Which self-report measure

							Which self-report measure?

      Lucy Ziegler, Dan Stark
    Strand 1 Scientific Session
         September 2008



        www.COMPASScollaborative.com
           The problem

Systematic review of 177 papers reported
                       57
     different self-report measures.




          www.COMPASScollaborative.com
Having 57 varieties of pickles and
chutneys is not usually a problem




         www.COMPASScollaborative.com
  But 57 varieties of self-report
measures is not such a good thing.




         www.COMPASScollaborative.com
           Why is it a problem?
• No consensus on which measure works for which
  population - especially palliative care.

• Difficult to compare cohorts and hinders meta-analysis

• Potential to choose poorly validated or inappropriate
  measure

• Grant applications - perceived lack of scientific rigour.



                 www.COMPASScollaborative.com
        A possible solution?
• Systematic review generated a list of
  potentially useful measures
• Each potential measure to be subjected to
  quality appraisal using QUADAS
  framework
• Evidence Report is prepared for each
  measure
• Report presented to Delphi panel

            www.COMPASScollaborative.com
         Delphi background




• Developed in 1953 to explore potential
  bombing strategies Soviet military leaders
  might use in the event of an atomic war.
• Entered academic mainstream in 1960’s
            www.COMPASScollaborative.com
         Delphi definition

      A method of structuring a group
communication process to allow a group of
individuals to deal with a complex problem
              (Linstone 1975)




          www.COMPASScollaborative.com
  Aim of Strand 1 Delphi Study

• To determine most appropriate measure
  for longitudinal study of distress in cancer
  and supportive palliative care




             www.COMPASScollaborative.com
   Delphi Process
               Evidence


Round                             Round
 One                               Two

              Consensus




             Focus Group
             Discussions



   www.COMPASScollaborative.com
          Different approaches
• Postal Questionnaire

• Web based

• Email

• Focus group


              www.COMPASScollaborative.com
Participant stakeholder groups
   Stakeholder Group                     Number
   Clinical academics                    6
   (oncology/psychiatry/ primary
   care)
   Psychologists                         6
   PAMs and Nurses                       6 -12
   Service users                         6
   Researchers / statisticians           6


          www.COMPASScollaborative.com
www.COMPASScollaborative.com
   Broad Question Categories
• Measure development process

• Psychometric properties

• Relevance to definition of distress

• Feasibility

                www.COMPASScollaborative.com
Advantages                           Disadvantages

• Unites empirical                    • Relies on the
  evidence and expert                   commitment of panel
  opinion                               members to remain
                                        involved throughout
                                        entire process.
• Limits biases

• Systematic


               www.COMPASScollaborative.com
       Areas For Discussion
• Which methodological approach to use.
   (web based / email / focus groups)

• Recruitment and constituency of panel
  members.

• Choice of measures for Delphi – do we
  need to consult on this and how?

            www.COMPASScollaborative.com
Measuring distress is not easy




        www.COMPASScollaborative.com
Least distressed (and fastest) man
    in the world, Usain Bolt




         www.COMPASScollaborative.com