Purpose: To examine the psychometric properties of the 9-item Fatigue Severity Scale (FSS) using a Rasch model application.
Methods: A convenience sample of HIV-infected adults was recruited, and a subset of the sample was assessed at 6-month intervals for 2 years. Socio-demographic, clinical, and symptom data were collected by self-report questionnaires. CD4 T-cell count and viral load measures were obtained from medical records. The Rasch analysis included 316 participants with 698 valid questionnaires.
Results: FSS item 2 did not advanced monotonically, and items 1 and 2 did not show acceptable goodness-of-fit to the Rasch model. A reduced FSS 7-item version demonstrated acceptable goodness-of-fit and explained 61.2% of the total variance in the scale. In the FSS-7 item version, no uniform Differential Item Functioning was found in relation to time of evaluation or to any of the socio-demographic or clinical variables.
Conclusion: This study demonstrated that the FSS-7 has better psychometric properties than the FSS-9 in this HIV sample and that responses to the different items are comparable over time and unrelated to socio-demographic and clinical variables.