Validity of patient-derived disability and clinical data in multiple sclerosis

Patient-derived historical data are widely employed to make fundamental management decisions in multiple sclerosis, although the validity of the information provided is unclear. The objectives of this study were to determine validity of patient-derived historical data and to describe the utility of a locally relevant, patient-administered questionnaire designed to ascertain current disability and other important disease milestones. A well-described cohort of 99 patients was identified for whom comparable, detailed, prospective longitudinal clinician-derived data were available. Patient-derived data were collected by completion of a standardized questionnaire or telephone interview for comparison. Reliability analysis for current Expanded Disability Status Scale (EDSS) demonstrated an intraclass correlation coefficient of 0.79 between questionnaire and clinician-derived data in 79 patients, with complete agreement in 75.9%. Intraclass correlation coefficient for year of disease onset, diagnosis and onset of secondary progression was 0.86, 0.91 and 0.78, respectively. Time to EDSS >4.0, 6.0 and 8.0 all had an intraclass correlation coefficient of >0.9. Less robust agreement was observed for current disease course (Kappa coefficient 0.71), initial relapse rate (intraclass correlation coefficient 0.37) and clinical features at disease onset (Kappa 0.25). We conclude that self-reported questionnaires can provide reliable current and retrospective data on time-to-disability milestones with high levels of correlation observed for some additional elements, supporting the use of selected components of patient-derived data in clinical practice and for epidemiological studies.