CMPH

Korean immigrants have been identified as one of the most vulnerable populations in the United States in terms of healthcare access. Prior to the Affordable Care Act (ACA), Korean immigrants had the highest uninsured rates among Asian ethnic groups, with 29.9% lacking health insurance [1]. Data from the California Health Interview Survey revealed that 36.1% of Korean Americans ages 18-64 in California were uninsured [2]. Although the implementation of the ACA led to significant strides in healthcare coverage for Korean immigrants, this group still faces challenges in consistently utilizing and comprehending healthcare plans. This study examines whether or not the type of healthcare plan explains the disparities that this population encounters.

Types of Health Insurance

Employer-Sponsored Health Insurance (ESHI) has played a pivotal role within the U.S. healthcare system [3] since its inception.

Initially, it served as a strategy for employers to attract employees during World War II without breaching wartime wage controls, with each employer offering highly individualized healthcare coverage. This led to the development of the complex healthcare system that exists today. Despite the widespread adoption of ESHI, the reliance on the private sector in the U.S. healthcare system has left a significant portion of the population uninsured [4,5] and therefore susceptible.

To address this issue, the ACA was enacted in 2010 [6]. This resulted in a notable decrease in uninsured rates through various provisions such as Medicaid expansions for low-income individuals and subsidies for private health insurance purchased on federal or state exchanges [7,8]. Following the ACA’s implementation in January 2014, uninsured rates among significantly from 20.1% in 2013 (as of the fourth quarter) to 15.1% in 2014, and further to 12.6 % in 2015 [9]; it remained at 12.1% in 2022 [10]. Particularly, the ACA increased health coverage among self-employed individuals and wage earners without ESHI plans, achieving coverage rates equal to or greater than that of employed individuals [11]. By expanding coverage for workers without ESHI during job transitions, parttime employment, early retirement, or self-employment, the ACA has weakened the traditional ties between employment and health insurance in the U.S. healthcare system [7]. However, questions still remain regarding the differences in the quality of healthcare coverage between ESHI and other types of health coverages. Better understanding these differences will help to equip the policy.

Korean Immigrants’ Healthcare Plan

Historically, studies have attributed the elevated uninsured rates among Korean immigrants to their notably higher rate of selfemployment compared to other racial/ethnic groups, especially before ACA [12,13]. Nearly 30% of Korean immigrants in the United States are self-employed, approximately three times the rate of White Americans [14,15]. Many Korean immigrants operate family-owned small businesses due to socio-cultural and economic factors, including perceived disadvantages and limited opportunities in traditional mainstream employment sectors [16,17]. The high rates of self-employment and employment in small firms among Korean immigrants restrict their access to ESHI, which is the primary source of coverage for most nonelderly immigrants, consequently contributing to their low rates of health coverage [18].

Since the ACA’s implementation, many self-employed Korean immigrants have gained access to healthcare. However, it remains unclear whether Korean immigrants who have healthcare through ACA adequately utilize health insurance compared to those with ESHI. While health insurance coverage has long been seen as a necessity for accessing healthcare and ensuring appropriate medical care [19], it is important to acknowledge that having health coverage does not always guarantee seamless access to services [20]. Racial and ethnic minorities and immigrants often face distinct challenges when seeking healthcare, such as cultural and language barriers that hinder effective use of health insurance [21-23]. It is especially noteworthy that self-employed Korean immigrants, historically marginalized within the US healthcare system, may face disproportionate obstacles to equitable access even if they are enrolled in insurance. For example, those without ESHI may encounter barriers due to potentially inferior coverage or higher out-of-pocket costs compared to those with ESHI.

Healthcare Plans and Social Networks

Social networks and social capital within immigrant communities have always been valuable resources for gaining information on various aspects of life. Korean immigrants, particularly those facing linguistic barriers and those unfamiliar with the U.S. systems, rely heavily on ethnic social networks to navigate the unfamiliar circumstances including healthcare services system [24]. However, the benefit of social networks may not apply to everyone within Korean immigrant communities, and the availability of information about health insurance and healthcare services within these networks can be significantly limited by the types of healthcare plans, potentially influencing healthcareseeking behaviors [25]. In a qualitative study, self-employed Korean immigrants reported that they often refrain from seeking information regarding health insurance within their ethnic networks due to the perceived inadequacy of useful information about health insurance benefits, costs, and usage [26]. Moreover, many selfemployed Korean immigrants are hesitant to seek information about health insurance from their ethnic social networks out of fear of revealing their socioeconomic conditions to their ethnic acquaintances, further impeding their access to relevant resources [26]. These findings suggest that individuals without ESHI may have reduced access to health discussion networks, as they share similar experiences regarding health insurance.

Current Study

While numerous studies on healthcare access for Korean immigrants have focused on identifying factors that contribute to health insurance coverage, there is a notable lack of research directly examining how the type of coverage shapes individuals’ experiences of healthcare plans as well as the impacts of access to health discussion networks where healthcare-related information is exchanged. Given the U.S. healthcare system’s heavy reliance on ESHI and the unique employment status of Korean immigrants, a significant research gap exists in understanding whether the absence of ESHI leads to discernible disparities in healthcare utilization within this population.

To address this gap, this study examines the association between access to ESHI, health insurance-related behaviors, and healthdiscussions within social networks. To be specific, the health insurance-related behaviors examined in this study are consistent utilization, level of satisfaction, and perceptions on affordability, accessibility, and comprehensibility as these are critical indicators that can influence healthcare-seeking behaviors [27-29]. We hypothesized that individuals without ESHI are more likely to exhibit negative health insurance-related behaviors, such as inconsistent utilization and lower satisfaction. Individuals without ESHI are also more likely to exhibit unfavorable perceptions regarding affordability, accessibility, and comprehensibility. In addition, we hypothesized that lack of ESHI leads to lower likelihood of having any health discussion networks, particularly within workplace settings.

A comprehensive understanding of the impact of ESHI, or lack of it, is essential for addressing significant disparities in healthcare access among disadvantaged population. Gaining insight into how access to or lack of ESHI shapes patients’ experiences as they navigate the healthcare delivery system, as well as other factors that influence their experience, is crucial for informing and guiding healthcare policy.

METHODS

Data and Sample

Data was drawn from surveys conducted on a community-based sample of Korean immigrants between the ages of 26 and 64. A purposive sampling method was employed to recruit this sample, targeting individuals meeting the age and ethnicity eligibility criteria. The respondents were recruited through 58 communitybased Korean churches in the Greater Los Angeles area between October and December 2018. The Greater Los Angeles area was chosen as the research setting due to its significant concentration of Korean immigrants [30]. This concentration provided a unique opportunity for researchers to collect relevant data on Korean immigrants residing in ethnic enclaves [31]. Local churches were selected as the primary data collection sites, recognizing their central role in developing social networks among Korean immigrants and where Korean immigrants can receive formal and informal services and information on health insurance and healthcare [32,33]. The survey was conducted at various locations and events, including churches and small group fellowship meetings. Following the acquisition of informed consent, participants were asked to complete a paper-based, selfadministered questionnaire in either Korean or English, based on their language preference. The questionnaires were distributed and collected by the researcher or designated contact persons within the churches, such as pastors and small group leaders. A total of 549 Korean immigrants completed the survey. This study focuses on the 386 Korean immigrants with health insurance at the time of the survey, excluding the participants uninsured. After deleting the observation with missing values in variables, the final sample size is 377. The project adhered to ethical guidelines and received approval from the Institutional Review Board at the [blind for review].

Measurement

Employer sponsored health insurance (ESHI)

Those who are currently insured were asked which type of health insurance they have. Those who marked employer-sponsored insurance were coded as 1, and all other answers were coded as 0. Other types of health insurance are Obamacare, governmentsupported programs including Medi-Cal, and private health insurance.

Health Insurance Utilization and Perceptions

Health insurance utilization

In the original questionnaire, participants were asked to indicate on a five-point scale how often they use health insurance when they need health care services: always, most of the time, sometimes, occasionally, and never. The answers were dichotomized into always (1) or not (0).

Health insurance satisfaction

Participants were asked how satisfied they were with their current health insurance using a five-point Likert scale question.

Perception on affordability

Participants were asked four questions that gauged how strongly participants agreed using a five-point Likert scale, and the mean of four answers was used to measure insurance affordability (α = .88). The statements included “Out-of-pocket cost (deductible, copayment) is affordable.”

Perception on comprehensibility

Three questions were used to assess how well participants understand the insurance policy (α = .89), including “It is easy to understand the policies of health insurance.”

Perception on accessibility

Two questions were asked to assess the perceived accessibility (α = .83) which are “it is easy to access the network hospitals included in my coverage,” and “it is convenient to choose health care providers that I want.”

Health Discussion Networks

Characteristics of social networks from which respondents obtained health information were inquired about using a name generator and a name interpreter [34]. Participants (egos) were asked for the initials of up to 5 people (alters) with whom they talk about their health-related concerns. Attributes of each alter, as perceived by the respondents, were asked, including the alters’ gender, ethnicity, relationships with respondent, frequency of contact, topic of health discussion, and levels of satisfaction with the health information obtained from each alter. In this study, all health discussion networks were measured in binary form. Those with access to the networks were marked “1” and those without were marked “0.”

Health Discussion Networks

If participants had any alter, they were recorded to have access to health discussion networks.

Health Discussion Networks at Work

Among those who have access to health discussion networks, if any of the alters were known from work, the participants were marked to have access to health discussion networks at work. Even if an alter is marked to be a contact from more than one context, the person was considered to have a health discussion network at work if the alter was marked to be known at work. Respondents who did not have any health discussion networks or who did not have any health discussion networks at work were marked as “0”.

Covariates

Age was calculated from the year of birth and the time the data was collected. Gender was asked about in a binary form considering the context of data collection. Female individuals were marked as “1” and male individuals were marked as “0.” Marital status was asked about in four categories (married, single, widowed, and divorced) and was dichotomized to married or not. Participants were asked whether they were born in the U.S., and immigration status was asked in four categories (U.S. citizen, permanent resident, visa holder, and others). Whether one has a college degree or not and whether one’s main occupation is self-employed or not were included as covariates considering their impact on health-related behavior.18 Whether participants pursued insurance due to the implementation of the ACA was recorded in a binary form.

Analyses

Descriptive analyses of outcome variables and covariates were conducted by whether or not one has ESHI. Logistic regression and Ordinary Least Squares (OLS) regression models were used to investigate the health insurance utilization and perception. For the binary dependent variable, logistic regression was used; for continuous outcome variables, OLS regression models were used. To investigate the relationship between the type of health insurance and the access to health discussion networks, logistic regression models were estimated. For a rigorousness test, treatment effect of ESHI was calculated using Inverse Probability Treatment Weighting (IPTW) regression. IPTW regression provides an analysis that accounts for selection bias into certain conditions such as having ESHI. To estimate IPTW treatment effect, several steps were taken. Propensity scores for ESHI were estimated using logistic regression with the same set of covariates. Then, only the participants with common propensity scores remained in the sample and 11 observations were dropped, resulting in 366 in the sample for the IPTW analysis. The average treatment effect was estimated using the weight calculated in the above process. All analyses were conducted using STATA 17.0 for Mac.

RESULTS

Descriptive analysis

The descriptive analyses of variables are presented in Table 1 by the type of health insurance. As shown in the sample size, only about half of those insured are insured through employer-sponsored health insurance. This is far less than national average which is 60% among those under the age of 65. In total, 73% of those with ESHI reported utilizing health insurance whenever they use healthcare services, while only 43% of those without ESHI reported so. The difference in the utilization of health insurance is reported at a significant level of 99.9% using the chi-square test. Means and standard deviation of perceived affordability, accessibility, and comprehensibility of health insurance are reported, as well as whether the differences in the outcomes are significant. Health insurance satisfaction, affordability, and accessibility significantly differ among the two groups.

Table 1: Descriptive Analysis by insurance type.

Is your health insurance sponsored by your employer?

Note. n.s. signifies the difference is not statistically significant when tested using t-test or chi-square test.

*** p<0.001, ** p<0.01, * p<0.5, +p<.01

More than 90% of participants had some access to health discussion networks among both groups. Among individuals with ESHI, 35% reported having access to health discussion networks at work, compared to only 23% among those without ESHI. The difference in access to health discussion at work was significantly different between the two groups.

ESHI and Health Insurance Utilization, Satisfaction, and Perceptions

The results of logistic and OLS regression examining the relationship between ESHI and health insurance utilization, satisfaction, and perception are presented in Table 2. ESHI is significantly related to consistent utilization of health insurance when Korean immigrants use healthcare services (OR = 2.76, [1.67, 4.54]). ESHI is marginally significantly related to satisfaction of health insurance (b = 0.20, p < .10). Having ESHI is related to the perceived accessibility to health insurance at the marginal level (b = 0.25, p <.10). Having ESHI is not significantly related to how participants perceive the affordability and comprehensibility of health insurance.

Table 2: Association between having employer sponsored health insurance (ESHI) and health insurance utilization and perception.

Note. Confidence interval or standard error in parentheses

*** p<0.001, ** p<0.01, * p<0.5, +p<.01

ESHI and Health Discussion Networks

The results of a logistic regression with all covariates considered show that ESHI is not statistically significantly related to the access to health discussion networks. On the contrary, ESHI is significantly related to having access to health discussion networks in the workplace (OR = 1.75, [1.03, 2.97]) (Table 3).

Table 3: Association between having employer sponsored health insurance (ESHI) and access to network.

Confidence interval in parentheses

*** p<0.001, ** p<0.01, * p<0.5, +p<.01

Rigorousness Test Using IPTW Regression

The results from IPTW are presented in Appendix 1 as a table. In establishing comparable samples between those with and without ESHI, 11 individuals from the sample were further dropped to ensure the propensity score of each group is within the range of the other group. As a result, the smallest and largest probabilities of having ESHI are within 0.2 standard deviation of the lowest and the highest probabilities. The means and the standard deviations of covariates after weighing are very close to 0 and 1, respectively, which shows that the groups are well-balanced in terms of the probability of having an ESHI.

The average treatment effect after weighting is consistent with the main results presented above using logistic and OLS regressions. Having ESHI is significantly related to effective utilization of health insurance, while it is not related to any other variables regarding the perceptions on health insurance. Also, having ESHI is related to the higher likelihood of having health discussion networks at work. However, it is not related to health discussion networks in general.

DISCUSSION

While Korean immigrants show advancement in other areas of socioeconomic statuses, they continue to be one of the least insured immigrant populations [20,26]. Although the proportion of those insured has increased as a result of the enactment of the ACA [1], our results show interesting patterns among this population that may inform health insurance policies and practice for immigrants: (1) many Korean immigrants do not consistently use health insurance for healthcare services even if insured, (2) many Korean immigrants experience unequal utilization rates based on health insurance type, yet (3) they have similar perceptions of the health insurance.

The results show that health insurance usage is not always warranted among Korean immigrants. Only half of the sample consistently used health insurance when they utilized healthcare services, even if insured. This is contrary to the understanding that enrollment in health insurance automatically leads to higher usage of health insurance and therefore healthcare services, as some studies have found [35,36]. While social policies like the ACA aim to enroll immigrants in health insurance so that they can utilize it when health issues occur, it may not facilitate effective utilization automatically. On the other hand, it only imposes the burden of enrollment and paying