The Michigan Retinal Degeneration Questionnaire: A Patient-Reported Outcome Instrument for Inherited Retinal Degenerations

Abstract

Purpose: To create a psychometrically validated patient-reported outcome measure for inherited retinal degenerations.

Design: Qualitative and quantitative patient-reported outcome (PROs) questionnaire development using item response theory validation.

Methods: One hundred twenty-eight patients with a diagnosis of an inherited retinal degeneration at the Kellogg Eye Center (University of Michigan) were recruited and administered a 166-item questionnaire comprising 7 expert-defined domains. The questionnaire was re-administered 4-16 days later to a subset of 25 participants to assess test-retest variability. Graded response models were fit by Cai's Metropolis-Hastings Robbins-Monro algorithm using the R (version 3.6.3) package mirt. Model data were fit to assess questionnaire dimensionality, to estimate item information, and to score participants. Poorly functioning items were removed, and the model was refit to create the final questionnaire.

Results: The psychometrically validated PROs measure was reduced to a 59-item questionnaire measuring 7 unidimesnional domains: central vision, color vision, contrast sensitivity, scotopic function, photopic peripheral vision, mesopic peripheral vision, and photosensitivity. A total of 39 items were removed because of poor factor loading, low item information, poor person-ability differentiation, or high item-level interdependence. This novel questionnaire produces a reliable domain score for person ability that does not show significant test-retest variability across repeated administration.

Conclusions: The final PRO questionnaire, known as the Michigan Retinal Degeneration Questionnaire, is psychometrically validated and available for use in the evaluation of patients with inherited retinal degenerations.

Figure 1:

Person-item Map for Each Domain