Answers are grounded in this paper and cite their sources.
Background: Burdensome fatigue is ubiquitous in multiple sclerosis. Of 18 patient-reported outcome measures used in trials, we shortlisted six of the better-developed according to external criteria and compared their performance and equivalence at item, group and individual person levels. Objectives: All six instruments were administered to 740 people with MS. Instrument performance was examined using Rasch measurement theory, separately and in a joint analysis, which places all item and person estimates on the same scale. Results: Response rate was 72.7% ( n = 538). Instruments differed in content and structure. All combined satisfactory with imperfect measurement characteristics. Item-level analyses (fit, subtests, principal component analyses) and group-level analyses (error-corrected r = 0.85–0.98) implied these instruments measured the same variable. At the person level, instruments generated statistically different estimates for up to 38% of people, depending on instrument item numbers and intercorrelations. Conclusion: Despite item- and group-level analyses implying equivalence, individual-level differences imply that instrument choice matters. This raises concerns regarding our understanding of fatigue and its measurement for any given context. Results highlight the relevance of instrument item number, which influences measurement error associated with individual person estimates and, therefore, the degree to which different patient-reported outcome measures produce statistically equivalent estimates; the value of modern psychometric methods; and should interest anyone choosing, using, developing and analysing patient-reported outcome measure instruments.
No takes yet. Share an insight, caveat, or question.
Hobart et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: