(1361) Using Artificial Intelligence to Promote Equitable Care for Complex Patients With Language Barriers

Introduction: Healthcare disparities among populations with language barriers can be partially mitigated by leveraging interpreter skills to reduce cultural, language and literacy barriers through improved bidirectional communication between patients/ families and their clinicians. Federal mandates entitle patients with language barriers to interpreter services. Much evidence supports the use of professional interpreters in all interactions with these patients, especially those with complex care or critical illness needs. However, due to interpreter shortages and clinician underuse of interpreters, patients with language barriers frequently do not get the language services they need or are entitled to.

Methods: Our study is developing, testing, and implementing an artificial intelligence (AI) algorithm to identify hospitalized patients with language barriers and complex care needs to increase the use of in-person interpreters and reduce the time to interpreter use. We conducted semi-structured interviews among stakeholders including intensive care and hospitalist physicians, advanced practice providers, nurses, and interpreters about their perceptions of the AI tool, potential risks and benefits, and suggestions to integrate the AI tool into clinical practice. Data was managed, coded, and analyzed using NVIVO software.

Results: We completed 49 interviews. Perceived benefits for the use of AI included to highlight the importance of in-person interpreters for this population; enhance the prioritization of interpreter use; and streamline the process for accessing interpreters. Perceived risks included the potential for feelings of stigmatization among patients; alert fatigue related to suboptimal integration of AI into the clinical workflow; and supply-demand issues. Participants contributed suggestions and preferences for integration into workflow that will be used during implementation.

Conclusions: Overall participants perceived more benefits than risks to the use of AI in this domain, some of which overlap with known hazards and values of AI but some unique and novel. The use of AI to identify and prioritize patients with language barriers for interpreter services was seen as valuable with the potential to improve standard of care and address healthcare disparities.