What Does It Mean to Speak a Patient’s Language?
Every day in multilingual clinics across the United States, physicians communicate with patients in languages that their health systems have never formally evaluated. A clinician may feel confident holding a conversation in Spanish or Mandarin, yet confidence does not always translate into clinical accuracy. For patients, the risks of miscommunication are significant. Misunderstandings can affect treatment plans, adherence, informed consent, and the overall experience of care. These stakes have made a longstanding question more urgent: how should health systems verify that clinicians are truly proficient in the languages they use with patients?
One established tool is the Clinician Cultural and Linguistic Assessment, known as the CCLA. Originally developed by Kaiser Permanente and now administered by a commercial testing company, the CCLA is a standardized oral exam that simulates a primary care encounter in a target language. It evaluates grammar, register, medical vocabulary, fluency, pronunciation, and cultural rapport. Clinicians who pass the exam are certified to provide language-concordant care without an interpreter. In theory, this strengthens equity. In practice, many physicians feel the test does not capture the nuances and emotional complexity of real clinical communication. Some find it stressful, while others describe it as disconnected from the back-and-forth problem-solving that characterizes most medical visits.
A recent study published in The Joint Commission Journal on Quality and Patient Safety offers a possible alternative. The researchers asked whether direct observation of real clinical encounters might provide a more authentic evaluation of non-English language skills. Instead of judging clinicians through a one-time exam, this approach looks at how they actually communicate with patients in the moment. The study also explored how clinicians feel about being evaluated through observation.
The research was part of the Language Access System Improvement project, or LASI, which focuses on strengthening communication in multilingual health systems. Guided by an advisory board of patients, clinicians, and community advocates, the team recruited eleven general internists who care for adults aged eighteen and older. The group reflected a wide linguistic range. Five physicians reported proficiency in Mandarin, four in Spanish, and one each in Cantonese and Russian. Nine of the eleven participants were women. They were categorized into two proficiency groups. Five clinicians had full proficiency, meaning they had previously passed the CCLA. Six had partial proficiency, meaning they described their skills as good or very good but had not taken or passed the CCLA.
Researchers conducted thirty to forty-five minute semistructured interviews that asked clinicians to reflect on their experiences with language assessment and to share their opinions of direct observation. The interviews were recorded, transcribed, and analyzed thematically.
One of the strongest qualitative themes was familiarity. Many clinicians noted that observation already plays a key role in medical training and evaluation. Residents are observed routinely, and peer review remains part of practice. Because of this, several participants viewed observation of language skills as a natural extension of existing norms. They also felt that real encounters reveal aspects of communication that standardized tests miss, such as tone, pacing, cultural connection, and the ability to adjust explanations based on patient cues.
Yet, discomfort surfaced as well. Some clinicians worried that having an observer in the room could make patients uneasy, particularly during sensitive conversations. Others said they might become self-conscious or distracted under scrutiny. Many raised concerns about fairness. A single observed interaction might not represent true proficiency. A routine follow-up could make any clinician seem fluent, while a particularly complex case might make a skilled speaker appear unsure. For this reason, clinicians recommended multiple observations or recorded encounters that could be reviewed more thoroughly.
Participants also offered ideas for improving the assessment process. They suggested clearer evaluation criteria, training for observers in both linguistic and cultural competence, and combining observation with other verification tools rather than relying on it alone.
The implications of this work extend beyond the study site. Direct observation has potential because it captures the lived realities of multilingual care. At the same time, it raises questions about bias, patient comfort, clinician workload, and the emotional pressures of being evaluated. The study highlights a gap between the practical need for accurate language assessment and the limitations of current tools to do so. This article asserts that communication in medicine is relational and deeply tied to trust, and any method of assessing language proficiency must reflect that responsibility.