Psychometric properties of the revised Piper Fatigue Scale in Dutch cancer patients were satisfactory

Abstract

Objective

To examine the psychometric properties of the revised Piper Fatigue Scale (PFS) in Dutch cancer patients.

Study Design and Setting

Participants were 64 patients with lung (any stage, nonsurgery) and breast (any stage) cancer, selected for curative high-dose radiotherapy (≥50 Gy). Prior to radiotherapy, patients completed the revised PFS (translated into Dutch), the Multidimensional Fatigue Inventory (MFI), and the Rotterdam Symptom Checklist (RSCL). Reliability and construct and criterion validity of the PFS were investigated.

Results

Overall, the structure of the Dutch revised PFS appeared appropriate, with corrected item–subscale correlations being higher than the correlations of the same item with the three other subscales; furthermore, internal consistency was excellent (Cronbach's α ≥0.90). Lung cancer patients were significantly more tired then breast cancer patients, supporting construct validity. Criterion validity was also satisfactory, with highest correlations of PFS total fatigue with the MFI subscale general fatigue (0.84) and RSCL overall quality of life score (0.74). Internal consistency was similar in Dutch and U.S. breast cancer patients; fatigue scores were lower on several PFS subscales in the Dutch study population.

Conclusion

Psychometric properties of the Dutch version of the revised PFS, as tested in cancer patients prior to radiotherapy, were satisfactory.

Introduction

Fatigue can be considered as a subjective experience that affects every person from time to time. For healthy people, it can be considered as a protective, sometimes even pleasant, regular response to physical or psychological stress, which helps to maintain a healthy balance between rest and activity. For patients with specific diseases, however, fatigue often becomes a major symptom causing distress; for cancer patients, fatigue has been described as a major concern during treatment, both in advanced stages and after curative treatment. The numerous current definitions of fatigue are ambiguous and inconsistent, and vary across research studies [1], [2]. Most authors consider fatigue in terms of a complex subjective and multifactor construct with physical, mental, and psychological dimensions that are often associated with diminished quality of life (QoL) [2], [3], [4], [5].

Fatigue is one of the most frequently reported symptoms in cancer patients [4], [6], [7], [8], [9]. Several authors have reported a high diversity of occurrence of fatigue across tumor sites [8], [10], [11]. Moreover, the prevalence of fatigue increases over the course of radiation therapy [8], [10], [12], [13], [14], [15]. Lung cancer patients undergoing radiotherapy had the highest pre- and post-treatment levels of fatigue relative to patients of other tumor types [8]. A previous study [16] showed that both patients and oncologists reported the presence of fatigue in 75% of the patients; however, patients and oncologists disagreed on its importance—61% of patients reported that fatigue affected their lives more than pain, but only 37% of oncologists reported that this was the case.

The measurement of fatigue with questionnaires has been the subject of many studies. The instruments used to evaluate cancer-related fatigue (CRF) are either on a single dimension, mostly assessing the presence or the intensity of fatigue (e.g., by means of a visual analog scale), or multidimensional, in which case fatigue is either incorporated as one of the domains of QoL or measured in an individual multidimensional CRF instrument.

Wu and McSweeney [5] evaluated the quality of existing instruments measuring CRF and concluded that the Piper Fatigue Scale (PFS) is the first validated and best-developed multidimensional measure of cancer-related fatigue. The PFS was developed from a thorough review of the literature on conception and measurement of symptoms in general and of fatigue and pain in particular [5]. The scale assumes that subjective perception is the key to understanding how fatigue varies between patients with cancer and healthy individuals [2]. With its strong theoretical foundation, Piper's integrated fatigue model makes the PFS unique among existing fatigue measures [5].

Initially, two forms of the PFS each used 0–100 visual analog scales to assess the feelings of temporal, intensity or severity, affective, sensory, evaluative, associated symptom, and relief components of fatigue experienced at two times: (a) 6 months prior to diagnosis or treatment and (b) at present. The early version of the PFS was criticized for its lack of clarity and length, limiting its application with patients who are very ill or tired [5]. The PFS was therefore revised as a shorter questionnaire, composed of 22 items on a 10-point numerical, self-report scale, based on data from a large cross-sectional, mailed survey [17].

The revised PFS has been described as one of the most well-developed and widely used instruments for assessing CRF [5]; it is used not only in the United States [18], [19], [20] but also in Europe, Asia, and Australia [21], [22], [23], [24]. To facilitate cross-cultural comparisons between different countries and continents, especially the United States and European countries, we decided to translate the revised PFS into Dutch and to perform a validation study in a population of Dutch cancer patients in an early stage of treatment (i.e., preceding radiotherapy). Validation of the revised PFS in a Dutch patient population was considered important, because cultural differences in psychological adjustment and/or in the response to self-report questionnaires were reported in earlier studies [25], [26], [27]. So far, two non-U.S. validation studies of the revised PFS have been reported, one from France [21] and one from China [22].

The aim of the present study was to assess the psychometric properties (structure, reliability, construct validity, and criterion validity) of the revised PFS in Dutch cancer patients. In addition, we compared levels of fatigue in a Dutch and an American population of breast cancer patients. A Dutch version of the revised PFS was created for the present study. Feasibility and acceptability of the Dutch revised PFS, compared to other QoL assessment tools, are reported elsewhere [28].