Validation of Catquest-9SF in Danish: developing a revised form of the Catquest-9SF - the Danish Catquest-7SF.

Abstract

The primary goal was to validate a Danish translated version of the Catquest-9SF by Rasch analysis. The secondary goal was to investigate whether preoperative Catquest-9SF scores, best-corrected visual acuity, comorbidity, gender, age or corneal astigmatism could predict improvements in subjective outcome. In a prospective trial, 250 patients eligible for cataract surgery were included. Patients filled out the translated Catquest-9SF questionnaire before surgery and again 3months after surgery. Both preoperative and postoperative questionnaires were included in the Rasch analysis. A multiple reverse stepwise regression model was used to investigate the correlation between preoperative measurements and subjective improvement. The preliminary Rasch analysis showed misfit of items 4 and 6. These items were removed, and the remaining seven items demonstrated a measurement precision of 2.78, a person reliability coefficient of 0.89, ordered response categories, infit of 0.69-1.22, outfit of 0.73-1.14, observed raw variance explained by measures of 70.4% and an eigenvalue of 1.7. Item 7 showed a mild DIF for gender (0.54 logits), and person mean Rasch score targeting was -1.69 logits. Preoperative Catquest score was the only parameter with a significant correlation to a gain in subjective outcome (p<0.001). A preoperative Catquest-9SF score of 0.5 carried a 95% likelihood of an increase in subjective outcome. The Danish version of the Catquest-9SF fit the Rasch model. Only preoperative Catquest-9SF score was correlated to subjective improvement, and a cut-off value of 0.5 predicted an improvement in subjective outcome with 95% probability.