Psychometric performance of the Functional Assessment of Chronic Illness Therapy (FACIT) Fatigue questionnaire among adults with paroxysmal nocturnal hemoglobinuria
摘要
Two Phase 3 clinical trials were conducted to investigate the efficacy and safety of iptacopan, a novel treatment for paroxysmal nocturnal hemoglobinuria (PNH), a rare acquired hemolytic disorder whose most prevalent symptom is fatigue. The FACIT-Fatigue questionnaire version 4 was administered to participants to document change in patient-reported fatigue after receiving treatment with iptacopan. The aim of this study is to further evaluate the psychometric properties of the FACIT-Fatigue as a measure of fatigue among adults with PNH.
MethodologyPost-hoc analyses were conducted using data from the two clinical trials at Baseline and Days 42, 126, 140, and 168. After description of its item properties (item distribution and item scores over time), the scores produced by the FACIT-Fatigue were evaluated to examine reliability (internal consistency and stability), validity (construct validity, known-groups, and responsiveness to change) and dimensionality. Meaningful within-patient change estimates were also developed.
ResultsAcross timepoints, participants endorsed the full range of the FACIT-Fatigue’s response options, and there was monotonicity of responses showing decreasing fatigue severity over time. Dimensionality analyses supported the use of the FACIT-Fatigue total score for measurement of fatigue. The total score demonstrated very high internal consistency reliability (Cronbach’s α ≥ 0.90) and good to excellent stability (test-retest intra-class correlation ≥ 0.89), correlated with other assessment scores as expected, and was able to distinguish between clinically distinct known groups. Anchor-based analysis suggested meaningful within-patient change for the FACIT-Fatigue total score (possible score range = 0–52) to be in the range of 7.5–9.5 points.
ConclusionsThe psychometric results reported here provide evidence suggesting that the scores produced by the FACIT-Fatigue are reliable and valid, and thus may be considered fit for purpose in measuring fatigue among adults with PNH. While unidimensional confirmatory factor analysis did not entirely reflecting a unidimensional model, use of a bifactor model did provide evidence of essential unidimensionality, supporting the conclusion that FACIT-Fatigue total score may be appropriate to use in evaluating treatment efficacy in this target patient population.