Language access in healthcare is often evaluated by whether an interpreter is available, yet interpreter presence alone does not establish that a patient’s account has been fully expressed, accurately conveyed, or incorporated into clinical interaction. Interpreter-mediated encounters can support participation while also redistributing conversational control, confidentiality, relational continuity, and dependence. To synthesize qualitative evidence on how patients, caregivers, interpreters, clinicians, and other relevant actors experience interpreter-mediated healthcare, with particular attention to patient voice, moral agency, relational trust, and the contextual conditions that shape these experiences. A qualitative evidence synthesis using meta-ethnographic principles was undertaken. Searches conducted on 22 September 2026 identified 90 records through five structured search runs and supplementary citation chaining. After removal of 26 duplicates, 64 unique records were screened; 32 reports were sought, 26 were available for full assessment, and 22 primary qualitative evidence units met the eligibility criteria. Included evidence was appraised using qualitative methodological domains without numerical scoring. First-order participant accounts and second-order author interpretations were preserved separately before reciprocal and refutational translation. Confidence in final synthesis findings was planned using GRADE-CERQual. The included literature showed that interpreter-mediated care was not experienced as a single, uniform intervention. Meaning was shaped by interpreter accessibility, continuity, dialect and linguistic fit, confidentiality, modality, clinician language capability, family involvement, time pressure, organizational workflow, and the extent to which interpreted speech preserved clinically and personally significant detail. Professional interpretation often supported understanding and participation, but residual problems remained when patients perceived information to be condensed, insufficiently contextualized, delayed, or mediated through relationships they did not fully control. Conversely, informal or partial language concordance could sometimes support rapport while introducing other risks. Interpreter-mediated care can enable patients to be heard, but the ethical significance of interpretation depends on whether language support preserves meaningful expression, supports rather than substitutes for patient participation, and operates within trustworthy and usable institutional arrangements. The evidence supports a context-sensitive account of mediated agency rather than a binary distinction between interpreted and uninterpreted care.
