Information Sharing by Caregiver Language in Pediatric Serious Illness Care Conferences

Abstract

Background: Children using a language other than English are known to experience poorer medical care, including at end-of-life. Information sharing is essential to achieve optimal communication outcomes, but little is known about effects of caregiver language. Objective: We explored differences in quantity and complexity of bidirectional information sharing in English and interpreted care conferences for seriously ill children. Design/Methods: Care conferences for pediatric patients with serious illness from linguistically diverse families at a single institution were audio-recorded, transcribed, and analyzed using content analysis. Provider and caregiver information-sharing statements were characterized as biomedical, logistical, or psychosocial; caregiver categories also included patient wellness and values/goals. Provider communication practices known to impact information sharing were also tabulated. Mann-Whitney-U tests compared median quantity of information shared by conference language. Results: We analyzed 29 care conferences from 2018-2021, including 11 (39%) professionally interpreted conferences. On average, providers made 156 statements per conference while parents made 71 (p<.001). Providers made significantly fewer biomedical statements in interpreted conferences than English ones (117 vs 176, p=0.01). Caregivers made a similar number of statements per conference, regardless of language (78 vs 66, p=.95). Meanwhile, use of provider practices supporting information sharing, such as gathering information preferences, were rare. Conclusions: Providers spoke far more than caregivers, and rarely used practices that support information sharing, regardless of language. They also shared significantly fewer biomedical statements in interpreted conferences than English ones. Strategies to reduce disparities in information sharing may support equitable outcomes for patients and families from all language backgrounds.