This study suggests that among insured Latinos, those with low English language proficiency have more negative experiences of primary care than their English language proficient counterparts, with decreased access (longer wait times and greater difficulty obtaining information or advice by phone) and less continuity. There was no association demonstrated between low English language proficiency and difficulty getting an appointment over the phone. We suspect that the reason for this lack of association is that this outcome captures an aspect of the primary care experience that is less complex than the other three measures, and therefore less likely to be sensitive to the strains imposed by language barriers. Both simple interventions (bilingual office staff) and minimal language proficiency on the part of subjects may be enough to allow them to circumvent difficulties in obtaining appointments, but not improve these other dimensions of primary care.
The findings of this study are in keeping with past work demonstrating an association between low English proficiency and less timeliness of care, as well as poorer communication with providers and less helpful staff.10
Consistent with past work, our results demonstrate an association between low proficiency in English and continuity of care.9
This body of work has implications both in terms of ongoing research, practice, and policy. Future studies should consider interventions that address the barriers faced by Latinos with limited English proficiency, including training more physicians who speak Spanish and who are culturally concordant,17
and better integration and availability of translation services.18
With regards to policy, only 9 states offered direct reimbursement for the cost of language interpreters in 2003, despite data suggesting that these services are cost-effective.19, 20,21
In addition, while addressing language barriers is mandated by federal law, there remains need for increased standardization of approaches to ensuring linguistic competence in healthcare.19,22
Policy makers ought to examine both the implementation of services aimed at mitigating the effects of language barriers as well as the implementation of existing federal and state legislation.
Insurance status was associated with the experience of primary care in this study. Subjects with Medicare or Medicaid were significantly more likely than those with private insurance to report having long waits, and they also experienced a non-significant trend towards having greater difficulty obtaining information or advice by phone. Past work has demonstrated that lack of insurance is an important mediator of the difference between Hispanics and whites in access to care,23
but less is known about the relationship between type of insurance and the experience of primary care among Latinos. A study of the elderly (including Latinos and African Americans) in California found that subjects with either Medicare coverage or Medicare plus Medicaid coverage were less likely than subjects with Medicare plus private supplemental insurance to report use of a number of preventative services.24
This study overcomes some of the limitations of previous work by the inclusion of a nationally representative sample of Latinos from a number of different ethnic subgroups and with various types of insurance. Since Latinos are the largest and fastest growing minority group in the United States, with a population that exceeds 42 million in 2005, it is particularly timely to focus on the experiences of this population.25
This study has several limitations. The outcomes are based on self-report. Past work has documented the presence of differences in reliability and validity between the responses to English and Spanish versions of surveys; however, the majority of the differences were due to the tendency for Spanish speakers to give more favorable responses.26
Such a bias, however, would have caused the results of our study to be under-estimated, suggesting that the strong association demonstrated between language proficiency and experience of primary care is robust. In addition, the study sample is drawn from subjects who all report some type of health insurance, preventing us from examining the impact of language proficiency among the uninsured. This analysis does not include information on language concordance between subjects and providers nor on the availability of interpreter services. Future work ought to include such variables in order to better elucidate the mechanisms underlying the relationship between language proficiency and the quality of primary care.
Among insured Latinos, low English language proficiency is associated with worse reports of the experience of primary care despite health insurance coverage. Reducing disparities in the quality of care for this growing segment of the US population requires that providers and payers address linguistic barriers to care.