Association of non-English language preference with postoperative hospital visits among California Medicaid enrollees

• Published in: Surgery 2024-12, Vol.176 (6), p.1711-1720
• Main Authors: Blegen, Mariah B., Zingmond, David S., Jackson, Nicholas J., Torres, Jesus R., Russell, Tara A., Maggard-Gibbons, Melinda, Russell, Marcia M.
• Format: Article
• Language: English
• Subjects: Adolescent; Adult; Aged; California; Communication Barriers; Emergency Service, Hospital - statistics & numerical data; Female; Humans; Language; Male; Medicaid - statistics & numerical data; Middle Aged; Patient Preference - statistics & numerical data; Patient Readmission - statistics & numerical data; Retrospective Studies; United States; Young Adult
• ISSN: 0039-6060; 1532-7361; 1532-7361
• DOI: 10.1016/j.surg.2024.08.016

Patients with non-English language preference encounter language barriers across phases of surgical care. Patients with a non-English language preference represent 35% of California households and are disproportionately insured by Medicaid. To determine whether language predicts surgical outcomes, we investigated the association of patient non-English language preference with postoperative emergency department visits and readmissions among California Medicaid enrollees. Our retrospective analysis of adult Medicaid enrollees undergoing 1 of 10 common inpatient operations using California hospital administrative data (2016–2019) modeled the association between non-English language preference and 30-day postoperative emergency department visits and readmissions using mixed effects logistic regression with hospital random intercept, adjusting for patient, operation, hospital, and community characteristics. Secondary analyses stratified by operation urgency and by insurance type in an all-payor cohort. Of 115,527 Medicaid enrollees, 17.2% had non-English language preference (n = 19,881), 66% were female (n = 73,653), and 40% were Hispanic/Latino (n = 45,541). Patients with non-English language preference experienced fewer postoperative emergency department visits (non-English language preference: 13.5%, English preference: 17.9%, P < .001) and readmissions (non-English language preference: 7.5%, English preference: 8.5%, P < .001), which persisted in adjusted models (adjusted odds ratio emergency department, 0.80, 95% confidence interval, 0.77–0.85; readmissions: adjusted odds ratio, 0.86, 95% confidence interval, 0.80–0.92). Non-English language preference was associated with fewer emergency department visits after elective (adjusted odds ratio, 0.80; 95% confidence interval, 0.73–0.88) and urgent/emergent surgery (adjusted odds ratio, 0.80; 95% confidence interval, 0.75–0.85) but not readmissions after elective surgery (adjusted odds ratio, 0.89; 95% confidence interval, 0.78–1.01). This pattern was only observed for Medicaid and not other insurance types. Patients with non-English language preference who receive Medicaid have fewer postoperative emergency department visits and readmissions, even after urgent surgery. Our findings suggest that patterns of health care seeking after surgery vary by patient language, and investigating explanatory mechanisms is needed. [Display omitted]