Psychometric properties of a short version of Lee Fatigue Scale used as a generic PROM in persons with stroke or osteoarthritis: assessment using a Rasch analysis approach

Abstract

Background: Fatigue is a common symptom associated with a wide range of diseases and needs to be more thoroughly studied. To minimise patient burden and to enhance response rates in research studies, patientreported outcome measures (PROM) need to be as short as possible, without sacrificing reliability and validity. It is also important to have a generic measure that can be used for comparisons across different patient populations. Thus, the aim of this secondary analysis was to evaluate the psychometric properties of the Norwegian 5-item version of the Lee Fatigue Scale (LFS) in two distinct patient populations. Methods: The sample was obtained from two different Norwegian studies and included patients 4-6 weeks after stroke (n = 322) and patients with osteoarthritis on a waiting list for total knee arthroplasty (n = 203). Fatigue severity was rated by five items from the Norwegian version of the LFS, rating each item on a numeric rating scale from 1 to 10. Rasch analysis was used to evaluate the psychometric properties of the 5-item scale across the two patient samples. Results: Three of the five LFS items ("tired", "fatigued" and "worn out") showed acceptable internal scale validity as they met the set criterion for goodness-of-fit after removal of two items with unacceptable goodness-of-fit to the Rasch model. The 3-item LFS explained 81.6% of the variance, demonstrated acceptable unidimensionality, could separate the fatigue responses into three distinct severity groups and had no differential functioning with regard to disease group. The 3-item version of the LFS had a higher separation index and better internal consistency reliability than the 5-item version.

The Job-Search Self-Efficacy (JSSE) Scale: an Item Response Theory Investigation

“…high job search self-efficacy versus low job search self-efficacy). These psychometric properties provide useful information about the utility of a measure such as the JSSE (see Bragstad et al, 2020).…”

Key considerations to reduce or address respondent burden in patient-reported outcome (PRO) data collection

“…The study suggested that trialists should consider using the entire library, especially for trials of experimental drugs to ensure they adequately capture potential adverse events 40 . Bragstad et al also highlighted that one challenge with using different short versions of a measure may be the difficulty of comparing the scores of the various versions with each other 41 . Finally, it was noted that while brevity should be a consideration, it should not outweigh the need to assess outcomes that participants consider important 23 .…”

“… When selecting PROMs, the characteristics of candidate measures should be considered as these may contribute to potential respondent burden. Various issues as highlighted in the previous sections need to be carefully considered 4 , 22 – 24 , 27 – 30 , 32 – 34 , 37 , 41 – 45 . The delivery of PROMs may influence respondent burden.…”

“…Despite evidence that the length of the PRO questionnaire may not lead to a burden or affect compliance rates, some authors suggest keeping PRO data collection as brief as possible, without compromising on reliability and validity, especially when patients are very ill or when their condition causes fatigue, lack of energy or tiredness [41][42][43] . Patient's health status and functional disability should be considered during the PROM selection process to ensure that length and content are tailored to limit response burden 44,45 .…”

“…Patient's health status and functional disability should be considered during the PROM selection process to ensure that length and content are tailored to limit response burden 44,45 . Several studies focused on the development of short forms of existing measures and found that this enhanced the feasibility of collecting PROs on a regular basis and potentially could reduce respondent burden 23,36,41,[46][47][48][49][50][51][52][53][54][55] . The use of statistical methods such as factor analysis and Rasch can assist with item reduction during the validation of PROMs.…”