Validation of the Catquest-9SF Questionnaire in Greek Language

Primary outcomes

1. Rasch Analysis: Response category ordering

   Time frame: 3 months

   Response category ordering is evaluated graphically with category probability curves by observing whether the four response categories - omitting 'Cannot decide' as missing - (three category thresholds) are utilized in ascending order.

2. Rasch Analysis: Item fit statistics [infit and outfit mean square (MNSQ)]

   Time frame: 3 months

   Item fit statistics are used for the evaluation of the unidimensionality. When data fit well, the items should contribute to a single construct (visual disability) indicating the unidimensionality of the questionnaire (or the subscale). Both infit and outfit mean squares have an expected value of 1.0, with an ideal range of 0.7 to 1.3, and an acceptable range of 0.5 to 1.5) A value < 0.7 indicates too little variance and easily predicted questions, while > 1.3 indicates too much variance.

3. Rasch Analysis: Principal components analysis (PCA)

   Time frame: 3 months

   PCA is used for the evaluation of the unidimensionality. Two criteria are used; the PCA assessment is performed by comparing the amount of variance explained empirically and by the model, and the amount of variance explained by the first contrast (additional dimension). For unidimensionality, the unexplained variance in the first contrast of the residuals should be < 2.0 eigenvalue units, and ≥ 60% of variance should ideally be explained by the measure, while a percentage of ≥ 50% is acceptable.

4. Rasch Analysis: Person Separation Index (PSI)

   Time frame: 3 months

   PSI contributes to the assessment of the measurement precision of the Greek Catquest-9SF. PSI implies the questionnaire's ability to differentiate along its scale. The greater the value of person separation, the greater the questionnaire precision (acceptable person separation > 3.0).

5. Rasch Analysis: Person Reliability (PR)

   Time frame: 3 months

   PR contributes to the assessment of the measurement precision of the Greek Catquest-9SF. PR (ranging from 0 to 1) should be > 0.9 to be considered adequate, indicating a high ability to distinguish between the strata of person ability.

6. Rasch Analysis: Differential Item Functioning (DIF)

   Time frame: 3 months

   DIF indicates the differences between different subgroups of patients, since the questionnaire item calibration should be comparable across different patient groups. DIF occurs if patients from different subgroups with the same underlying true ability have an unequal probability of giving of a response to an item. DIF is assessed by evaluating age, gender, existence of ocular comorbidities, first vs second eye cataract surgery and binocular distance BSCVA. DIF of >1.0 logits is significant, DIF with values between 0.5 and 1.0 is minimal, and DIF <0.5 logits is considered absent.

7. Rasch Analysis: Τargeting

   Time frame: 3 months

   Τargeting of the questions to the patient sample, which indicates how well the items' difficulties match the person ability, is assessed to establish if the questions are appropriate for patients before and following cataract surgery. This can be assessed graphically by observing the spread along the person-item map and by the difference between the item and the person mean values. A well-targeted instrument requires the ability of the patients and the difficulty of the questions to be centered on the same mean (targeting of 0). In general, a mean difference in magnitude of ≥1 logit indicates significant mistargeting, while a difference of <1.0 logit indicates good targeting, with targeting of 0 being ideal. A positive value indicates that the items are too difficult relative to person's ability, while a negative value indicates that the items are too easy. A person-item map also visualizes the item hierarchy of difficulty, ranging from a minimum to maximum difficulty of performance.