ORIGINAL ARTICLE
Factors Associated With Having a Usual Source of Care in an Ethnically Diverse Sample of Asian American Adults Eva Chang, PhD, MPH,* Kitty S. Chan, PhD,* and Hae-Ra Han, PhD, MSN, FAANw
Background: Despite significant population increases, how Asian Americans ethnic subgroups vary in having a usual source of care (USC) is poorly understood. Objectives: To examine how having a USC varies among Asian American ethnic subgroups (Chinese, Filipinos, Japanese, Koreans, Vietnamese, and South Asians), and the potential factors influencing variation in having a USC. Research Design: Data were from 2005 and 2009 California Health Interview Survey. Logistic regressions and pair-wise comparisons were used to compare odds of having a USC among Asian ethnic adults (18–64 y) and to examine ethnicity-specific associations with immigration-related factors (English proficiency, length of residence, and living in an ethnically concordant neighborhood) and key enabling (employment, income, insurance), predisposing (education), and need (health status) factors. Models also adjusted for other sociodemographic factors. Results: Significant differences in the magnitude of the variation and factors influencing having a USC were found across Asian subgroups. Korean and Japanese adults had 52%–69% lower adjusted odds of having a USC compared with Chinese. Among all Asian subgroups, uninsured adults had 85%–94% lower adjusted odds of having a USC. Patterns of associations with USC and key factors varied by specific Asian subgroup. Conclusions: Patterns of associations for USC varied by Asian subgroup, although uninsurance persisted significantly across all subgroups. Persistent variation and heterogenous associations suggest that targeted, ethnicity-specific policies and outreach are needed to improve having a USC for Asian American ethnic adults. Key Words: access to care, usual source of care, Asian American ethnicity, immigration From the *Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health; and wDepartment of CommunityPublic Health, Johns Hopkins School of Nursing, Baltimore, MD. Funding for this work was provided by the Agency for Healthcare Research and Quality (R36 HS021684-01). The authors thank the University of California, Los Angeles, Center for Health Policy Research for providing access to the California Health Interview Survey confidential data. The findings in this article were presented at the 2014 AcademyHealth Annual Research Meeting, San Diego, CA. The authors declare no conflict of interest. Reprints: Eva Chang, PhD, MPH, Group Health Cooperative, Group Health Research Institute, 1730 Minor Ave, Ste 1600, Seattle, WA 98101. E-mail: [email protected]. Copyright r 2014 by Lippincott Williams & Wilkins ISSN: 0025-7079/14/5209-0833
Medical Care
Volume 52, Number 9, September 2014
(Med Care 2014;52: 833–841)
A
sian Americans are the fastest growing group in the United States, increasing by 46% between 2000 and 2010.1 Much of the Asian American population growth is attributed to immigration; approximately 67% are foreignborn.2,3 Overall, Asian Americans consist of 4.8% of the general US population, and Chinese, as the largest group, make up 23% of the Asian American population, followed by Asian Indian (19%), Filipino (17%), Vietnamese (11%), Korean (10%), and Japanese (5%).1,3 As the Asian American population continues to grow in the United States, understanding how Asian ethnic subgroups may differentially interact with the health care system will be valuable for addressing potential disparities in access to health care. Access to health care is a complex concept including multiple dimensions such as utilization of health care services and the presence or absence of resources that facilitate access.4,5 Having a usual source of care (USC), or a health care provider or place to which a person usually goes to when sick or in need of medical advice, is a key facilitator of entry into the health care system and of timely use of services that are applicable to the general population. In contrast, utilization measures, like frequency of physician visits, are patient and population-specific.5 Having a USC has been shown to be associated with better health outcomes and chronic disease management, and more timely and appropriate utilization of preventive and primary care.6–8 This is especially salient for Asian Americans because several ethnic subgroups have higher risks for many leading causes of death, including cancer, heart disease, stroke, and diabetes, that benefit from having a USC.9–11 For example, diabetes prevalence and incidence among South Asians and Filipinos are higher than rates observed in other high-risk minorities (Latinos and African Americans).11 Moreover, Asian American diabetics are less likely to get eye and foot examinations compared with whites.12 Having a USC has been shown to increase utilization of these diabetes-related services, but Asian Americans are less likely to have a USC.12,13 Improving rates of having a USC among Asian Americans may increase early detection of disease and facilitate treatment of medical care. Variations in having a USC exist among Asian ethnic subgroups. Pooled data from the 2004–2006 National Health Interview Survey found that age-adjusted percentages ranged from 73% among Koreans to 88% among Japanese.14 However, previous studies found mixed associations when www.lww-medicalcare.com |
833
Medical Care
Chang et al
comparing having a USC among Asian subgroups.15–17 Furthermore, these studies included all subgroups in 1 model and used different referent subgroups, making it difficult to make direct, cross-study comparisons. In addition to the lack of clarity in relative rates, it is also unclear why Asian subgroups have varied rates of having a USC. Few studies have examined the relationships between access to care and key resources, such as education and income, by Asian ethnic subgroups.18–20 Studies examining health insurance and service utilization have found distinct ethnic heterogeneity in associations with sociodemographic factors,20,21 yet no study has examined whether this heterogeneity persists in and what factors affect having a USC. Similarly, given high percentages of immigrants in the Asian American population, it is important to better understand how immigration-related factors may influence having a USC. Low English proficiency can make finding information about accessing care difficult, whereas short length of residence may limit understanding of the health care system.19,22 In contrast, Asian Americans living in ethnically concordant neighborhoods may benefit from cultural goods and social networks that characterize these neighborhoods.23,24 Predominantly immigrant and low-income, ethnic neighborhoods have been found to have a positive influence on health care access among Latinos and African Americans.23,25,26 This study describes variation in having a usual USC among Asian American ethnic subgroups (Chinese, Filipinos, Japanese, Koreans, Vietnamese, and South Asians), and examines the potential factors influencing these differences. In addition, the study utilized restricted data from the California Health Interview Survey (CHIS) to examine the influence of living in an ethnically concordant neighborhood, a factor that has not been examined in access for Asian American populations. An understanding of ethnicity-specific factors associated with having a USC will help develop targeted policies and interventions to improve access to care for Asian American ethnic subgroups.
METHODS Data are from the 2005 and 2009 CHIS, a cross-sectional, random-digit-dial telephone survey conducted biennially and representing the noninstitutionalized population in California. CHIS 2007 was not included because of concerns related to phrasing changes before the USC question; CHIS 2009 used the original phrasing (CHIS Data Access Center, written communication, March 19, 2012). As the largest statewide survey, CHIS oversamples several of the smaller ethnic groups and conducts interviews in English, Spanish, Chinese (Mandarin and Cantonese), Korean, and Vietnamese.27,28 Comparable to other California surveys, adult response rates for 2005 and 2009 were 26.9% and 17.7%, respectively.27,29 Two survey years were pooled to ensure statistical power to assess cross-ethnicity differences. CHIS confidential data were linked to tract-level data from 2010 US Census to derive the independent variable, living in an ethnically concordant neighborhood. Each census tract was defined as a neighborhood; there were 8057 neighborhoods.
834 | www.lww-medicalcare.com
Volume 52, Number 9, September 2014
This study included 7566 adults (18–64 y) who selfidentified as Asian, with 1918 Chinese, 882 Filipinos, 467 Japanese, 1138 Koreans, 1552 Vietnamese, 740 South Asians, and 869 Other Asians. Elderly adults (65+ y) were excluded because their universal access to Medicare may disproportionately affect the relationships under study, particularly those related to insurance.
Measures The primary outcome was having a USC other than the emergency room. Respondents were coded to have a USC if they answered affirmatively to the question “Is there a place that you usually go to when you are sick or need advice about your health?” and did not specify emergency room to the question “What kind of place do you go to most often—a medical doctor’s office, a clinic or hospital clinic, an emergency room, or some other place?” The primary independent variable was Asian ethnicity: Chinese, Filipino, Japanese, Korean, Vietnamese, South Asian, or Other Asian. South Asians included individuals identified as Bangladeshi, Pakistani, Sri Lankan, and Asian Indian. Other Asians included multiethnic Asians and Asians who did not identify with the 6 categories. Following a modified Andersen health behavioral model,30,31 key variables were categorized as predisposing characteristics, enabling resources, need, and immigrationrelated factors. The key predisposing characteristic was educational attainment, categorized as less than high school, high school degree, some college, college graduate (Bachelor’s), and some graduate school. Key enabling resources were annual household income, current employment status, and current health insurance status. Household income was adjusted by household size and categorized as: 40% of Koreans and Vietnamese reported limited English proficiency. Chinese had the highest proportion living in an ethnically concordant neighborhood. The majority of South Asians was male, whereas the majority of all other groups was female. Japanese were the oldest sampled (60% were 45–64 y old), whereas Other Asians were the youngest (58.5% were 18–34 y old). Other Asians had the lowest proportion of married respondents, and Vietnamese had larger households than other subgroups.
Variation in Having a USC by Ethnic Subgroup Pair-wise comparisons showed significant variation in the odds of having a USC among Asian ethnic subgroups (Table 2). The crude odds ratios suggested that significant differences exist among subgroups. Generally, Chinese and South Asians had significantly higher crude odds of having a USC than other subgroups, whereas Koreans had significantly lower crude odds than other subgroups. Although most significant differences attenuated after adjustment, several important differences persisted or appeared. Koreans consistently had significantly lower adjusted odds of having a USC than other subgroups except for Japanese. Compared with Chinese and Filipinos, Koreans had 69% and 52% lower r
2014 Lippincott Williams & Wilkins
Asian Americans and Usual Source of Care
adjusted odds of having a USC, respectively, whereas Vietnamese, South Asians, and Other Asians had almost 2–3 times greater adjusted odds than Koreans. Japanese had 52% lower adjusted odds of having a USC than Chinese. Conversely, South Asians had 95% greater adjusted odds of having a USC than Japanese.
Ethnicity-specific Associations With Having a USC Among aggregated Asian Americans, highly educated, self-employed, insured, high English proficiency, US-born, older, female, and married adults had significantly higher odds of having a USC (Table 3). Across subpopulations, uninsured adults consistently had significantly lower odds of having a USC than adults with employment-based insurance, ranging from 85% among aggregated Asian Americans to 94% among Koreans. Conversely, health status was not found to be associated with having a USC for any subpopulation. Other than these 2 factors, the factors associated with USC differed across Asian subgroups. Among Chinese adults, education, employment, other public insurance, length of residence, living in an ethnically concordant neighborhood, age, gender, and marital status also were associated significantly with having a USC. Chinese with some college education had 68% lower odds of having a USC than those with a graduate degree, whereas those with other public insurance also had 76% lower odds. Whereas neighborhood was not significant for aggregated Asian Americans, Chinese in an ethnically concordant neighborhood had 57% lower odds of having a USC than those not in a concordant neighborhood. Self-employed and mid-tenure immigrant Chinese had significantly higher odds of having a USC than employee and US-born Chinese, respectively. Being unmarried and in the 25–34 and 45–54 age groups were associated with lower odds of having a USC, whereas female gender was associated with higher odds. Medicaid insurance, income, and age also were associated with having a USC among Filipino adults. Medicaid recipients had 75% lower odds of having a USC than those with employment-based insurance, and Filipino households with an income between 100% and 199% FPL had 58% lower odds compared with households with 300%+ FPL. Like aggregated Asian Americans, odds of having a USC progressively increased with age. Having a USC was associated with length of residence, age, marital status, and household size for Japanese adults. Recent/mid-tenure Japanese immigrants had 81% lower odds of having a USC. Respondents 35–44 years old had higher odds of having a USC than 18–34-year-olds, but no trend was observed. Being unmarried and living in household with Z5 people was associated with lower odds of having a USC. Among Korean adults, uninsurance was the only factor significantly associated with having a USC. Age and gender also were associated with having a USC for Vietnamese adults. Vietnamese female individuals had >3 times higher odds of having a USC than male individuals, whereas adults 35–44 years old had 79% lower odds than adults 18 to 24 years old. www.lww-medicalcare.com |
835
Sample size Having a usual source of care other than ER No 17.5 Yes 82.5 Predisposing characteristics Educational attainment Some graduate school or 19.7 more College degree (Bachelor’s) 34.6 Some college 18.2 High school degree 19.9 Less than high school 7.6 Enabling resources Employment status Employee 64.1 Self-employed 9.2 Unemployed 6.2 Not in labor force 20.5 Insurance status Employment -based 64.0 Other private 9.2 Medicaid (Medi-Cal) 8.1 Other public 2.7 Uninsured 16.0 Household income 300%+ FPL 60.4 200%–299% FPL 13.1 100%–199% FPL 15.5 < 100% FPL 11.0 Need Health status Excellent/Very good/Good 85.0 Fair/Poor 15.0 Immigration-related factors z English proficiency High 82.1 Limited 17.9 y Length of residence in the US US born 28.4 Recent immigrant 8.0 Mid-tenure immigrant 21.5 Long-tenure immigrant 42.1 Lives in an ethnically 8 concordant neighborhood No 94.0 Yes 6.0 Other predisposing/demographic characteristics Gender Male 47.1 Age (y) 18–24 17.6 25–34 24.5 35–44 24.8 61.4 12.3 16.6 9.8
83.7 (80.6–86.5) 88.0 (84.4–90.9) 92.8 (89.2–95.3) 80.3 (75.6–84.3) 64.1 (59.0–69.0) 96.8 (95.3–97.8) 88.9 (85.4–91.7) 16.3 (13.5–19.4) 12.0 (9.1–15.6) 7.2 (4.7–10.8) 19.7 (15.7–24.4) 35.9 (31.0–41.0) 3.2 (2.2–4.7) 11.1 (8.3–14.6) 73.1 (69.0–76.8) 97.5 (95.8–98.6) 91.9 (86.8–95.1) 59.3 (52.2–66.0) 57.0 (51.5–62.4) 97.0 (95.3–98.2) 92.2 (88.9–94.6) 26.9 (23.2–31.0) 2.5 (1.4–4.2) 8.1 (4.9–13.2) 40.7 (34.0–47.8) 43.0 (37.6–48.5) 3.0 (1.8–4.7) 7.8 (5.4–11.1)
(58.2–62.5) (11.7–14.6) (14.0–17.2) (9.5–12.8) (83.5–86.3) (13.7–16.5) (80.7–83.4) (16.6–19.3)
47.5 (42.2–52.8) 16.0 (12.7–20.0)
63.5 12.6 8.0 2.3 13.6
6.1 (4.2–8.8)
(61.7–66.3) (7.9–10.6) (7.2–9.2) (2.0–3.5) (14.4–17.8)
22.1 (17.5–27.6) 17.0 (13.5–21.1)
836 | www.lww-medicalcare.com (51.8–64.4) 67.5 (61.8–72.8) 63.8 (58.3–69.0) (5.1–12.8) 8.0 (6.0–10.4) 8.9 (6.7–11.7) (3.1–6.0) 4.0 (2.7–6.0) 11.1 (7.7–15.6) (24.4–34.6) 20.5 (16.0–25.8) 16.2 (12.0–21.5)
r
14.6 (11.3–18.6) 17.5 (13.8–22.0) 8.4 (4.6–14.8) 20.3 (14.4–27.8) 19.8 (13.5–28.1) 14.5 (10.3–20.2) 28.8 (24.6–33.4) 23.3 (19.8–27.3) 25.2 (20.8–30.3) 9.6 (6.4–14.1) 26.9 (19.8–35.4) 15.8 (11.8–20.9) 32.3 (27.3–37.8) 29.7 (24.3–35.7) 25.8 (21.7–30.3) 22.0 (18.3–26.3) 22.0 (17.8–26.9) 25.0 (21.0–29.4) 24.3 (19.2–30.4) 31.2 (26.2–36.7) 22.5 (18.8–26.7)
(16.1–19.3) (22.7–26.4) (23.2–26.5)
100 0 45.5 (41.7–49.4) 48.3 (44.0–52.7) 43.0 (37.1–49.1) 34.8 (28.7–41.4) 47.9 (42.1–53.7) 56.9 (51.9–61.6) 47.8 (42.5–53.1)
96.8 (95.1–97.9) 90.0 (81.9–94.7) 98.8 (95.8–99.6) 3.2 (2.1–4.9) 10.0 (5.3–18.1) 1.2 (0.4–4.2)
(46.6–47.6)
100 0
85.7 (82.0–88.7) 97.1 (95.1–98.2) 14.3 (11.3–18.0) 2.9 (1.8–4.9)
(7.9–18.1) 50.6 (45.5–55.7) (9.7–16.6) 1.9 (0.9–3.8) (34.9–46.8) 11.6 (7.9–16.8) (29.7–39.5) 35.9 (30.9–41.2)
(92.6–95.1) (4.9–7.4)
12.1 12.8 40.7 34.4
21.7 8.3 24.6 45.4
(26.5–30.4) (6.8–9.2) (19.7–23.4) (40.0–44.2)
(19.2–34.6) 14.5 (8.7–23.2) (7.7–18.4) 6.9 (4.7–10.0) (19.1–29.9) 21.9 (18.0–26.3) (32.7–43.0) 56.6 (49.9–63.1)
(49.0–59.7) (10.4–17.6) (10.2–16.9) (14.1–24.7)
< 0.01
< 0.01
< 0.01
< 0.01
< 0.01
< 0.01
< 0.01
26.2 12.1 24.1 37.7
54.4 13.6 13.2 18.8
< 0.01
< 0.01
< 0.01
< 0.01
Pw
Medical Care
(17.9–26.0) 34.3 (29.7–39.4) 69.6 (63.4–75.1) (5.9–11.6) 7.5 (5.4–10.4) 2.4 (1.2–4.7) (20.7–28.9) 13.6 (11.0–16.7) 6.2 (4.1–9.4) (40.7–50.2) 44.5 (39.8–49.4) 21.8 (16.6–28.0)
(56.8–65.7) 61.3 (56.4–66.1) 79.6 (73.9–84.3) 58.5 (51.7–65.1) 33.9 (28.1–40.3) 78.2 (74.2–81.7) (10.1–14.9) 15.3 (12.2–19.0) 11.1 (7.6–15.9) 14.5 (11.3–18.5) 15.0 (9.2–23.6) 8.4 (6.2–11.2) (13.0–20.9) 15.1 (12.0–18.9) 6.4 (3.8–10.5) 17.3 (12.7–23.0) 28.3 (22.5–35.1) 7.4 (5.1–10.5) (7.3–12.9) 8.2 (5.2–12.7) 3.0 (1.6–5.5) 9.7 (6.1–15.1) 22.7 (18.8–27.2) 6.1 (4.0–9.1)
(59.0–67.7) 68.1 (63.0–72.9) 84.2 (79.5–88.0) 43.1 (35.6–50.9) 53.5 (47.5–59.5) 76.0 (71.4–80.1) 61.2 (55.6–66.5) (9.2–17.1) 7.0 (4.9–10.0) 6.1 (3.8–9.8) 13.6 (9.8–18.6) 5.5 (3.6–8.5) 8.4 (6.1–11.5) 8.0 (5.6–11.2) (5.7–11.3) 7.1 (5.0–9.9) 2.5 (1.4–4.5) 3.9 (2.8–5.6) 19.9 (16.6–23.6) 3.6 (2.1–6.2) 10.2 (7.7–13.4) (1.1–4.6) 4.6 (3.0–6.9) 1.1 (0.3–3.7) 4.2 (1.7–9.8) 1.9 (1.1–3.2) 0.4 (0.1–1.0) 2.3 (1.3–3.7) (11.1–16.6) 13.2 (9.4–18.3) 6.1 (3.8–9.6) 35.3 (28.7–42.5) 19.1 (14.9–24.2) 11.5 (8.7–15.1) 18.4 (14.3–23.3)
(57.2–66.0) 74.8 (69.9–79.2) 62.5 (56.2–68.3) 46.3 (39.3–53.5) 58.2 (8.5–16.1) 4.7 (3.2–6.7) 9.8 (6.9–13.8) 16.4 (12.5–21.3) 8.2 (2.8–7.7) 7.6 (4.7–12.0) 3.4 (1.8–6.3) 8.6 (4.4–16.1) 4.3 (18.1–26.0) 12.9 (9.6–17.2) 24.4 (19.2–30.5) 28.7 (23.0–35.1) 29.3
(28.1–36.4) 41.9 (36.9–47.1) 37.5 (32.0–43.4) 43.9 (37.3–50.8) 26.1 (19.6–33.8) 30.8 (25.7–36.4) 28.7 (23.8–34.1) (12.5–19.0) 24.1 (20.3–28.4) 23.0 (17.9–29.1) 15.3 (10.4–22.0) 19.5 (14.9–25.2) 9.7 (7.2–12.9) 21.2 (17.3–25.7) (15.4–21.3) 21.3 (17.1–26.1) 15.0 (10.7–20.8) 18.0 (14.4–22.3) 29.9 (25.4–34.8) 9.7 (7.1–13.1) 27.0 (21.8–32.9) (6.6–14.4) 4.9 (2.5–9.3) 2.2 (0.7–6.5) 5.7 (3.8–8.6) 18.4 (14.7–22.9) 2.3 (1.1–4.8) 7.2 (4.8–10.7)
7.8 (5.7–10.5)
61.7 11.8 4.7 21.8
869
Other Asian
(62.1–66.1) (7.9–10.5) (5.1–7.6) (19.1–21.9)
740
South Asian
32.1 15.5 18.2 9.9
1552
Vietnamese
(32.7–36.6) (16.6–19.9) (18.4–21.4) (6.3–9.1)
1138
Korean
24.4 (21.1–27.9)
467
Japanese
% (95% CI)
(18.3–21.1)
882
Filipino
12.7 (10.4–15.3) 17.8 (13.6–22.9) 12.9 (8.9–18.2) 34.4 (27.6–41.8) 20.5 (14.9–27.4) 11.5 (8.5–15.5) 20.6 (16.0–26.0) 87.3 (84.7–89.6) 82.2 (77.1–86.4) 87.1 (81.8–91.1) 65.6 (58.2–72.4) 79.5 (72.6–85.1) 88.5 (84.5–91.5) 79.4 (74.0–84.0)
1918
Chinese
(15.7–19.5) (80.5–84.3)
7566
Aggregate Asian Americans*
TABLE 1. Characteristics of Asian American Adults (18–64 y), California Health Interview Survey, 2005 and 2009
Chang et al Volume 52, Number 9, September 2014
2014 Lippincott Williams & Wilkins
r
6.3 16.0 22.7 29.7 25.3
(4.6–8.7) (12.8–19.8) (18.5–27.7) (25.5–34.2) (20.9–30.3)
12.6 28.8 27.8 22.9 7.9
(10.0–15.9) (24.0–34.0) (22.3–34.0) (18.0–28.8) (5.4–11.3)
7.9 22.7 23.1 26.6 19.7
(4.8–12.9) (17.3–29.1) (18.9–28.0) (21.9–31.9) (12.7–29.1)
3.2 15.5 19.1 28.1 34.1
(2.1–4.9) (9.6–24.1) (14.5–24.7) (23.0–33.8) (28.6–40.1)
7.0 20.1 27.0 28.1 17.7
(4.3–11.2) (15.7–25.5) (22.5–32.1) (24.1–32.5) (14.1–22.1)
2014 Lippincott Williams & Wilkins < 0.01
< 0.01
Ref. 1.47 (0.38) 0.41 (0.09)z 0.84 (0.22) 1.66 (0.39)w 0.84 (0.18)
Ref. 0.73 (0.22) 0.48 (0.15)w 0.95 (0.31) 1.43 (0.41) 1.00 (0.25)
Adjusted
Ref. 0.28 (0.07)z 0.57 (0.17) 1.13 (0.31) 0.57 (0.15)w
Crude
Ref. 0.65 (0.22) 1.29 (0.45) 1.95 (0.63)w 1.36 (0.41)
Adjusted
Ref. 2.03 (0.51)z 4.01 (0.94)z 2.02 (0.45)z
Crude
Ref. 1.97 (0.65)w 2.99 (0.96)z 2.08 (0.64)w
Adjusted
Korean [OR (SE)]
Ref. 1.97 (0.50)z 0.99 (0.24)
Crude
Ref. 1.51 (0.50) 1.06 (0.28)
Adjusted
Vietnamese [OR (SE)]
Ref. 0.50 (0.12)z
Crude
Ref. 0.70 (0.20)
Adjusted
South Asian [OR (SE)]
*Adjusted model controls for living in an ethnically concordant neighborhood, educational attainment, employment status, insurance, household income, health status, language proficiency, length of residence, age, gender, marital status, household size, and survey year. w P < 0.05. z P < 0.01. OR indicates odds ratio.
Ref. 0.65 (0.16) 0.48 (0.13)z 0.31 (0.08)z 0.61 (0.19) 0.93 (0.25) 0.65 (0.15)
Adjusted
Crude
Ref. 0.66 (0.12)w 0.98 (0.23) 0.28 (0.05)z 0.56 (0.12)z 1.11 (0.23) 0.56 (0.11)z
Crude
Japanese [OR (SE)]
Volume 52, Number 9, September 2014
Chinese Filipino Japanese Korean Vietnamese South Asian Other Asian
Filipino [OR (SE)]
Chinese [OR (SE)]
TABLE 2. Pair-wise Comparison of Asian American Ethnic Subgroups in Having a Usual Source of Care Other Than the Emergency Room, California Health Interview Survey, 2005 and 2009*
6.8 (5.1–9.1) 25.7 (21.3–30.5) 19.3 (15.7–23.5) 19.8 (15.9–24.3) 28.5 (23.1–34.5)
7.2 20.3 24.5 24.5 23.4
6.8 19.9 23.3 26.3 23.8
(5.0–10.4) (17.1–24.0) (21.6–27.7) (21.6–27.7) (19.1–28.2)
63.8 (59.3–68.1) 53.2 (47.9–58.4) 65.5 (58.9–71.4) 58.9 (51.2–66.1) 59.4 (52.5–65.9) 68.8 (62.5–74.5) 42.7 (37.7–47.8)
58.8 (56.7–60.8) (5.9–7.8) (18.2–21.7) (21.6–25.1) (24.6–28.1) (21.7–25.9)
22.5 (19.3–26.1) 20.1 (16.9–23.7) 33.3 (27.7–39.4) 17.2 (14.4–20.5) 25.9 (21.9–30.3) 14.2 (11.2–17.8) 13.6 (10.9–17.0) 13.8 (11.8–16.2) 15.1 (12.2–18.5) 26.7 (21.5–32.7) 10.7 (8.4–13.4) 14.2 (11.4–17.5) 7.8 (5.6–10.8) 5.4 (4.1–7.0)
20.3 (18.9–21.7) 12.8 (11.8–13.8)
All percentages are weighted and might not add to 100 because of rounding. *Aggregate Asian Americans include Chinese, Filipino, Japanese, Korean, Vietnamese, South Asian, and Other Asians. w The w2 test of significance among Asian ethnic subgroups. z High English proficiency responses included English only, very well/well and limited English proficiency responses included not well/poor. y Recent immigrants have been in the US for
Factors Associated With Having a Usual Source of Care in an Ethnically Diverse Sample of Asian American Adults Eva Chang, PhD, MPH,* Kitty S. Chan, PhD,* and Hae-Ra Han, PhD, MSN, FAANw
Background: Despite significant population increases, how Asian Americans ethnic subgroups vary in having a usual source of care (USC) is poorly understood. Objectives: To examine how having a USC varies among Asian American ethnic subgroups (Chinese, Filipinos, Japanese, Koreans, Vietnamese, and South Asians), and the potential factors influencing variation in having a USC. Research Design: Data were from 2005 and 2009 California Health Interview Survey. Logistic regressions and pair-wise comparisons were used to compare odds of having a USC among Asian ethnic adults (18–64 y) and to examine ethnicity-specific associations with immigration-related factors (English proficiency, length of residence, and living in an ethnically concordant neighborhood) and key enabling (employment, income, insurance), predisposing (education), and need (health status) factors. Models also adjusted for other sociodemographic factors. Results: Significant differences in the magnitude of the variation and factors influencing having a USC were found across Asian subgroups. Korean and Japanese adults had 52%–69% lower adjusted odds of having a USC compared with Chinese. Among all Asian subgroups, uninsured adults had 85%–94% lower adjusted odds of having a USC. Patterns of associations with USC and key factors varied by specific Asian subgroup. Conclusions: Patterns of associations for USC varied by Asian subgroup, although uninsurance persisted significantly across all subgroups. Persistent variation and heterogenous associations suggest that targeted, ethnicity-specific policies and outreach are needed to improve having a USC for Asian American ethnic adults. Key Words: access to care, usual source of care, Asian American ethnicity, immigration From the *Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health; and wDepartment of CommunityPublic Health, Johns Hopkins School of Nursing, Baltimore, MD. Funding for this work was provided by the Agency for Healthcare Research and Quality (R36 HS021684-01). The authors thank the University of California, Los Angeles, Center for Health Policy Research for providing access to the California Health Interview Survey confidential data. The findings in this article were presented at the 2014 AcademyHealth Annual Research Meeting, San Diego, CA. The authors declare no conflict of interest. Reprints: Eva Chang, PhD, MPH, Group Health Cooperative, Group Health Research Institute, 1730 Minor Ave, Ste 1600, Seattle, WA 98101. E-mail: [email protected]. Copyright r 2014 by Lippincott Williams & Wilkins ISSN: 0025-7079/14/5209-0833
Medical Care
Volume 52, Number 9, September 2014
(Med Care 2014;52: 833–841)
A
sian Americans are the fastest growing group in the United States, increasing by 46% between 2000 and 2010.1 Much of the Asian American population growth is attributed to immigration; approximately 67% are foreignborn.2,3 Overall, Asian Americans consist of 4.8% of the general US population, and Chinese, as the largest group, make up 23% of the Asian American population, followed by Asian Indian (19%), Filipino (17%), Vietnamese (11%), Korean (10%), and Japanese (5%).1,3 As the Asian American population continues to grow in the United States, understanding how Asian ethnic subgroups may differentially interact with the health care system will be valuable for addressing potential disparities in access to health care. Access to health care is a complex concept including multiple dimensions such as utilization of health care services and the presence or absence of resources that facilitate access.4,5 Having a usual source of care (USC), or a health care provider or place to which a person usually goes to when sick or in need of medical advice, is a key facilitator of entry into the health care system and of timely use of services that are applicable to the general population. In contrast, utilization measures, like frequency of physician visits, are patient and population-specific.5 Having a USC has been shown to be associated with better health outcomes and chronic disease management, and more timely and appropriate utilization of preventive and primary care.6–8 This is especially salient for Asian Americans because several ethnic subgroups have higher risks for many leading causes of death, including cancer, heart disease, stroke, and diabetes, that benefit from having a USC.9–11 For example, diabetes prevalence and incidence among South Asians and Filipinos are higher than rates observed in other high-risk minorities (Latinos and African Americans).11 Moreover, Asian American diabetics are less likely to get eye and foot examinations compared with whites.12 Having a USC has been shown to increase utilization of these diabetes-related services, but Asian Americans are less likely to have a USC.12,13 Improving rates of having a USC among Asian Americans may increase early detection of disease and facilitate treatment of medical care. Variations in having a USC exist among Asian ethnic subgroups. Pooled data from the 2004–2006 National Health Interview Survey found that age-adjusted percentages ranged from 73% among Koreans to 88% among Japanese.14 However, previous studies found mixed associations when www.lww-medicalcare.com |
833
Medical Care
Chang et al
comparing having a USC among Asian subgroups.15–17 Furthermore, these studies included all subgroups in 1 model and used different referent subgroups, making it difficult to make direct, cross-study comparisons. In addition to the lack of clarity in relative rates, it is also unclear why Asian subgroups have varied rates of having a USC. Few studies have examined the relationships between access to care and key resources, such as education and income, by Asian ethnic subgroups.18–20 Studies examining health insurance and service utilization have found distinct ethnic heterogeneity in associations with sociodemographic factors,20,21 yet no study has examined whether this heterogeneity persists in and what factors affect having a USC. Similarly, given high percentages of immigrants in the Asian American population, it is important to better understand how immigration-related factors may influence having a USC. Low English proficiency can make finding information about accessing care difficult, whereas short length of residence may limit understanding of the health care system.19,22 In contrast, Asian Americans living in ethnically concordant neighborhoods may benefit from cultural goods and social networks that characterize these neighborhoods.23,24 Predominantly immigrant and low-income, ethnic neighborhoods have been found to have a positive influence on health care access among Latinos and African Americans.23,25,26 This study describes variation in having a usual USC among Asian American ethnic subgroups (Chinese, Filipinos, Japanese, Koreans, Vietnamese, and South Asians), and examines the potential factors influencing these differences. In addition, the study utilized restricted data from the California Health Interview Survey (CHIS) to examine the influence of living in an ethnically concordant neighborhood, a factor that has not been examined in access for Asian American populations. An understanding of ethnicity-specific factors associated with having a USC will help develop targeted policies and interventions to improve access to care for Asian American ethnic subgroups.
METHODS Data are from the 2005 and 2009 CHIS, a cross-sectional, random-digit-dial telephone survey conducted biennially and representing the noninstitutionalized population in California. CHIS 2007 was not included because of concerns related to phrasing changes before the USC question; CHIS 2009 used the original phrasing (CHIS Data Access Center, written communication, March 19, 2012). As the largest statewide survey, CHIS oversamples several of the smaller ethnic groups and conducts interviews in English, Spanish, Chinese (Mandarin and Cantonese), Korean, and Vietnamese.27,28 Comparable to other California surveys, adult response rates for 2005 and 2009 were 26.9% and 17.7%, respectively.27,29 Two survey years were pooled to ensure statistical power to assess cross-ethnicity differences. CHIS confidential data were linked to tract-level data from 2010 US Census to derive the independent variable, living in an ethnically concordant neighborhood. Each census tract was defined as a neighborhood; there were 8057 neighborhoods.
834 | www.lww-medicalcare.com
Volume 52, Number 9, September 2014
This study included 7566 adults (18–64 y) who selfidentified as Asian, with 1918 Chinese, 882 Filipinos, 467 Japanese, 1138 Koreans, 1552 Vietnamese, 740 South Asians, and 869 Other Asians. Elderly adults (65+ y) were excluded because their universal access to Medicare may disproportionately affect the relationships under study, particularly those related to insurance.
Measures The primary outcome was having a USC other than the emergency room. Respondents were coded to have a USC if they answered affirmatively to the question “Is there a place that you usually go to when you are sick or need advice about your health?” and did not specify emergency room to the question “What kind of place do you go to most often—a medical doctor’s office, a clinic or hospital clinic, an emergency room, or some other place?” The primary independent variable was Asian ethnicity: Chinese, Filipino, Japanese, Korean, Vietnamese, South Asian, or Other Asian. South Asians included individuals identified as Bangladeshi, Pakistani, Sri Lankan, and Asian Indian. Other Asians included multiethnic Asians and Asians who did not identify with the 6 categories. Following a modified Andersen health behavioral model,30,31 key variables were categorized as predisposing characteristics, enabling resources, need, and immigrationrelated factors. The key predisposing characteristic was educational attainment, categorized as less than high school, high school degree, some college, college graduate (Bachelor’s), and some graduate school. Key enabling resources were annual household income, current employment status, and current health insurance status. Household income was adjusted by household size and categorized as: 40% of Koreans and Vietnamese reported limited English proficiency. Chinese had the highest proportion living in an ethnically concordant neighborhood. The majority of South Asians was male, whereas the majority of all other groups was female. Japanese were the oldest sampled (60% were 45–64 y old), whereas Other Asians were the youngest (58.5% were 18–34 y old). Other Asians had the lowest proportion of married respondents, and Vietnamese had larger households than other subgroups.
Variation in Having a USC by Ethnic Subgroup Pair-wise comparisons showed significant variation in the odds of having a USC among Asian ethnic subgroups (Table 2). The crude odds ratios suggested that significant differences exist among subgroups. Generally, Chinese and South Asians had significantly higher crude odds of having a USC than other subgroups, whereas Koreans had significantly lower crude odds than other subgroups. Although most significant differences attenuated after adjustment, several important differences persisted or appeared. Koreans consistently had significantly lower adjusted odds of having a USC than other subgroups except for Japanese. Compared with Chinese and Filipinos, Koreans had 69% and 52% lower r
2014 Lippincott Williams & Wilkins
Asian Americans and Usual Source of Care
adjusted odds of having a USC, respectively, whereas Vietnamese, South Asians, and Other Asians had almost 2–3 times greater adjusted odds than Koreans. Japanese had 52% lower adjusted odds of having a USC than Chinese. Conversely, South Asians had 95% greater adjusted odds of having a USC than Japanese.
Ethnicity-specific Associations With Having a USC Among aggregated Asian Americans, highly educated, self-employed, insured, high English proficiency, US-born, older, female, and married adults had significantly higher odds of having a USC (Table 3). Across subpopulations, uninsured adults consistently had significantly lower odds of having a USC than adults with employment-based insurance, ranging from 85% among aggregated Asian Americans to 94% among Koreans. Conversely, health status was not found to be associated with having a USC for any subpopulation. Other than these 2 factors, the factors associated with USC differed across Asian subgroups. Among Chinese adults, education, employment, other public insurance, length of residence, living in an ethnically concordant neighborhood, age, gender, and marital status also were associated significantly with having a USC. Chinese with some college education had 68% lower odds of having a USC than those with a graduate degree, whereas those with other public insurance also had 76% lower odds. Whereas neighborhood was not significant for aggregated Asian Americans, Chinese in an ethnically concordant neighborhood had 57% lower odds of having a USC than those not in a concordant neighborhood. Self-employed and mid-tenure immigrant Chinese had significantly higher odds of having a USC than employee and US-born Chinese, respectively. Being unmarried and in the 25–34 and 45–54 age groups were associated with lower odds of having a USC, whereas female gender was associated with higher odds. Medicaid insurance, income, and age also were associated with having a USC among Filipino adults. Medicaid recipients had 75% lower odds of having a USC than those with employment-based insurance, and Filipino households with an income between 100% and 199% FPL had 58% lower odds compared with households with 300%+ FPL. Like aggregated Asian Americans, odds of having a USC progressively increased with age. Having a USC was associated with length of residence, age, marital status, and household size for Japanese adults. Recent/mid-tenure Japanese immigrants had 81% lower odds of having a USC. Respondents 35–44 years old had higher odds of having a USC than 18–34-year-olds, but no trend was observed. Being unmarried and living in household with Z5 people was associated with lower odds of having a USC. Among Korean adults, uninsurance was the only factor significantly associated with having a USC. Age and gender also were associated with having a USC for Vietnamese adults. Vietnamese female individuals had >3 times higher odds of having a USC than male individuals, whereas adults 35–44 years old had 79% lower odds than adults 18 to 24 years old. www.lww-medicalcare.com |
835
Sample size Having a usual source of care other than ER No 17.5 Yes 82.5 Predisposing characteristics Educational attainment Some graduate school or 19.7 more College degree (Bachelor’s) 34.6 Some college 18.2 High school degree 19.9 Less than high school 7.6 Enabling resources Employment status Employee 64.1 Self-employed 9.2 Unemployed 6.2 Not in labor force 20.5 Insurance status Employment -based 64.0 Other private 9.2 Medicaid (Medi-Cal) 8.1 Other public 2.7 Uninsured 16.0 Household income 300%+ FPL 60.4 200%–299% FPL 13.1 100%–199% FPL 15.5 < 100% FPL 11.0 Need Health status Excellent/Very good/Good 85.0 Fair/Poor 15.0 Immigration-related factors z English proficiency High 82.1 Limited 17.9 y Length of residence in the US US born 28.4 Recent immigrant 8.0 Mid-tenure immigrant 21.5 Long-tenure immigrant 42.1 Lives in an ethnically 8 concordant neighborhood No 94.0 Yes 6.0 Other predisposing/demographic characteristics Gender Male 47.1 Age (y) 18–24 17.6 25–34 24.5 35–44 24.8 61.4 12.3 16.6 9.8
83.7 (80.6–86.5) 88.0 (84.4–90.9) 92.8 (89.2–95.3) 80.3 (75.6–84.3) 64.1 (59.0–69.0) 96.8 (95.3–97.8) 88.9 (85.4–91.7) 16.3 (13.5–19.4) 12.0 (9.1–15.6) 7.2 (4.7–10.8) 19.7 (15.7–24.4) 35.9 (31.0–41.0) 3.2 (2.2–4.7) 11.1 (8.3–14.6) 73.1 (69.0–76.8) 97.5 (95.8–98.6) 91.9 (86.8–95.1) 59.3 (52.2–66.0) 57.0 (51.5–62.4) 97.0 (95.3–98.2) 92.2 (88.9–94.6) 26.9 (23.2–31.0) 2.5 (1.4–4.2) 8.1 (4.9–13.2) 40.7 (34.0–47.8) 43.0 (37.6–48.5) 3.0 (1.8–4.7) 7.8 (5.4–11.1)
(58.2–62.5) (11.7–14.6) (14.0–17.2) (9.5–12.8) (83.5–86.3) (13.7–16.5) (80.7–83.4) (16.6–19.3)
47.5 (42.2–52.8) 16.0 (12.7–20.0)
63.5 12.6 8.0 2.3 13.6
6.1 (4.2–8.8)
(61.7–66.3) (7.9–10.6) (7.2–9.2) (2.0–3.5) (14.4–17.8)
22.1 (17.5–27.6) 17.0 (13.5–21.1)
836 | www.lww-medicalcare.com (51.8–64.4) 67.5 (61.8–72.8) 63.8 (58.3–69.0) (5.1–12.8) 8.0 (6.0–10.4) 8.9 (6.7–11.7) (3.1–6.0) 4.0 (2.7–6.0) 11.1 (7.7–15.6) (24.4–34.6) 20.5 (16.0–25.8) 16.2 (12.0–21.5)
r
14.6 (11.3–18.6) 17.5 (13.8–22.0) 8.4 (4.6–14.8) 20.3 (14.4–27.8) 19.8 (13.5–28.1) 14.5 (10.3–20.2) 28.8 (24.6–33.4) 23.3 (19.8–27.3) 25.2 (20.8–30.3) 9.6 (6.4–14.1) 26.9 (19.8–35.4) 15.8 (11.8–20.9) 32.3 (27.3–37.8) 29.7 (24.3–35.7) 25.8 (21.7–30.3) 22.0 (18.3–26.3) 22.0 (17.8–26.9) 25.0 (21.0–29.4) 24.3 (19.2–30.4) 31.2 (26.2–36.7) 22.5 (18.8–26.7)
(16.1–19.3) (22.7–26.4) (23.2–26.5)
100 0 45.5 (41.7–49.4) 48.3 (44.0–52.7) 43.0 (37.1–49.1) 34.8 (28.7–41.4) 47.9 (42.1–53.7) 56.9 (51.9–61.6) 47.8 (42.5–53.1)
96.8 (95.1–97.9) 90.0 (81.9–94.7) 98.8 (95.8–99.6) 3.2 (2.1–4.9) 10.0 (5.3–18.1) 1.2 (0.4–4.2)
(46.6–47.6)
100 0
85.7 (82.0–88.7) 97.1 (95.1–98.2) 14.3 (11.3–18.0) 2.9 (1.8–4.9)
(7.9–18.1) 50.6 (45.5–55.7) (9.7–16.6) 1.9 (0.9–3.8) (34.9–46.8) 11.6 (7.9–16.8) (29.7–39.5) 35.9 (30.9–41.2)
(92.6–95.1) (4.9–7.4)
12.1 12.8 40.7 34.4
21.7 8.3 24.6 45.4
(26.5–30.4) (6.8–9.2) (19.7–23.4) (40.0–44.2)
(19.2–34.6) 14.5 (8.7–23.2) (7.7–18.4) 6.9 (4.7–10.0) (19.1–29.9) 21.9 (18.0–26.3) (32.7–43.0) 56.6 (49.9–63.1)
(49.0–59.7) (10.4–17.6) (10.2–16.9) (14.1–24.7)
< 0.01
< 0.01
< 0.01
< 0.01
< 0.01
< 0.01
< 0.01
26.2 12.1 24.1 37.7
54.4 13.6 13.2 18.8
< 0.01
< 0.01
< 0.01
< 0.01
Pw
Medical Care
(17.9–26.0) 34.3 (29.7–39.4) 69.6 (63.4–75.1) (5.9–11.6) 7.5 (5.4–10.4) 2.4 (1.2–4.7) (20.7–28.9) 13.6 (11.0–16.7) 6.2 (4.1–9.4) (40.7–50.2) 44.5 (39.8–49.4) 21.8 (16.6–28.0)
(56.8–65.7) 61.3 (56.4–66.1) 79.6 (73.9–84.3) 58.5 (51.7–65.1) 33.9 (28.1–40.3) 78.2 (74.2–81.7) (10.1–14.9) 15.3 (12.2–19.0) 11.1 (7.6–15.9) 14.5 (11.3–18.5) 15.0 (9.2–23.6) 8.4 (6.2–11.2) (13.0–20.9) 15.1 (12.0–18.9) 6.4 (3.8–10.5) 17.3 (12.7–23.0) 28.3 (22.5–35.1) 7.4 (5.1–10.5) (7.3–12.9) 8.2 (5.2–12.7) 3.0 (1.6–5.5) 9.7 (6.1–15.1) 22.7 (18.8–27.2) 6.1 (4.0–9.1)
(59.0–67.7) 68.1 (63.0–72.9) 84.2 (79.5–88.0) 43.1 (35.6–50.9) 53.5 (47.5–59.5) 76.0 (71.4–80.1) 61.2 (55.6–66.5) (9.2–17.1) 7.0 (4.9–10.0) 6.1 (3.8–9.8) 13.6 (9.8–18.6) 5.5 (3.6–8.5) 8.4 (6.1–11.5) 8.0 (5.6–11.2) (5.7–11.3) 7.1 (5.0–9.9) 2.5 (1.4–4.5) 3.9 (2.8–5.6) 19.9 (16.6–23.6) 3.6 (2.1–6.2) 10.2 (7.7–13.4) (1.1–4.6) 4.6 (3.0–6.9) 1.1 (0.3–3.7) 4.2 (1.7–9.8) 1.9 (1.1–3.2) 0.4 (0.1–1.0) 2.3 (1.3–3.7) (11.1–16.6) 13.2 (9.4–18.3) 6.1 (3.8–9.6) 35.3 (28.7–42.5) 19.1 (14.9–24.2) 11.5 (8.7–15.1) 18.4 (14.3–23.3)
(57.2–66.0) 74.8 (69.9–79.2) 62.5 (56.2–68.3) 46.3 (39.3–53.5) 58.2 (8.5–16.1) 4.7 (3.2–6.7) 9.8 (6.9–13.8) 16.4 (12.5–21.3) 8.2 (2.8–7.7) 7.6 (4.7–12.0) 3.4 (1.8–6.3) 8.6 (4.4–16.1) 4.3 (18.1–26.0) 12.9 (9.6–17.2) 24.4 (19.2–30.5) 28.7 (23.0–35.1) 29.3
(28.1–36.4) 41.9 (36.9–47.1) 37.5 (32.0–43.4) 43.9 (37.3–50.8) 26.1 (19.6–33.8) 30.8 (25.7–36.4) 28.7 (23.8–34.1) (12.5–19.0) 24.1 (20.3–28.4) 23.0 (17.9–29.1) 15.3 (10.4–22.0) 19.5 (14.9–25.2) 9.7 (7.2–12.9) 21.2 (17.3–25.7) (15.4–21.3) 21.3 (17.1–26.1) 15.0 (10.7–20.8) 18.0 (14.4–22.3) 29.9 (25.4–34.8) 9.7 (7.1–13.1) 27.0 (21.8–32.9) (6.6–14.4) 4.9 (2.5–9.3) 2.2 (0.7–6.5) 5.7 (3.8–8.6) 18.4 (14.7–22.9) 2.3 (1.1–4.8) 7.2 (4.8–10.7)
7.8 (5.7–10.5)
61.7 11.8 4.7 21.8
869
Other Asian
(62.1–66.1) (7.9–10.5) (5.1–7.6) (19.1–21.9)
740
South Asian
32.1 15.5 18.2 9.9
1552
Vietnamese
(32.7–36.6) (16.6–19.9) (18.4–21.4) (6.3–9.1)
1138
Korean
24.4 (21.1–27.9)
467
Japanese
% (95% CI)
(18.3–21.1)
882
Filipino
12.7 (10.4–15.3) 17.8 (13.6–22.9) 12.9 (8.9–18.2) 34.4 (27.6–41.8) 20.5 (14.9–27.4) 11.5 (8.5–15.5) 20.6 (16.0–26.0) 87.3 (84.7–89.6) 82.2 (77.1–86.4) 87.1 (81.8–91.1) 65.6 (58.2–72.4) 79.5 (72.6–85.1) 88.5 (84.5–91.5) 79.4 (74.0–84.0)
1918
Chinese
(15.7–19.5) (80.5–84.3)
7566
Aggregate Asian Americans*
TABLE 1. Characteristics of Asian American Adults (18–64 y), California Health Interview Survey, 2005 and 2009
Chang et al Volume 52, Number 9, September 2014
2014 Lippincott Williams & Wilkins
r
6.3 16.0 22.7 29.7 25.3
(4.6–8.7) (12.8–19.8) (18.5–27.7) (25.5–34.2) (20.9–30.3)
12.6 28.8 27.8 22.9 7.9
(10.0–15.9) (24.0–34.0) (22.3–34.0) (18.0–28.8) (5.4–11.3)
7.9 22.7 23.1 26.6 19.7
(4.8–12.9) (17.3–29.1) (18.9–28.0) (21.9–31.9) (12.7–29.1)
3.2 15.5 19.1 28.1 34.1
(2.1–4.9) (9.6–24.1) (14.5–24.7) (23.0–33.8) (28.6–40.1)
7.0 20.1 27.0 28.1 17.7
(4.3–11.2) (15.7–25.5) (22.5–32.1) (24.1–32.5) (14.1–22.1)
2014 Lippincott Williams & Wilkins < 0.01
< 0.01
Ref. 1.47 (0.38) 0.41 (0.09)z 0.84 (0.22) 1.66 (0.39)w 0.84 (0.18)
Ref. 0.73 (0.22) 0.48 (0.15)w 0.95 (0.31) 1.43 (0.41) 1.00 (0.25)
Adjusted
Ref. 0.28 (0.07)z 0.57 (0.17) 1.13 (0.31) 0.57 (0.15)w
Crude
Ref. 0.65 (0.22) 1.29 (0.45) 1.95 (0.63)w 1.36 (0.41)
Adjusted
Ref. 2.03 (0.51)z 4.01 (0.94)z 2.02 (0.45)z
Crude
Ref. 1.97 (0.65)w 2.99 (0.96)z 2.08 (0.64)w
Adjusted
Korean [OR (SE)]
Ref. 1.97 (0.50)z 0.99 (0.24)
Crude
Ref. 1.51 (0.50) 1.06 (0.28)
Adjusted
Vietnamese [OR (SE)]
Ref. 0.50 (0.12)z
Crude
Ref. 0.70 (0.20)
Adjusted
South Asian [OR (SE)]
*Adjusted model controls for living in an ethnically concordant neighborhood, educational attainment, employment status, insurance, household income, health status, language proficiency, length of residence, age, gender, marital status, household size, and survey year. w P < 0.05. z P < 0.01. OR indicates odds ratio.
Ref. 0.65 (0.16) 0.48 (0.13)z 0.31 (0.08)z 0.61 (0.19) 0.93 (0.25) 0.65 (0.15)
Adjusted
Crude
Ref. 0.66 (0.12)w 0.98 (0.23) 0.28 (0.05)z 0.56 (0.12)z 1.11 (0.23) 0.56 (0.11)z
Crude
Japanese [OR (SE)]
Volume 52, Number 9, September 2014
Chinese Filipino Japanese Korean Vietnamese South Asian Other Asian
Filipino [OR (SE)]
Chinese [OR (SE)]
TABLE 2. Pair-wise Comparison of Asian American Ethnic Subgroups in Having a Usual Source of Care Other Than the Emergency Room, California Health Interview Survey, 2005 and 2009*
6.8 (5.1–9.1) 25.7 (21.3–30.5) 19.3 (15.7–23.5) 19.8 (15.9–24.3) 28.5 (23.1–34.5)
7.2 20.3 24.5 24.5 23.4
6.8 19.9 23.3 26.3 23.8
(5.0–10.4) (17.1–24.0) (21.6–27.7) (21.6–27.7) (19.1–28.2)
63.8 (59.3–68.1) 53.2 (47.9–58.4) 65.5 (58.9–71.4) 58.9 (51.2–66.1) 59.4 (52.5–65.9) 68.8 (62.5–74.5) 42.7 (37.7–47.8)
58.8 (56.7–60.8) (5.9–7.8) (18.2–21.7) (21.6–25.1) (24.6–28.1) (21.7–25.9)
22.5 (19.3–26.1) 20.1 (16.9–23.7) 33.3 (27.7–39.4) 17.2 (14.4–20.5) 25.9 (21.9–30.3) 14.2 (11.2–17.8) 13.6 (10.9–17.0) 13.8 (11.8–16.2) 15.1 (12.2–18.5) 26.7 (21.5–32.7) 10.7 (8.4–13.4) 14.2 (11.4–17.5) 7.8 (5.6–10.8) 5.4 (4.1–7.0)
20.3 (18.9–21.7) 12.8 (11.8–13.8)
All percentages are weighted and might not add to 100 because of rounding. *Aggregate Asian Americans include Chinese, Filipino, Japanese, Korean, Vietnamese, South Asian, and Other Asians. w The w2 test of significance among Asian ethnic subgroups. z High English proficiency responses included English only, very well/well and limited English proficiency responses included not well/poor. y Recent immigrants have been in the US for
