Language Proficiency and Survival in Pancreatic Cancer: a Propensity Score–Matched Analysis

Abstract

BackgroundLimited English proficiency has been shown to negatively affect health outcomes. However, as of now, little is known about survival rates of patients with limited English proficiency (LEP) and pancreatic ductal adenocarcinoma (PDAC) when compared to patients with English proficiency (EP) in an urban, non-safety net setting. We aimed to compare survival rates between patients with LEP and those with EP who had a diagnosis of PDAC. MethodsA single-institution retrospective propensity-matched cohort study of patients with biopsy-proven PDAC was undertaken. Demographics, clinical characteristics, and language information were collected for all participants. Patients were classified as having LEP or EP based on their preferred speaking language at the time of admission and matched on baseline characteristics using propensity scores. Survival analysis methods were used to study survival rates in patients with PDAC based on their EP status. ResultsOf 739 included patients, 71 (9.48%) had LEP, mean age was 68.4 ± 10.9, and 51.8% were female. Both groups of patients were comparable for age, gender, marital status, and time to treatment. LEP status was associated with higher odds of death in both unmatched (HR 1.65, 95% CI 1.22–2.22) and matched (HR 1.60, 95% CI 1.03–2.47) analyses. Additionally, patients with LEP had significantly decreased odds of receiving cancer-directed treatment and increased odds of advanced stage cancer at presentation. ConclusionsIn this cohort of patients with PDAC, LEP predicted worse survival. The results of this study suggest that, after accounting for interpreter use, other factors contribute to this disparity. Such factors, as yet unmeasured, may include health literacy and cultural expectations, for which further investigation is warranted to better understand and limit this survival disparity.