Purpose: To describe patterns of clinical Spanish use by pediatric residents, and to compare self-assessment of language proficiency against an objective language test.
Method: In 2010, the authors e-mailed a survey to all 247 pediatric residents at three institutions, inviting those with any level of Spanish language ability to participate. Participants completed a survey reporting Spanish proficiency, interpreter use, and comfort using Spanish in a range of clinical scenarios. Clinical scenarios were grouped and analyzed by degree of complexity. Self-reported Spanish proficiency was compared with tested proficiency, as measured by a 20-minute telephone assessment of general language ability. Scores were categorized as "not proficient," "proficient," and "highly proficient."
Results: Of the 247 residents, 78 (32%) participated, self-reporting a range of Spanish skills; 23% of those reported spoken proficiency ("proficient" or "fluent"). Participants at all levels of proficiency reported using Spanish without interpretation, including 63% of those who were not proficient. The majority (56%) of nonproficient residents reported comfort using Spanish in straightforward clinical scenarios, and 10% reported comfort in clinical scenarios with legal implications. Self-reported proficiency had a positive predictive value of 67% for testing at a proficient level and 22% for testing at a highly proficient level.
Conclusions: Regardless of level of Spanish proficiency, pediatric residents provide clinical care to patients in Spanish. Self-reported Spanish proficiency does not reliably predict tested ability, especially when using stringent criteria to define proficiency. Provider language "credentialing" is an important step in implementing a policy to improve care for limited English proficiency patients.