the aggregated data was used as a uniform scoring. The uniform scoring will be included in the EUREQUO system as a conversion algorithm from raw scoring and for future use by any clinic collecting Catquest-9SF data irrespective of language.

Results

Altogether 1338 completed preoperative questionnaires were included in the final Rasch analysis. The data fit the Rasch model. The mean age of patients was 74 and 59% were females. Item B (satisfaction with vision) was the most difficult item and the mean item scoring varied in the six language specific analyses between -1.65 and -2.86 logits. In the uniform scoring item B got a score of – 2.25 logits. For the scoring conversion algorithm item B will score -4.83, -2.37, -0.67 and 1.75 corresponding to the raw score 1, 2, 3 and 4, respectively, for the four response options.

Conclusion

A uniform Rasch scoring for different language versions of Catquest-9SF is presented. The advantage of such a scoring is that surgeons don’t need to do a new Rasch analysis for new studies. Another advantage is the possibility to compare patient-reported outcomes in different cohorts irrespective of language. A disadvantage is that the uniform scoring is not as exact as a new cohort specific Rasch scoring would be.