Objective
To examine the association of patient-provider racial and ethnic concordance on healthcare use within Hispanic ethnic subgroups.
Methods
We estimate multivariate probit models using data from the Medical Expenditure Panel Survey, the only national data source measuring how patients use and pay for medical care, health insurance, and out-of-pocket spending. We collect and utilize data on preventive care visits, visits for new health problems, and visits for ongoing health problems from survey years 2007–2017 to measure health outcomes. Additionally, we include data on race and ethnicity concordance, non-health-related socioeconomic and demographic factors, health-related characteristics, provider communication characteristics, and provider location characteristics in the analysis. The sample includes 59,158 observations: 74.3% identified as Mexican, 10.6% identified as Puerto Rican, 5.1% identified as Cuban, 4.8% identified as Dominican, and 5.2% classified in the survey as Other Hispanics. Foreign-born respondents comprised 56% of the sample. A total of 8% (4678) of cases in the sample involved Hispanic provider-patient concordance.
Results
Hispanic patient-provider concordance is statistically significant and positively associated with higher probabilities of seeking preventive care (coef=.211, P<.001), seeking care for a new problem (coef=.208, P<.001), and seeking care for an ongoing problem (coef=.208, P<.001). We also find that the association is not equal across the Hispanic subgroups. The association is lowest for Mexicans in preventive care (coef=.165, P<.001) and new problems (coef=.165, P<.001) and highest for Cubans in preventive care (coef=.256, P<.001) and ongoing problems (coef=.284, P<.001). Results are robust to the interaction of the Hispanic patient-provider concordance for the Hispanic patient categories and being foreign-born.
Conclusions
In summary, racial disparities were observed in health utilization within Hispanic subgroups. While Hispanic patient-provider concordance is statistically significant in associating with healthcare utilization, the findings indicate that this association varies across Hispanic subpopulations. The observations suggest the importance of disaggregating Hispanic racial and ethnic categories into more similar cultural or origin groups. Linked with the existence of significant differences in mortality and other health outcomes across Hispanic subgroups, our results have implications for the design of community health promotion activities which should take these differences into account. Studies or community health programs which utilize generalized findings about Hispanic populations overlook differences across subgroups which may be crucial in promoting healthcare utilization.
