|Home | About | Journals | Submit | Contact Us | Français|
US Department of Health and Human Services reported that the lack of English language proficiency and the shortage of providers who possessed appropriate language skills were identified as major barriers to mental health service use for approximately half of the population of Asians and Pacific Islanders. The aim of this study was to examine the predictors of lifetime mental health service use in relation to English language proficiency among Asian Americans.
Data from 2095 Asian participants from the National Latino and Asian American Study were analyzed using logistic regression.
Respondents with better English language proficiency and with a mental health diagnosis were more inclined to use mental health services. Participants who were born in the USA, who were widowed, separated or divorced, who sought comfort from religion, who reported worse physical and mental health self-ratings were more likely to use mental health services. The lack of health insurance coverage was not a significant predictor.
The public health implications for behavioral health include the need to educate health-care providers working with Asian Americans regarding the benefits derived from seeking services and making interpreter services available in a culturally sensitive environment.
The existence of barriers to mental health service utilization represents a critical public health issue in the USA. Only 37% of people with major depressive disorder accessed mental health services within the first year of an episode.1 Approximately 88% of those with major depressive disorder eventually engage in mental health treatment at some point; however, the average treatment delay for those who do not obtain assistance during the first year can range from 6 to 8 years.1 Even longer treatment delays can occur for other disorders.2,3
When examining the mental health service use by people of color, the underutilization becomes even more problematic. People of color in the USA utilize mental health services at lower rates than whites.4–6 The findings from one study indicated that Asian Americans demonstrated lower rates of any type of mental health-related service use than did the general population.7 Asians who eventually use mental health services are more severely ill than whites who use the same services. Moreover, families may tend to discourage the use of mental health facilities among family members until the members become unmanageable.8 Unlike other ethnic minority groups in the USA, the delay of service use among Asian Americans may be due to culturally specific factors, including language barriers between mental health service users and providers.
The ability of nearly 50% of Asian American/Pacific Islanders to use the mental health-care system is limited due to lack of English proficiency, as well as the shortage of providers who possess appropriate language skills.6 The extent of interpreter services for Asian Americans varies by geography and provider type. Where the population of Asian Americans reaches a specific threshold, such as 10% in California, interpreter services are available by in-house staff or on a consultative basis. Approximately 80% of American hospitals encounter limited-English-proficient (LEP) patients; 43% of the hospitals encounter LEP patients daily, 20% encounter them weekly and 17% encounter them monthly (Armada and Hubbard, unpublished work). A prerequisite for health-care organizations and clinicians to address language factors in health-care delivery is to identify those patients most likely to benefit from language assistance.9 In 2009, the Joint Commission, the National Committee on Quality Assurance and the National Quality Forum all issued draft statements proposing new cultural competence standards for hospitals, health plans and other health-care organizations. All three sets of guidelines endorsed the collection of race, ethnicity and language data. To encourage health-care organizations to track language-associated disparities and address these, the Joint Commission requires the maintenance of records on patients' language and communication needs.10,11 In addition, the Department of Justice guidance on implementing the Civil Rights Act (Title VI) for LEP persons stipulates that language assistance can take many forms, including language concordant clinicians and staff, and professional interpreter services provided either in person or remotely via telephone or videoconferencing.12
This study's conceptual framework was based on the Andersen–Newman13,14 behavioral model of health services use. In this model, an individual's propensity to use health care is influenced by predisposing, enabling and need characteristics (Fig. 1). In the traditional Andersen–Newman health behavior model, literacy was included in the predisposing domain that cannot be changed by intervention. However, among the immigrant population, the nature of literacy can be improved. Thus, English language proficiency was included separately for the purpose of this study. We hypothesized that Asian Americans with low levels of English language proficiency would be less likely to use mental health-care services after controlling for various predisposing, enabling and need factors.
Data were from the National Latino and Asian American Study (NLAAS). The NLAAS is a nationally representative community household survey that estimates the prevalence of mental disorders and rates of mental health utilization by Latinos and Asian Americans in the USA.15 Data were collected between 2001 and 2003 during face-to-face interviews. The inclusion criteria were: (i) 18 years of age or older; (ii) living in the non-institutionalized population of the continental USA or Hawaii; and (iii) Latino, Hispanic or Spanish descent, or Asian descent. A more detailed description of the development of the NLAAS was published by Alegria et al.16 The sample of 2095 included 520 Vietnamese, 508 Filipino, 600 Chinese and 467 Other Asian.17
On the basis of the Andersen–Newman health behavior model, independent variables were selected. The predisposing domain included age, gender, ethnicity, marital status (married or cohabiting, never married, and widowed, separated or divorced), education level [categorized as less than high school graduate (0–11), high school graduate (12), some college (13–15) and college graduate or beyond ≥16 years], employment status (0, currently not working; 1, currently working), age at immigration (categorized as US born, <12, 13–17, 18–34 and 35+ years) and household size. The enabling domain included household income, health insurance coverage (without health insurance versus with health insurance plan from at least one of the followings: military, employer or union, purchased from insurance company, Medicare, Medicare supplemental or Medigap, government assistant program, state health insurance program for uninsured people or other) and religion (during difficult times people seek comfort in religion, often, rarely or sometimes and never). The need domain included self-rated perceived physical and mental health status (from poor to excellent), and the presence of a mental health diagnosis per the DSM-IV in the respondents' lifetime.18
English language proficiency was assessed using questions that assessed speaking, writing and reading skills (from 1 = poor to 4 = excellent), which were then collapsed into one variable. The dependent variable was the use of mental health services during the respondents' lifetime in the USA. The services included counseling service experience with (i) psychiatrist, (ii) medical doctor, (iii) psychologist, (iv) social worker, (v) counselor, (vi) any other type of mental health professional (e.g. psychotherapist or mental health nurse) and (vii) nurse, occupational therapist or other non-MD health professional. This variable was coded dichotomously (0, none; 1, at least once). The nine respondents who last used mental health services prior to immigration were coded as non-users.
This study is using a secondary data set; thus, IRB approval exemptions were granted. IRB exemptions were obtained by authors from their own institutions.
Descriptive statistics and logistic regression analyses19,20 were conducted using SPSS, version 16, with the complex samples add-on module. Sampling weight, strata and cluster variables were applied to conduct the data analyses. Observations were weighted to ensure national representativeness. Standard errors were adjusted by taking the complex survey sampling frame into consideration.
Table 1 presents the socio-demographic characteristics of the weighted sample. The mean age was estimated as being 41 years old. The average number of household members was 2.86. Fifty-three percent would be identified as female. The weighted proportion of married people was 70%, and about 76% were foreign born. Approximately 43% had an education level of 16 years or more, whereas 66.5% were employed. Further, 22.3% reported that they have used at least one mental health service. Regarding health insurance coverage, 87% of the population were estimated to have some form of health insurance. For the religiosity, 33% reported never seeking comfort through religion.
Table 2 presents the results of the logistic regression. The respondents who speak, write and read English were more likely to use at least one mental health service once in their lifetime in the USA [odds ratio (OR) = 1.15; 95% confidence interval (CI) = 1.03, 1.28; P = 0.011]. The comparison among four ethnicity groups was not statistically significant (P = 0.103). Overall, marital status was statistically significant (P = 0.01).
Compared with married respondents, divorced/separated/widowed respondents were more likely to use the service (OR = 2.17; CI = 1.49, 3.16; P ≤ 0.001). Overall, education level was statistically significant (p = 0.023). Compared with respondents who finished college, respondents who did not finish college were more likely to use the service. The age at immigration to the USA was a significant predictor. Compared with respondents who were born in the USA, those who immigrated to the USA between the ages of 18 and 34 or 35 years old or older were less likely to use mental health services. No difference was noted between respondents born in the USA and those who immigrated to the USA at an age younger than 12 or who immigrated to the USA between the ages of 13 and 17. Household size was not statistically significant; however, a trend toward significance was noted (P = 0.089).
Although household income and health insurance coverage among the respondents were not significant, religion was statistically significant with P < 0.001. Compared with respondents who never sought comfort in religion during difficult times, respondents who often sought comfort were more likely to use mental health services (OR = 2.13; CI = 1.37, 3.31; P = 0.001), whereas respondents who sometimes or rarely sought comfort were also more likely to use mental health services (OR = 2.16; CI = 1.45, 3.23; P < 0.001).
Compared with the respondents who rated their physical health as excellent, the odds for respondents (who rated their physical health as poor) to use mental health service were 3.42 times as high as the odds of not using mental health services. Compared with the respondents who rated their mental health as excellent, the odds for respondents (i) who rated their mental health as poor to use mental health service available were 8.43 times as high, (ii) who rated their mental health as fair to use at least one service are 3.10 times as high and (iii) who rated their mental health as good were 2.47 times as high. Compared with respondents who had no mental health problems based on a DSM-IV diagnosis in their lifetime, the odds for those who had any mental health problem to use at least one service in their lifetime in the USA were 6.27 times as high as the odds of not using mental health services.
Respondents with better English language proficiency and respondents with mental health diagnosis during their lifetime were more inclined to use mental health services in their lifetime in the USA. This study found the following variables in the framework as being important predictors: marital status, education level, age at immigration, religion, self-rated physical health and self-rated mental health.
Relative to the education level and mental health service use, respondents who did not finish college were more likely to use the service. This finding requires further study. The attainment of education is critically important to Asian Americans. Respondents who could not finish college may incur more stress associated with saving face, which may lead to the use of more mental health-care services. Alternatively, due to mental health problems, the respondent might be unable to finish college. Individuals with high levels of psychopathology have impaired information-processing skills, which are a critical component of academic performance and success.21
Participants, who were born in the USA, who were widowed, separated or divorced, who sought comfort from religion, who reported worse physical and mental health self-ratings, were more likely to use mental health services. The lack of health insurance coverage was not a significant predictor.
There have been several studies to explore the issue of the underutilization of health services due to LEP skills. Underutilization of physical health service due to language and communication barriers has been well documented with Chinese in the UK,22 Australia23 and with other linguistic minority populations in the USA.24–26 In the mental health service use, Lau and Zane27 reported that for respondents with one episode of mental disorder in Los Angeles, there was a language barrier between providers and respondents. Another noted barrier was the limited number of providers with proper language skills to assist the Asian American population.28 However, these studies focused primarily on one episode of mental disorder in a shorter time period or focused on physical rather than mental health.
The objective of this study was to examine the predictors of lifetime mental health service utilization in relation to English language proficiency among Asian Americans. This study presents the first nationally representative study of Asian Americans in this particular area. As reported by Sue,29 Asians can be viewed as the model minority; therefore, limited resources are made available in the mental health community. Further, if the population does not request mental health services because they see their behavioral health issues as somatic, this can delay their treatment.
Surprisingly, this study found a higher level of health insurance coverage than did the previous studies. The enabling factors of influence include lack of health insurance coverage or inability to pay for care.30 Access to mental health care often depends on health insurance coverage, and about 21% of Asians and Pacific Islanders lack health insurance.31,32 Within the subgroups, the rate varies significantly; e.g. 20% of Chinese and Filipino Americans and 34% of Korean Americans lack insurance. The rate of Medicaid coverage for most Asians and Pacific Islanders is well below that of whites, possibly due, in part, to mistaken concerns among immigrants that enrolling themselves and their children in Medicaid could jeopardize their application for citizenship. Our study did not find a significant association between the health insurance coverage and mental health service use. The possible explanation for this might be that the outcome variable of this study is lifetime mental health service use, and the questions related to health insurance coverage may provide health insurance coverage at the time the respondents were interviewed; thus, this cannot answer whether the respondent has insurance during his or her lifetime. Families USA reported that the number of persons without health insurance increases if we consider the long-term period rather than short-term period.33,34 Thus, even people with health insurance during the survey period are not likely to have health insurance at a specific point during their lifetime, especially among the immigrant population.
First, some important variables could be included in the analyses, such as mental health belief that influences mental health care. Asian Americans demonstrated that they were less likely than whites to mention their mental health problems to a friend or relative (12% versus 25%), psychiatrist or mental health specialist (4% versus 26%) or physician (3% versus 13%).35 Attitudinal beliefs influence an individual's decision to disclose mental health issues. Sue et al.36 found evidence that the reluctance to use services is attributable to factors, such as the shame and stigma accompanying use of mental health services, cultural perceptions of mental health and treatment that may be inconsistent with Western forms of treatment, and the cultural or linguistic inappropriateness of services. Asian American population grew up in families with both parents, had limited exposure to drugs, gangs and crime in their neighborhoods, performed well in high school, had low incarceration rates and low rates of teenage pregnancy.37 Although these environments may be seen initially as protective from a public health perspective, the mental health hazards in the population require oversight.
Second, within the constraints of the NLAAS data set structure, we were unable to look in detail at within-group variability characteristics that are well recognized among Asian Americans in the USA. Asian Americans/Pacific Islanders are diverse in ethnicity and include as many as 43 different ethnic groups, which utilize more than 100 languages and dialects.6 Yet, much of the research in this area has focused on the Asian American/Pacific Islander population as a whole. Research has suggested that Asians Americans/Pacific Islanders appear to suffer less from mental health disorders and have higher rates of economic and educational success than other people of color. However, considering the Asian American population as a single group can lead to the conclusion that Asians/Pacific Islanders are a ‘model minority’, glossing over within group variations.36
Third, the utilization of service is based on respondents' lifetime experiences, whereas their health status variable indicated their current mental health status, rather than their past mental health status. Only those people who answered ‘fair’ and ‘poor’ to the mental health status question were included in the study analysis. The research excluded those who used health-care services during their lifetime and who currently have a good or excellent health status. The sample may include people who had good or excellent health status in the past and did not use health services. The time difference in the two variables might create selection bias of our sample.
In conclusion, there must be a reciprocal and concerted effort among the providers, patients and public health community to quantify the extent of this isolation and then develop plans that reduce barriers to care. The US federal Civil Rights laws protect Asian Americans relative to accessing health services and the US Health Insurance Portability and Accountability Act protects the confidentiality of the services they access. The US federal government requires that all hospitals provide patients with interpreters if needed.38 The US government has not set formal standards for interpreter services and does not require quality assessment of these services. However, the fact that the federal government requests that these services are provided strongly encourages health-care providers to offer these services.
Further areas of study should include some of the following: Does the Asian population recognize that English language proficiency is associated with mental health? Do ESL courses facilitate more effective integration of Asian Americans into the population and thus serve to reduce mental health issues? Unfortunately, there have been eliminations of ESL classes due to current US economic downturn. Thus, the provision of interpreter services in health-care settings may be more critical.
This study was partially supported by Emerging Scholars Interdisciplinary Network University of Michigan, Research Group grant and by National Institute on Aging (NIA) Fellowship (T32-AG000117).