Measuring Disability in Canadian Questionnaires: A First Look

1
Measuring Disability in Canadian Questionnaires
A First Look

• Sally Kader
• Health Statistics Division
• Statistics Canada

2
Objectives

• To assess the differences in estimates of
  disability from filter (screening) questions in
  different surveys
• To examine the association between disability as
  measured by the filter questions and other
  measures of health or functional limitations

3
Surveys with disability questions

• Census
• Health Surveys Canadian Community Health Survey
  (CCHS), National Population Health Survey (NPHS)
• Post-Censal Surveys Participation and Activity
  Limitations Survey (PALS), Aboriginal Peoples
  Survey (APS)
• Labour Surveys Labour Force Survey (LFS), Survey
  of Labour and Income Dynamics (SLID), Workplace
  and Employee Survey
• Other Surveys National Longitudinal Survey of
  Children and Youth, General Social Survey (GSS)

4
Sources of data

• Health
• 2000 CCHS
• 1998 NPHS
• 1996 NPHS

• Non-health
• 2001 Census
• 1999 SLID
• 2000 GSS

5
Filter questions Census prior to 2001

• Is this person limited in the kind or amount of
  activity he/she can do because of a long-term
  physical condition, or mental condition or a
  health problem
• ...at home?
• at school or at work?
• in other activities, for example, in
  transportation to and from work, or in leisure
  time activities?
• Does this person have any long-term disabilities
  or handicaps?
• Responses
• No, not limited
• Yes, limited

6
Filter questions Census 2001

• Do you have any difficulty hearing, seeing,
  communicating, walking, climbing stairs, bending,
  learning or doing any similar activities?
• Does a physical condition or mental condition or
  health problem reduce the amount or the kind of
  activity you can do
• at home?
• at work or at school?
• in other activities, for example, transportation
  or leisure?
• Responses
• Yes, often
• Yes, sometimes
• No, never

7
Prevalence of population identified as
disabled, by survey
8
Selected health status measures

• Self-perceived health (SPH)
• NPHS and CCHS only
• A global evaluation of ones health on a 5-point
  scale.
• Response categories excellent, very good,
  good, fair or poor.
• Indicator scored as 1 if response is fair or
  poor.

9
Selected health status measures

• Health Utilities Index (HUI)
• NPHS and CCHS only
• A set of 30 questions covering 8 components
  vision, hearing, speech, mobility, dexterity,
  pain, emotion, cognition.
• Each component has 2 levels none/mild and
  moderate/severe.
• Indicator scored as 1 if response indicates
  moderate/severe impairment in any component.

10
Selected health status measures

• Activities of daily living (ADL and IADL)
• NPHS and CCHS only
• A set of 6 questions about needing help in
  preparing meals, shopping for groceries or other
  necessities, housework, heavy chores, personal
  care, moving inside house.
• Response categories Yes/No.
• Indicator scored as 1 if Yes response to any of
  the questions.

11
Prevalence under disability filter questions and
selected health status measures
12
Defining disagreement between filter questions
and selected health status measures

• False negatives
• No to all the filter questions
• A positive score on another health measure
• False positives
• Yes to at least one of the filter questions
• A zero score on health measure under
  consideration

13
Method used to adjust prevalence estimates of
disability

• Adjusted prevalence
• NPHS and CCHS only
• Exclude false positives
• Estimate prevalence based on filter questions only

14
Percentage of disagreement between filter
questions and other selected health status
measures
Data Source NPHS 1996
15
Percentage of disagreement between filter
questions and other selected health status
measures
Data Source NPHS 1998
16
Percentage of disagreement between filter
questions and other selected health status
measures
Data Source CCHS 2000
17
Summary

• As expected, the redesigned filter questions
  yield higher estimates of the prevalence of
  disability than do former filter questions or
  other measures of health or functional
  limitation.
• CCHS estimates of prevalence of disability, based
  on the redesigned filter questions, are higher
  than those obtained from other surveys

18
CCHS Higher rates

• Possible explanations
• New questions
• Question placement
• Proxy rates
• Larger percentage of respondents choosing
  sometimes
• Health context

19
Limitations

• Proxy reporting rates vary
• Method of questionnaire administration
  (in-person, telephone or self-completed)

20
Next steps

• Evaluate the possible effects of
• Health context of survey
• Proxy reporting
• Method of questionnaire administration
  (in-person, telephone or self-completed)
• Question placement in survey
• Short time span