Research That Matters (January 17 - 20, 2008)

PURPOSE. This study examines the effect of limited English proficiency on health insurance coverage in California. In 2005, over 1 in 10 Californians were limited English proficient (LEP) or spoke English “not well” or “not at all”, including 35% of Spanish-speakers and 24% of those who speak an Asian language. Health insurance is a key predictor of health access, but navigating the complex health insurance system is already difficult for many individuals, without language being an additional barrier. While it is well-documented that immigration plays a significant role in predicting health care coverage, less attention has been given to a person's ability to speak and understand English. METHOD. Using 2005 California Health Interview Survey data, bivariate and regression analyses were used to determine whether English proficiency predicts current health coverage (logistic: yes vs. no), insurance type (multinomial: private vs. public vs. uninsured), and the risk of intermittent insurance coverage (ordered probit: 0 months, 1 to 5 months, 6 to 11 months, and 12 months). The primary predictor—English language proficiency—was categorized as those who speak English only, very well/well, and not well/not at all. Interaction terms for English proficiency and race/ethnicity were also included to determine whether the effects differed across racial/ethnic groups. Key demographic, socioeconomic, and health-related factors were also controlled for in each model. The final sample consisted of 32,852 nonelderly adults aged 18 to 64. RESULTS. Forty-four percent of LEP adults were uninsured, compared to 10% of English speakers. Roughly one-quarter of LEP adults had public insurance (e.g. Medi-Cal), while only 30% had employer-based or privately-purchased insurance. About half of the LEP population was uninsured for six or more months in the past year, compared to only 10% of English speakers. After adjusting for socio-demographic factors, the regression analyses showed that LEP individuals were less likely than English speakers to be insured (OR = 0.55, p < .001) and more likely to receive public insurance than private insurance (OR = 1.66, p < .001). Being LEP also increases the risk of being uninsured at some point during the past year by 9 percent and for the entire year by 6 percent (both p < .001). While the interaction effects of English proficiency and race/ethnicity were largely nonsignificant, the predicted probabilities of being insured for LEP Latinos and Asians were at least 30 percentage points lower than their English-speaking counterparts. CONCLUSION. The findings suggest that health insurance in California remains inaccessible to those who cannot speak or understand English. After adjusting for demographic and socioeconomic characteristics, LEP adults still face a greater risk of being uninsured and having intermittent coverage throughout the year. To ensure that benefits are accessible to LEP individuals, both private and public health insurance plans should develop language access policies and services, e.g. translated materials, bilingual staff, community-based outreach, use of ethnic media, and community workers. Simplification of the application process and reporting requirements, e.g. in the Medi-Cal program, could also encourage LEP individuals to apply and remain insured.