Psychosocial difficulties in head and neck cancer the development of an evidence based measurement i

1
Psychosocial difficulties in head and neck
cancer the development of an evidence based
measurement instrument

• Lucy Ziegler
• Rob Newell

2
Justification for development of new instrument.

• Psychosocial consequences of disease and
  treatment are often extensive and profound.
• QoL measures currently used to identify and
  monitor psychosocial difficulties do not reflect
  extent and range of psychosocial issues that
  exist.

• Existing measures are based on professional
  opinion and patient consultation and ranking of
  item relevance.
• Traditional concept of QoL measurement may not be
  the best measurement technique in the head and
  neck cancer population?

3
4 phase study design

• Systematic review of psychosocial difficulties.
• Systematic review of existing measurement
  instruments
• Development of conceptual model and development
  of new measurement instrument.
• Validation study

4
Systematic review 1

• Psychosocial issues identified were related to
• Fear of recurrent disease
• Eating and communication
• Maintaining Social interaction
• Pain
• Disfigurement
• Personality
• Depression and anxiety

5
Systematic review of existing measurement
instruments

• Necessary to assess each existing instrument
  against standard criteria or format.
• No appraisal tool available.
• Developed appraisal tool to provide a standard
  format to evaluate the following elements of
  existing instruments

6
Measurement instrument appraisal criteria.
7
Developing New questionnaire

• Informed by systematic review of existing
  questionnaires
• Learned from systematic review 1 what issues the
  questionnaire should focus on.
• Format must be easy to understand and quick to
  complete.
• Reading age of 12 or less

8
Questionnaire items

• Eating (both functional and social)
• Role in family
• Employment
• Appearance
• Others perception of appearance
• Pain

• Relationships
• Fear of cancer coming back
• Meeting new people
• Support from family and friends
• anger
• Anxiety and depression

9
Questionnaire Layout
1. I worry about the cancer
coming back.
Disagree 0 1 2 3
4 5 6 7
8 9 10 Agree

10
Validation Study

• Types of validity explored
• Convergent validity
• Construct validity.
• Internal consistency
• Internal association
• Practical utility and application.

11
Method

• Inclusion criteria, must have diagnosis of head
  and neck cancer
• Exclusion criteria less than 3 months post
  diagnosis
• 210 questionnaires sent out

• Response rate of 73 achieved
• 154 returned

12
Internal consistency and internal association of
questionnaire

• Factor analysis led to the identification of 4
  sub scales of the questionnaire
• Interaction
• Appearance
• Emotional
• Functional

13
Factor analysis also led to the following
elements being discarded from the questionnaire

• Taste
• Tobacco and alcohol use,
• Loneliness
• Family support

14
Convergent validity

• Extent to which a questionnaire yields the same
  results as a questionnaire measuring similar
  phenomena.
• Hospital Anxiety and Depression Scale (HADS)
  (Zigmond and Snaith 1983 ) chosen as a well
  established measure that focuses on similar
  issues yet is not disease specific.
• High levels of convergent validity were seen
  between the anxiety subscale of HADS and Emotion
  and Appearance subscales of HNCPIQ.
• The Depression subscale of the HADS and the
  subscales of Function and Interaction showed
  equally high convergence.

15
Test-retest

• 40 respondents were sent the same questionnaire 2
  weeks after completing the first to determine the
  extent to which the questionnaire was able to
  yield consistent results under consistent
  conditions.
• Intraclass correlation coefficient .972
• (0.7 considered high correlation)

16
Feasibility

• Questionnaire must be feasible for clinical
  practice and acceptable to respondents
• 3 additional questions at end of questionnaire
  about feasibility (only for validation study)
• Was it upsetting to complete? Did it take too
  long? Was it easy to understand?
• stats

17
Results relating to feasibility

• 92 of respondents claimed they did not find the
  questionnaire upsetting to complete
• 98 thought the questionnaire did not take too
  long to complete.
• 96 thought the questionnaire was easy to
  understand.

18
Summary

• The head and neck psychosocial impact
  questionnaire has demonstrated good initial
  reliability and validity.
• Clinical applicability
• To further our understanding of the existence and
  severity of psychosocial difficulties.
• To determine need for psychosocial intervention.
• To facilitate further research study of this area.

19
Conclusion

• The study represents a different approach to
  instrument development than previously used in
  this field.
• The study challenges quality of life measurement
  as the only, and perhaps best approach for
  capturing information about psychosocial
  wellbeing in the head and neck cancer population.