Assessment of fatigue in the cancer patient by benbenzhou

VIEWS: 203 PAGES: 26

More Info
									Assessment of Fatigue in the
      Cancer Patient

      Andrew Bottomley Ph.D
     EORTC Quality of Life Unit
         EORTC Brussels

                 Why Assess Fatigue?

   At the research level to evaluate treatments e.g. The
    EPO story

   Design new approaches/monitor effectiveness e.g.
    Group therapy, dietary intervention

   Improve clinician’s knowledge/awareness of patients
    subjective experience better satisfaction

   Individual patient care

   Guiding the patient for appropriate individual strategies

 How to Assess Fatigue?

         The clinician ?

    The patients caregiver?

The patient with a questionnaire?

             Key Properties :-
        Selecting a Fatigue Measure
       Consistency, repeatability and reproducibility
        of the empirical measure
       Applied repeatedly to the same subject would
        yield consistent results each time

   Validity
       Refers to the extent a measure is actually
        measuring the phenomenon being examined


    Several types of reliability critical to
    establish before using a tool with cancer

   Test-retest reliability
        Demonstrates that scores should be
         stable over a short time.

        Limitations: memory and temporal effect

          Responsiveness to clinical change
      Relative to change in Karnofsky P.S, toxicity
       response to treatment. Example:
Before treatment      During treatment   After treatment

                                            (response) in
                                            fatigue levels

    Types of validity critical to establish before using a
    QL tool with cancer patients

   Content validity
        The extent to which the issue is covered - did
         the measure fully cover the issues for fatigue

   Construct validity
        What is the theory behind the measure? What
         accounts for the performance of the tool?


   Convertant validity
       Two measures with the same concepts should
        correlate highly with each other or yield similar
        results even if different instruments

   Discriminate validity
       Two instruments, while measuring the same
        construct should not correlate highly if they
        measure different concepts

   Cross-cultural validity
       A measure must be valid for a given culture and
        must have undergone translation and validation
Instrument dimensions of equivalence
 across language and cultural groups

   Content      Content is relevant
   Semantic     Meaning is the same
   Technical    Method of assessment is
   Criterion    Interpretation remains the
   Conceptual   Instrument measures the
                 same theoretical construct

Forward Trans                            Forward Trans

               Translation Coordinator

                    Accepted Final
Back Trans 1                             Back Trans 1

               Translation Coordinator


                     Pilot Testing

                      Final Report
           Failure to Select Robust
                Measures Can:
   Lead to incorrect interpretation of results,
    possibly recommending a treatment when it
    may not be appropriate (or vice-versa)

   Be ethically challenged as wasting patient’s
    time collecting data if of no value (or negative

   Lead to bias in the literature

   Encourage others to follow
Lacking a fatigue tool for cancer
         clinical trials.
 Fatigue specific                   General scales
Rhoten Fatigue Scale              Profile of Mood States
Pearson-Byars fatigue checklist

Piper Fatigue Self Report         EORTC QLQ-C30

Multidimensional Fatigue          FACT- G (with F)

The Fatigue Symptom Inventory

The Schwartz Cancer
Fatigue Scale

VAS Scales

        Selected Measures of Fatigue
                        Single item scale
Rhoten Fatigue Scale
(Rhoten 1982)

   Strengths :
       Brief and easy to complete

   Weaknesses:
       Difficult to evaluate statistical reliability
       Not as reliable as multidimensional measures
       Unidimensional
       Few translations

Pearson-Byars Fatigue Checklist
(Pearson and Byars 1955)

   Strengths :
       Brief = 13 items

   Weaknesses:
       Very dated – 1950
       Not used much with cancer patients
       Validated on USA population
       Old words “pooped” and “quite fresh”
Multidimensional Fatigue Inventory
(Smets et al 1995)
   Strengths :
       Designed for cancer patients
       Quite brief – 20 items (10 min)
       Multidimensional
       Good internal consistency (alpha 0.8)

   Weaknesses:
       Limited use in publications to date
       More work on divergent validity/test-retest needed
       Availability of translations?
Piper Fatigue Self Report
(Piper et al 1998 revised)

   Strengths :
       Multidimensional
       New 22 item version may be more suitable

   Weaknesses:
       Perhaps too much burden for patient (too long)
       Some evidence of difficulties completing an old version
       Validated on US population

The Fatigue Symptom Inventory                (Hann et al 1998)

   Strengths:
       Developed for cancer patients
       Relatively brief - 13 items
       Good evidence of validity and ability to discriminate

   Weaknesses:
       only validated in the USA
       Translations do not exist
       Poor test-retest over 2-4 weeks, unsuitable for
        monitoring change
       Some suggestion of limited acceptability in patients

The Brief Fatigue Inventory
(Medoza et al 1999)
   Strengths:
       Developed for cancer patients
       Brief - 9 items

   Weaknesses:
       USA validated only
       No translations available
       Short time frame

Categorical Linear Analogue Scale (C- LASA)                        fatigue

    General tools asking one or more questions on                    100

    fatigue (or other issues) using a 1-100 mm scale                9      0

                                                                    8      0

   Strengths:                                                      7      0

                                                                    6      0
       Simple to complete                                          5      0

       Can be sensitive: detecting small changes                   4      0

                                                                    3      0
   Weaknesses:                                                     2      0

       Difficult to score with possible errors                     1      0

       Psychometric properties difficult to assess
                                                               Lowest fatigue
       Difficulties if patients in poor physical condition
       Clinical interpretation can be difficult
            Two Selected General Measures
             of QL With Fatigue Subscales
EORTC QLQ-C30 (Aaronson et al 1993)
   Strengths:
       Brief multidimensional measure generic measure
       Translated into 38 languages
       Well established general tool
       Fatigue assessed by 3 items - good psychometric properties
         Were you tired, have you felt weak, Did you need a rest
         Categorised as Not at all (1), A little, quite a bit, Very much (4)

   Weaknesses:
       Designed for group use
       Unidimensional Fatigue subscale
            Using the EORTC QLQ-C30
Two approaches: (depending on the question)

   Undertake full evaluation with 30 items for
    overall impact on QL
       Advantages - full assessment of QL
       Disadvantages - extra time demand on patient

   Undertake screening with the Fatigue sub scale
       Advantages:      Quick, Simple and Easy to assess
       Disadvantages:   Limited QL information, guide only

     Scoring and Interpretation of the
    three Item QLQ-C30 Fatigue Scale
   Estimate the average of the items scale
   Use linear transformation to standardise raw score
                 Raw score = ( Q10+ Q12 + Q18 ) / 3
                 FA score = { (RawScore-1) /3 } x 100
 Challenge for all QL measures

 High score = high level of fatigue

 Use with reference values as a comparison

 10 point increase from reference values (mean)
   suggests a jump up in category = more detailed
                     Reference Data Norms QLQ-C30
                            Fatigue Sub Scale
                                  39                 43                   45
             40                                                                                39
Fatigue 30                      29                                                                              29
(Mean) 10
0-100          0
                       General                Cancer             Cardiac      Physical      Chronic
                      Population                                problems      disease    disease (skin
                                                                 (angina,    (arthritis,  problems,
                                                               myocardinal sciatica limp   diabetes,
                                                               infarction) impairment)     allergies)

Data from: Hjermestad et al. Using reference data on quality of life. Eur. J. Cancer 1998; 34 (9) :1381-89
FACT- Fatigue (Cella et al 1997)
   Strengths:
       Fatigue scale is brief =13 items
       Designed for cancer patients with fatigue
       Good psychometric properties FACT-F
       Available in 25 languages
       Easy to complete

   Weaknesses:
       Fatigue Subscale : limited data from studies
       Unidimensional
        Scoring and interpretation of the 13
            item FACT - Fatigue Scale
   Scoring
       Items are scored 0 (not at all) to 4 (very much)
       Total score X (13) /number of answer items then linear
        transformation (logit)
       Displayed on 0 -100 scale
       High score is good function (0= worst and 100= best)

   Interpretation - Cut off points (raw scores 0-54)
      Score of 17 and below
          a difference of 2 SD worse than the normal general population
          identify some 32% of all cancer patients
       Score of 23 and below
            just below average score for cancer patients
            Useful for separating those more fatigued than less fatigued


   Fatigue is an important issue to assess in
    cancer patients
   Many measures exist to assess fatigue but
    psychometric properties vary
   Two measures stand out as valuable and
    culturally valid: EORTC QLQ-C30 and the
   The research issue/clinical question should
    dictate the measure selected

To top