Is adjustment of National Eye Institute Visual Function Questionnaire scores for general health necessary in randomized trials?

Abstract

To assess whether treatment comparison of National Eye Institute Visual Function Questionnaire (NEI-VFQ) scores in a clinical trial is influenced by general health to warrant adjusting for it. Two randomized pilot trials. Patients enrolled in two randomized pilot trials of submacular surgery versus observation for choroidal neovascularization had quality of life interviews (NEI-VFQ and the Short Form-36 Health Survey) 24 months after enrollment. Information on comorbidities was collected through chart reviews. Data from 120 patients were analyzed using linear regression methods. Adjustment for comorbidities did not change the magnitude of the treatment effect on NEI-VFQ scores. However, adjustment for Short Form-36 physical and mental component summaries produced changes in the estimated treatment effect when NEI-VFQ scores were compared. Adjustment of NEI-VFQ scores for general health may be advisable. The Short Form-36 summary scores may be appropriate for this purpose.