Abstract
Purpose :
Inherited retinal dystrophy (IRD) clinical trials are in need of a validated patient-reported outcomes (PRO) measure. Through an extensive process of qualitative interviews and quantitative survey data analysis, this study presents a PRO questionnaire developed in accordance with U.S. Food and Drug Administration guidelines. This objective and psychometrically validated PRO measure is designed to detect a therapeutic effect of treatments for patients with IRDs.
Methods :
Participants with a diagnosis of cone, cone-rod, rod-cone, or macular dystrophy were recruited between December 2016 and November 2019. Open-ended, in-depth interviews (n= 21) were performed and analyzed using Atlas.ti software to draft content items. Cognitive (n= 16) and pilot interviews (n=12) were conducted to refine questions and solicit patient feedback. A 166-item questionnaire comprising 7 expert-defined domains was administered to 102 participants during a 35-50 minute telephone or in-person interview. The questionnaire was re-administered after two weeks to 24 participants to assess repeatability. The study was approved the University of Michigan Institutional Review Board.
Graded Response Models (GRM) from Item Response Theory (IRT) were fit to the data to assess questionnaire dimensionality, to estimate item information, and to score participants. Wright Maps were used to visualize item difficulty and person ability. Analysis performed in R 3.6.1 with mirt package.
Results :
The original 14 questions pertaining to central vision had two dimensions we labeled ‘central vision disability’ and ‘worry’. The two worry questions were removed from the central vision scale along with one low-information question. Estimates of central vision disability from the full and reduced questionnaires were nearly identical. Local dependence was reduced by the modifications but was not entirely gone.
The marginal reliability of the scale was 0.94. The correlation between first and second administrations was 0.95. There was no evidence of mean change between administrations. Central vision disability score was highly correlated (0.70; 95% CI: 0.58, 0.79) with objective visual acuity (better eye logMAR).
Conclusions :
The final PRO questionnaire, known as the Michigan Retinal Degeneration Questionnaire, will be available in 2020 and ready for use as an outcome measure in upcoming IRD therapeutic trials.
This is a 2020 ARVO Annual Meeting abstract.