404 - Lost in Translation? Association between patients' preferred language and wait times in the pediatric ED

Publication Number: 404.5669

Taylor E. Anderson, Monroe Carell Jr. Children's Hospital at Vanderbilt, Nashville, TN, United States; Donald H. Arnold, Vanderbilt University School of Medicine, Nashville, TN, United States; Adriana Bialostozky, Monroe Carell Jr. Children's Hospital at Vanderbilt, nashville, TN, United States; Barron Frazier, Monroe Carell Jr. Children's Hospital at Vanderbilt, Nashville, TN, United States

Background: Providing equitable care is a priority of pediatric emergency medicine. Patients whose preferred language differ from that of their care team may experience disparities in access and quality of care.

Objective: This study aimed to investigate associations between patients’ preferred language and wait times to see a provider in the pediatric emergency department (ED). The primary outcome was the time from patient rooming to initial provider evaluation, and secondary outcomes included time intervals from ED arrival to triage, ED arrival to rooming, ED length of stay (LOS), and associations between other demographic data including sex, race, and insurance status.

Design/Methods: This was a retrospective study of patients arriving to the ED from January 1st, 2022 to December 31st, 2023. Data from 111,095 patient charts were reviewed, and 84,200 patients were included in multivariable analyses. Exclusion criteria included age over 18 years and patients with incomplete data. Bivariate comparisons were made using Wilcoxon Rank Sum or Kruskal-Wallis tests. Multivariable models were used to provide adjusted associations between relevant covariates and outcomes of interest. Analyses were performed using STATA/IC v14.2.

Results: Patient demographics of all charts reviewed are reported in Table 1. Fifty-three percent of the cohort identified as white, with 77 different preferred spoken languages, and 81% of patients preferring English. Bivariate analyses indicated significant differences in wait time from rooming to initial provider evaluation for sex, race, and language (p < 0.05) (Table 2). Multivariable linear regression including the covariates age, sex, race, primary payer, language, and Emergency Severity Index score showed patients preferring non-English language compared to English-speaking patients waited an adjusted average of 1.1 minutes longer to begin triage [95% CI 0.7, 1.5], 5.7 minutes longer to be roomed [95% CI 4.6, 6.9], and 3.8 minutes longer for first provider evaluation after rooming [95% CI 3.3, 4.2] (p < 0.001). Notably, patients identified as non-white had shorter average LOS by 16.6 minutes [95% CI 13.3, 19.9] (p < 0.001) (Table 3).

Conclusion(s): These findings underscore the necessity of recognizing and accommodating patients’ preferred language to ensure equitable care in a pediatric ED. Although the differences in wait times are small on the individual patient level, these differences over a large cohort of children highlights a potential disparity in the quality of care provided to children from diverse backgrounds or who prefer other languages than English.

Demographic characteristics of patients aged 0 to 18 years who arrived to the pediatric ED from January 2022 through December 2023 with time to event data
Time intervals reported in minutes with median and interquartile range [IQR]. Abbreviations: ESI, Emergency Severity Index; LOS, length of stay.

Bivariate analyses of time interval from patient placement in room to initial provider evaluation
Time intervals reported in minutes with median and interquartile range [IQR]. Abbreviation: ESI, Emergency Severity index. *Wilcoxon rank sum test. **Kruskal-Wallis test.

Multivariable linear regression analyses of adjusted patient wait times and multivariable logistic regression analyses of adjusted odds ratios of admission and leaving without being seen by ED provider
Wait times are reported in minutes as beta coefficient and 95% confidence interval [CI]. Logistic regression analyses are reported as adjusted odds ratio with 95% CI. Values in bold are statistically significant with p-value <0.05. Abbreviations: LOS, length of stay; LWBS, left without being seen.