The service patterns of a racially, ethnically, and linguistically diverse housestaff.

PURPOSE: To explore whether racial and ethnic concordance in the service patterns of residents is as disproportionate as it is among practicing physicians, and to examine the effect of residents' second-language proficiencies on these patterns.
METHOD: The authors conducted a cross-sectional study of billing data from 13,681 patient visits to resident continuity clinics at a Northern California pediatric medical center between May 1998 and October 1999. For each racial or ethnic group of patients, the proportion of patient visits to race-concordant residents was compared with the proportion of visits to all residents of other races or ethnicities. Multivariate logistic regression analyses (adjusted for confounders and residents' second-language proficiency) were used to measure the concordance of race or ethnicity in patient-resident pairs.
RESULTS: For all visits made to African-American, Asian, and Latino residents, the percentage of race-concordant visits exceeded the percentage of race-discordant visits for each patient group. In adjusted regression models, African-American, Asian, and Latino patients were more likely to visit residents of the same race or ethnicity. White patients were not more or less likely to visit white residents. When adjusted for residents' second-language proficiencies, Latino patients remained more likely to visit Latino residents and Asian patients remained more likely to see Asian residents. In analyses restricted to residents reporting fluency in Spanish, Latino patients were more likely to visit Latino residents.
CONCLUSION: African-American, Asian, and Latino pediatric residents disproportionately served more patients from their own racial or ethnic backgrounds. This service pattern was not completely explained by physicians' second-language proficiencies. A resident's race or ethnicity may reflect a unique set of skills that is highly valued by patients or health care systems.