Original article
Measurement of fatigue: determining minimally important clinical differences

https://doi.org/10.1016/S0895-4356(01)00469-3Get rights and content

Abstract

The purpose was to determine the minimally important clinical difference (MICD) in fatigue as measured by the Profile of Mood States, Schwartz Cancer Fatigue Scale (SCFS), General Fatigue Scale, and a 10-point single-item fatigue measure. The MICD is the smallest amount of change in a symptom (e.g., fatigue) measure that signifies an important change in that symptom. Subjects rated the degree of change in their fatigue over 2 days on a Global Rating Scale. 103 patients were enrolled on this multisite prospective repeated measures design. MICD was determined following established procedures at two time points. Statistically significant changes were observed for moderate and large changes in fatigue, but not for small changes. The scales were sensitive to increases in fatigue over time. The MICD, presented as mean change, for each scale and per item on each scale is: POMS = 5.6, per item = 1.1, SCFS = 5.0, per item = 0.8, GFS = 9.7, per item = 1.0, and the single item measure of fatigue was 2.4 points. This information may be useful in interpreting scale scores and planning studies using these measures.

Introduction

The evaluation and management of fatigue has become a major focus of oncology research in the pst few years. Interest in cancer-related fatigue has grown as the incidence (90–100%) and intensity of the symptom has been recognized 1, 2. A major limitation in clinical application of research on fatigue is the lack of information about the meaning of changes of various magnitudes on different fatigue instruments. Recent advances in the measurement of fatigue have paved the way for determining the magnitude of change in a fatigue score that constitutes a clinically important difference. The purpose of this study was to establish the minimally important clinical difference (MICD) of four self-report measures of fatigue in adults undergoing cancer treatment.

The MICD is defined as the smallest change in measures of fatigue that signifies an important difference in a patient's symptom 3, 4. Determining the MICD of scores on a structured instrument is a challenge. Is a change in self-reported levels of fatigue of one point on a 10-point scale clinically important? How does this degree of change compare to the change observed on other measures of fatigue? Translating changes in scores on specific instruments into equivalent and clinically relevant, meaningful information from a patient and clinician perspective is key to interpreting results across studies. Establishing equivalence across fatigue scales will provide valuable information for comparing results across trials, calculating sample size, and developing clinical recommendations for the management of fatigue.

The primary focus of research on MICD has been quality of life measures. Clinically important changes in quality of life scores have been examined in samples of people with chronic heart disease [3], asthma [6], chronic obstructive lung disease 5, 6, 7, and arthritis [8]. Results of these studies document ranges of change in scores on quality of life measures that correspond to moderate and large changes in physical and emotional function [3]. The approach used in many of these studies to determine whether a difference is clinically important is to ask whether the patient feels better or worse than previously over a time period when there is a presumed “known” effect due either to intervention or the passage of time [9]. Although some authors have argued that this approach is a transgression from classical psychometric theory, the method is gaining acceptance as a means to determine change and compare, or calibrate, the amount of change among different instruments 4, 10, 11.

In studies examining the efficacy of different interventions on fatigue and on the relationship of fatigue to other variables, investigators use a variety of instruments to measure fatigue, making comparisons among studies awkward or impossible [12]. Assessment of the changes in fatigue, measured in meaningful terms, is necessary for clinical interpretation of study results, and for comparison of the magnitude of benefit among treatments [13]. The purpose of this study was to determine the MICD in fatigue as measured by the following instruments: Profile of Mood States, Schwartz Cancer Fatigue Scale, General Fatigue Scale, and a 10-point single item fatigue measure.

Section snippets

Design and sample

A multisite, prospective, repeated measures design was used to make comparisons within subjects over time to identify the MICD in patients receiving. Institutional Review Board approval was obtained at each site before the study began. Clinic nurses or social workers identified consecutive eligible subjects at each site. Research associates explained the study and obtained written informed consent. Subjects completed all fatigue measures before chemotherapy (baseline) and 2 days after

Results

All scales detected an increase in fatigue from baseline to Time 2 (Table 2). The scales all performed well, demonstrating strong internal consistency reliability exceeded (α > .85), sensitivity to change over two days (P < .01), and a high completion rate (>92%). The scales all showed small, but statistically significant correlations with the GRS. Nearly 43% of patients reported that their fatigue did not change on the GRS (GRS = 0). The mean change reported by these subjects on the four

Discussion

The pattern of change observed in the fatigue scores by GRS responses raises several important substantive, conceptual, and methodologic issues. The results of the analyses that take the direction of change reported on the GRS into consideration suggest that subjects' view increases and decreases in fatigue differently. Although it makes sense that an increase in the intensity of a symptom is more likely to capture a person's attention thana decrease in the same symptom, the design of this

Limitations

There are numerous limitations to this study and methodological approach. Although these relatively stable findings are consistent with MICD reported in patients with other chronic diseases 6, 20, 21, 22, 23, 24, the issue of stability may be problematic. The key issues are that patients are looking backward in time to make their judgements about change on a single item measure, the GRS, that has simply been accepted as a reliable and valid measure of change [4]. This retrospective rating may

Conclusions

In conclusion, the results of this study may be important to both researchers and clinicians, but need replication. The four fatigue measures consistently demonstrated that changes of great than 0.8 per item indicates an important change in fatigue. However, it should be noted that changes on each scale reflect the scales' scoring per item (e.g., POMS and SCFS items range from 0–4, GFS items range from 1–10) and for the total scale (e.g., SCFS scores range from 6–36, GFS scores range from

Acknowledgements

This work was funded by a grant from the Oncology Nursing Foundation Fatigue in Research and Education Program (FIRTM). There are no financial, personal, academic or intellectual relationships or commitments that might bias this work.

References (27)

  • G.H. Guyatt et al.

    Proceedings of the international conference on the measurement of quality of life as an outcome in clinical trials

    Controlled Clin Trials

    (1991)
  • D.A. Redelmeier et al.

    Assessing the clinical importance of symptomatic improvements

    Arch Intern Med

    (1993)
  • Piper B. Measuring fatigue. In: Stromberg MF, Olsen SJ, editors. Instruments for clinical health-care research. 2nd ed....
  • Cited by (81)

    • Oncology: Symptom Management

      2016, Physician Assistant Clinics
    • A case-control, mono-center, open-label, pilot study to evaluate the feasibility of therapeutic touch in preventing radiation dermatitis in women with breast cancer receiving adjuvant radiation therapy

      2015, Complementary Therapies in Medicine
      Citation Excerpt :

      Secondary objectives were evaluated at the baseline and last of these visit. This consisted of the quality of life (EORTC QLQ C30),7–10 mood (Profile of Mood States (POMS))11–15 and fatigue (Brief Fatigue Inventory (BFI)).16–20 Patients and health care providers were not blinded to whether the patient received TT or not.

    View all citing articles on Scopus
    View full text