Early Intervention: An Outcomes Based Evaluation of Disparity in Access

1
Early Intervention An Outcomes Based
Evaluation of Disparity in Access

• Taletha M. Derrington, M.A.
• and Beppie J. Shapiro, Ph.D.
• Center on Disability Studies, College of
  Education,
• University of Hawaii
• www.seek.hawaii.edu,
• taletha_at_hawaii.edu, beppie_at_hawaii.edu,

2
Definitions

• Early Intervention Part C of IDEA, a mandated
  system of services for babies under age 3 with
  special needs (EI)
• Child find Efforts to ensure that babies with
  special needs are identified and referred to
  early intervention

3
Context

• Required Child Find function
• Community programs
• No history of evaluation

4
Context

• Infant Toddler Development Programs
• Delays in ?2 domains
• Public Health Nursing Sections
• Medical condition or single delay
• Service areas
• Geographically defined for rural areas
• Parental choice for urban areas (2/3 state
  population)

5
Why Did We Study Disparity?

• National focus on disparities in health care
• - Minority ethnicity
• - Low income
• - Recent immigrants
• - Limited English proficiency
• Homelessness
• Uninsured

6
What Demographics Predict Disparity?

• Ethnicity vs. SES

7
Unfortunate Coincidence

• Family demographics predict child delays
• Same family demographics predict less access to
  services

8
Processes Studied for Equity in Access
9
Demographics Studied for Equity in Access

• Enrollment

• Referral

• Low-Income
• Uninsured
• Immigrant
• Limited English proficiency
• Military
• Homeless

10
Metric for Equity in Access

• Ideally compare served with in population
  (prevalence)
• Problem prevalence either unknown or based on
  served
• Assume prevalence of EI eligible conditions
  evenly spread across all sub-populations

referred or enrolled in population
11
How we measured prevalence

• Census is best population wide data
• But census does not give statistics for children
  aged 0 3
• So we had to estimate statistics for children 0
  3 from Census statistics for children aged 0 18
  or 0-5

12
Example

• 45,412 children aged 0 3/ 295,767 aged birth to
  18 .15 or 15
• If census reports 1000 children 0 18 are poor,
  we calculate 1000 X .15 150 children 0 3 are
  poor.
• Note new assumption same among poor as among
  total population
• Expect 15 of babies referred to EI to be poor.

13
Data Sources

• Intake records at EI programs (1997)
• 4 ITDPs
• 2 PHNs
• Study-specific questions added to intake
  (1996-97)
• 6 ITDPs
• 5 PHNs
• State information referral line
• Statewide EI management information system (1997)

14
Sample Sizes - Referral
15
Sample Sizes - Enrollment
16
Data Analysis

• Determine if observed and population s differ
  using chi squared
• If so, calculate the effect size using Relative
  Risk

17
And We Found
18
Income/Public Insurance
Referral
Enrollment
Public Insurance
Poor
Public Insurance
19
Uninsured Children
Referral
Enrollment
20
Immigrants
Referral
21
Limited English Proficiency
Referral
Enrollment
22
Children in Military Families
Referral
Enrollment
23
Where Do We Go From Here?

• Limitations
• 1997 data same in 2005?
• Estimations for population comparison data

24
Uninsured Children

• 56 less likely to be referred
• 66 less likely to be enrolled
• Disparity may be over-estimated
• Still a cause for concern

25
Limited English Proficiency

• Self-report a limitation for both study and
  population figures
• Equity in referral
• Disparity in enrollment possible for families who
  speak only some English
• Need for interpreter not recognized by program
  staff?
• What happens between referral enrollment?

26
Children in Military Families

• Equity in referral
• Disparity in enrollment
• Coordination with military Exceptional Family
  Member Program
• What happens between referral enrollment?

27
Homelessness

• How can you study this without turning away needy
  families due to stigma?

Data
Privacy
28
Further Study

• Multiple risk factors
• Increased risk, over-representation, or
  over-referral?

29
Group Discussion

• How can we address demographically based access
  barriers?
• Uninsured
• Limited English Proficiency
• Military dependents
• What can we do to address difficult-to-study
  demographics?
• How can or should we use data collected several
  years before its publication?

30
MAHALO!

• Please complete an evaluation
• for this session

31
Contact Reference

• Taletha M. Derrington, M.A. and Beppie J.
  Shapiro, Ph.D.
• Center on Disability Studies, College of
  Education,
• University of Hawaii
• www.seek.hawaii.edu,
• taletha_at_hawaii.edu, beppie_at_hawaii.edu,
• Shapiro, B. Derrington, T. (2004). Equity and
  Disparity in Access to Services An
  Outcomes-Based Evaluation of Early Intervention
  Child Find. Topics in Early Childhood Special
  Education, 24(4), 199-212.