eP293 - Language Concordance and Colorectal Surgery Outcomes: A Multilingual Perspective in an Urban Community Hospital

Purpose/Background: Limited English proficiency (LEP) has been linked with decreased education, household income, insurance coverage, decreased access to healthcare, and overall poorer health. Perioperative outcomes related to LEP have been published for a wide variety of surgical specialties, but are uncommonly described for colorectal surgery. In our urban community hospital, which serves a diverse patient population with a diverse and multilingual workforce, we hypothesize that language concordance (LC) between patient and provider is a meaningful measurement. We sought to define the frequency and impact of LC in our hospital.Methods/Interventions: This is a retrospective cohort study. All patients undergoing colectomy between 2015 and 2021 at a single institution were included. LC was based on patients’ primary language and operating surgeon’s language proficiencies. LC was categorized as English proficient, LEP language-concordant (LC), and LEP language-discordant (LDC). Statistics were conducted on SAS Academic.Results/Outcomes: A total of 1002 patients were included (53% male). Average age was 64.9 years (SD 15.1). There were 683 elective cases, 230 non-elective cases, and 89 emergent cases. Most common primary languages were English (57.9%, n=580), Mandarin-Chinese (17.6%, n=176), and Spanish (12.3%, n=123). Of patients with non-English language preference, most were language discordant (70.6%, n=298). LC was most common for English (100%) and Mandarin-Chinese speakers (66%) (p < 0.001). Both groups of LEP patients (LC and LDC) had shorter LOS, but LC (-0.4931[95% CI -0.6472, -0.3390]) group had a significant improvement from LDC (-0.2106[-0.3176, -0.1036], p < 0.001). LC patients had the lowest incidence of any 30-day complications (5.6%) compared to LDC (7.1%) and EP groups (12.2%) (X2 8.8, p=0.012). In a multivariable model, language concordance was still significant predictor of shorter LOS (p=0.029) after controlling for age (p=0.11), surgeon (p=0.005), BMI (p=0.67), elective/emergent status (p=0.018), health literacy (p=0.003), indication (p=0.57), ASA class (p=0.92), wound class (p=0.39), discharge destination (p < 0.001), preop ERAS (p < 0.001), any complication (p < 0.001), and ostomy creation (p < 0.001). Due to covariance between language concordance and race (p < 0.001), race was not included in the multivariable model. Group differences included older age for LDC patients (0.043 [0.0092, 0.077], p=0.004), and lowest neighborhood health literacy scores for LC patients (p=0.036). Additionally, English-speakers were an overall sicker cohort with higher ASA class (X2 43.6, p < 0.001), less likely to undergo elective surgery (X2 11.1, p=0.026), and were less likely to be discharged home (X2 25.1, p=0.002). All disparate variables from bivariate analyses were included in the multivariable model.Conclusion/

Discussion: Language concordance between patient and surgeon has not been previously studied in colorectal surgery and may provide improved perioperative outcomes to patients who speak another language. Investigating effects of language concordant care in colorectal surgery may provide new approaches to limit healthcare disparities.