100 - Influence of Preferred Language on Time to Diagnosis of Pediatric Leukemia

Publication Number: 100.5130

Rachel Kim, UCSF Benioff Children's Hospital San Francisco, San Francisco, CA, United States; Eleanor Kim, UCSF Benioff Children's Hospital San Francisco, San Francisco, CA, United States; Maria I. Castellanos, University of California, San Francisco, School of Medicine, San Francisco, CA, United States

Background: Survival disparities of acute leukemia among children and adolescents from marginalized communities or socioeconomically disadvantaged backgrounds are well documented. However, there is limited knowledge about how social determinants of health (SDoH) influence and possibly prolong time to diagnosis of leukemia. A better understanding of how SDoH interact with timely diagnosis may aid in addressing survival disparities. In particular, families who use a language other than English may experience difficulties navigating the health care system due to language and communication barriers.

Objective: In this study, we examine how preferred spoken language influences time to diagnosis from symptom onset of pediatric leukemia. We hypothesize that families who use a language other than English experience longer time to diagnosis.

Design/Methods: We conducted a retrospective chart review of pediatric patients diagnosed with acute leukemia at two urban academic pediatric hospitals between 2013-2023. We collected SDoH data including preferred spoken language, insurance type, and country of birth. Patients were then subdivided into groups: English or other language. We collected clinical data including symptom characteristics, and physical and diagnostic findings at diagnosis. Time from symptom onset to diagnosis of leukemia was calculated as the difference between the earliest reported date of symptoms attributable to leukemia and the date of the encounter in which leukemia was diagnosed. Analyses were conducted with Mann-Whitney U tests and chi-squared tests on R.

Results: Our study included 282 patients, of which 197 (70%) spoke English (Table 1). The second most common language was Spanish (29%), followed by Cantonese and Mandarin. Median time to diagnosis from symptom onset was 20.5 days with interquartile ranges 7-34.8 days. There was no significant difference in median time to diagnosis between patients whose families speak English compared to those who speak another language (21 vs 14 days, p=0.7, Figure 1). There was no significant difference in number of medical encounters prior to diagnosis (p=0.6) and presence of tumor lysis syndrome on diagnosis of leukemia (p=0.1) based on primary language spoken.

Conclusion(s): While some studies identify concerns for longer time to diagnosis in populations with language barriers, we found no significant difference based on primary language in our study population. This likely reflects that other SDoH may play a stronger role in explaining disparate outcomes. Further studies are needed to determine the association between other SDoH and their impact on time to diagnosis.

Table 1
Table 1.pdfDemographic, social determinants of health, and diagnostic encounter data based on primary language. p-values calculated based on comparison of patients who speak English versus patients who speak a language other than English.

Figure 1
Figure 1.pdfBox plot of median time to diagnosis from symptom onset based on primary language.

Table 1
Table 1.pdfDemographic, social determinants of health, and diagnostic encounter data based on primary language. p-values calculated based on comparison of patients who speak English versus patients who speak a language other than English.

Figure 1
Figure 1.pdfBox plot of median time to diagnosis from symptom onset based on primary language.