Differences in Access to Care for Asian and White Adults

1
Differences in Access to Care for Asian and White
Adults

• Merrile Sing, Ph.D.
• September 8, 2008

2
Policy Context

• Many Asians face significant linguistic and
  cultural barriers
• 25 of Asians live in linguistically isolated
  households (Census 2000)
• 63 of Asians are immigrants (Census 2000)
• Some Asian American subgroups are at greater risk
  than non-Hispanic Whites for certain diseases,
  such as diabetes, stomach and liver cancer,
  hepatitis B, and tuberculosis

2
3
Research Objectives

• To estimate adjusted differences in access to
  care between non-Hispanic White and Asian adults
• To identify factors that have the greatest
  marginal effects on improving access to care

3
4
Previous Research

• Moy et al. (2008). Community Variation
  Disparities in Health Care Quality Between Asian
  and White Medicare Beneficiaries.
• Miltiades and Wu (2008). Factors Affecting
  Physician Visits in Chinese and Chinese Immigrant
  Samples.
• Snyder et al. (2000). Access to Medical Care
  Reported by Asians and Pacific Islanders in a
  West Coast Physician Group Association
• AHRQ (2007), National Healthcare Disparities
  Report

4
5
Study Design

• Data are from the Medical Expenditure Panel
  Survey (MEPS) Area Resource File, 2002 - 2005
• MEPS contains a nationally representative sample
  of households in the U.S. civilian,
  non-institutionalized population
• Sample includes non-Hispanic adults age 18 and
  older
• There are 3,779 Asians and 52,498 Whites
• Andersen typology of access to care is used
• Outcome variables are binary
• Usual source of care (excluding emergency room)
• At least one office visit during past year

5
6

• Access to Care

6
7
Andersen TypologyControl variables

• Access depends on
• Predisposing characteristics
• Enabling Resources
• Illness level or perceived need

7
8
Predisposing Characteristics

• Demographic
• Age, sex, marital status
• Social structure
• Education
• Acculturation
• Difficulty speaking English
• In linguistically isolated family
• Immigrant lt 5 years in U.S.
• Immigrant 5 14 years in U.S.
• Attitudes
• Overcome illness without medical professional
• More willing to take risk
• Always uses seat belt

8
9
Enabling Resources

• Family
• Income
• Insurance coverage
• Community
• Urban-rural (using Metropolitan Statistical
  Areas)
• Census Region (4)
• Active non-federal MDs/ 1,000 population (county)
• Number of Federally Qualified Health Centers
  (county)
• Percent Asian population in county

9
10
Illness/Perceived Need

• Self-rated general health
• Poor mental health (Mental Component Summary)
• Number of chronic conditions

10
11

• Methods

11
12
Estimation Methods

• Unadjusted differences in means
• Adjusted differences (multivariate logistic
  regressions)
• Marginal effects estimated by method of recycled
  predictions
• Standard errors estimated using balanced repeated
  replicates

12
13
Marginal effects onAccess to care

• Which factors have the greatest marginal effects
  on improving access to care?
• Predisposing conditions
• with and without acculturation variables
• Enabling resources
• Perceived need
• All control variables

13
14

• Unadjusted
• Differences

14
15
Access to CareAdults Age 18


() Significantly different from White at
0.05 (0.01) level or better Source MEPS 2002
2005, adults eligible for access supplement
15
16
AcculturationImmigrants



() Significantly different from White at
0.05 (0.01) level or better Source MEPS 2002
2005, Adults eligible for Access Supplement
16
17
AcculturationEnglish Language


() Significantly different from White at
0.05 (0.01) level or better Source MEPS 2002
2005, Adults eligible for Access Supplement
17
18

• Factors Associated with Access to Care

18
19
Variables associated withUsual Source of Care

• Marginal effect
• Asian - 0.039 (0.019)
• Enabling Predisposing
• Income immigrant lt 5 yrs in U.S.
• Insurance status immigrant 5 - 14 yrs
  in U.S.
• MSA Difficulty w/ English
• Census Region Asian Difficulty w/English
  
• family size
• Perceived need age
• number of chronic cond. gender
• self-rated health
  marital status
• attitudes
• Year 2004 - Year 2005 - Source MEPS 2002 - 2005

19
20
Variables associated withOffice Visit(s)

• Marginal effect
• Asian - 0.077 (0.015)
• Enabling Predisposing
• Income immigrant lt 5 yrs in U.S.
• Insurance status Difficulty w/
  English
• MSA
• Census Region education
• Active MDs/ 1000 pop. family size
• age
• Perceived need gender
• number of chronic cond. marital status
• self-rated general health attitudes
• self-rated mental health
  
• Year 2004 Source MEPS 2002 - 2005

20
21

• Estimated Marginal Effects

21
22
Marginal Effects on Access to Care

• Unadjusted Usual Source of Care
  Office Visit(s)
• White 0.811 (0.004) 0.784
  (0.003)
• Asian 0.701 (0.013) 0.630
  (0.011)
• Difference -0.110 -
  0.154
• Adjusted differences Marginal effects
  controlling for
• Usual Source of Care Office
  Visit(s)
• Predisposing (w/o acculturation) - 0.115
  - 0.143
• Predisposing (w/ acculturation) - 0.055
  - 0.102
• Enabling - 0.078 - 0.123
• Perceived need - 0.068 - 0.098
• All variables - 0.039 - 0.077

22
23
Conclusions

• Asian adults were less likely than Whites to have
  a usual source of care or an office visit, after
  controlling for predisposing and enabling
  characteristics and perceived need
• Greatest Marginal
  Effects on Access to Care
• Predisposing Enabling Perceived
• w/ acculturation
  Need
• Usual Source of Care v
• Office Visit v

23
24
Policy Relevance

• Findings suggest areas to focus on for improving
  access to care for Asian adults
• Translating general medical information and
  Medicaid applications into Asian languages may
  improve access to care for some Asians
• Educating providers about differences in culture
  and disease incidence for Asians compared with
  non-Hispanic Whites

24