Multiple Sclerosis vision questionnaire (MSVQ-7): Reliability, validity, precision and discrimination

Visual dysfunction is common in people with Multiple Sclerosis (pwMS), associated with a variety of visual symptoms. Capturing the patient experience of these complex patterns of visual pathology is challenging. A valid and reliable patient reported measure, capable of detecting clinically significant change, would have considerable research and clinical benefits. We examined the properties of the MS Vision Questionnaire (MSVQ-7) in a large MS population.

Data were collected from participants in the UK-wide Trajectories of Outcome in Neurological Conditions-MS (TONiC-MS) study: MS subtype and Expanded Disability Status Scale (EDSS) band from the clinical team, as well as serial packs including the MSVQ-7 and questionnaires on depression, anxiety and stigma. A calibration sample of 1000 pwMS contributing several years of follow-up were split into training and validation samples for a Confirmatory Factor Analysis and Rasch analysis. The Minimal Detectable Change (MDC) was computed as well as the Minimal Clinically Important Change (MIC), by an anchor-based method, for different MS subtypes.

The MSVQ-7 is unidimensional and can be fit to the Rasch model with a solution discarding 3% of variance. Providing all 7 items are answered, the total can be converted to an interval-level metric for calculation of change scores and other parametric analyses. The % of missing values did not exceed 1.7%. Among 5478 pwMS, 80% reported visual problems. MSVQ-7 scores were categorised as mild for 36.1%, moderate for 33.6% and severe for 10.3%, and varied by MS subtype. In the follow-up sample of 2227 pwMS, 42.5% changed MSVQ-7 category between baseline and first follow-up (mean 22.6 months). The MIC exceeded the MDC so clinically significant change exceeds measurement error. While MDC was identical for relapsing and progressive MS, MIC varied by MS subtype, with smaller MIC in relapsing MS. Over one-quarter of the follow-up sample reported a clinically significant change in MSVQ-7: 12.2% improved and 13.5% deteriorated. For pwMS recruited within 2 years of diagnosis, 17.3% reported significant change on follow-up, all improving. MSVQ-7 scores showed strong associations with anxiety, depression and stigma (effect sizes>0.8). Duration, EDSS band and MS subtype all had effect sizes 0.2-0.49. A multinomial logistic regression exploring vision disturbance and depression, adjusted for age, gender, MS subtype, duration and disability, showed vision is the strongest significant predictor of depression. Each unit increase in interval MSVQ-7 increases risk by 10% of 'possible' and by 17% of 'probable' depression.

The MSVQ-7 is a brief self-report measure of visual problems for pwMS. It can easily be converted to interval-level measurement for change scores or power calculations and has good precision and discrimination. Visual problems were reported by 80% of pwMS and changed over time, evidencing the need for regular monitoring. MIC varied by MS subtype, indicating that perception of impact changes over the disease course. Visual dysfunction significantly affects depression risk and perceived stigma, highlighting the importance of routine assessment of visual problems in comprehensive care. The MSVQ-7 has strong psychometric properties for adoption as a measure for vision in clinical and research settings.