Rasch analysis of the Fatigue Severity Scale in multiple sclerosis

Abstract

Background The 9-item, Fatigue Severity Scale (FSS-9) has been widely used as an outcome measure in multiple sclerosis (MS). Modern psychometric theory, in the form of the Rasch measurement model, has set new quality standards for outcome measures by appraising a broad range of measurement properties in addition to the reliability and validity emphasized by classical test theory. Objective To appraise the FSS-9 by application of the Rasch model. Method The FSS-9 was posted to patients with clinically definite MS in two centers in the United Kingdom. Analysis was based on 416 records (55% response). Results The 9-item scale failed to meet Rasch model expectations. Two items had poor discrimination across the scale, and two further items showed bias for factors such as age. Removal of these four items provided a valid 5-item Rasch scale that satisfied strict tests of unidimensionality. Conclusion Summating the nine items of the FSS-9 is invalid. Five items (FSS-5), which seem to be measuring the social impact of fatigue, provide a strictly unidimensional Rasch scale. Studies using the FSS-9 may need to be re-evaluated using the FSS-5, preferably using the Rasch transformed scores.