Caregiver and Pediatric Clinician Perspectives on Artificial Intelligence for Language Services.

Abstract

Objective: Children and caregivers who use languages other than English (LOE) for pediatric healthcare experience inequities and poorer overall health. Advancement of large language models has raised questions about future use of artificial intelligence (AI) for language access. This study explored perspectives of caregivers who use LOE and pediatric clinicians on using AI to address unmet needs in healthcare translation and interpretation.

Methods: We conducted semi-structured individual interviews with caregivers of pediatric patients who use LOE and pediatric clinicians about the use of AI language technologies in healthcare. A phone interpreter or fluent team member was used for LOE interviews. Recordings were transcribed, translated as needed, coded, and analyzed using inductive thematic analysis.

Results: 20 caregivers using 11 different LOE and 22 clinicians participated. Themes were consistent across participants, though clinicians were more familiar with the concept of AI. Participants reported use of technologies (e.g., Google Translate) for written communication and situations where verbal interpretation was perceived to be inadequate. They were concerned about AI accuracy, privacy, and loss of empathy, but hoped it could provide real-time document translation and more convenient verbal communication. Ease of use, validation, and equitable creation and access were critical for use.

Conclusions: Caregivers and clinicians were open to using AI to fill gaps in translation and interpretation; however, robust validation of AI technology and complementary use with human interpreters and translators is needed. Future research, practice, and policy should focus on integrating AI while investing in human translation and interpretation.