Language matters: Impact of medical interpretation on lung cancer screening adherence.

Abstract

e13752 Background: Landmark trials like NLST, NELSON, DLCST, DANTE, PLCO, and LUSI have shown benefits of lung cancer screening (LCS) in early detection and improved survival. Despite this strong evidence, LCS uptake remains suboptimal, particularly in underserved populations, including individuals with limited English proficiency (LEP). While LEP has been linked to lower screening rates for breast and colorectal cancers, its impact on LCS remains largely unexamined. Our study seeks to evaluate LCS adherence by comparing LEP patients requiring a medical interpreter to patients not requiring an interpreter. Methods: In this retrospective observational study, we identified patients managed by a single English-speaking navigator in the LCS Program at Sylvester Comprehensive Cancer Center who underwent at least one low-dose computed tomography (LDCT) from 1/01/2022 to 4/01/2024. Adherence was defined as completing LDCT(s) within a specific timeframe based on the Lung Imaging Reporting and Data Systems (Lung-RADS) Category (1 or 2: 9-15 months, 3: 3-9 months, 4A: 2-4 months). Patients were categorized into four groups based on their initial Lung-RADS category. Interpreter users were identified from clinic notes. Adherence rate was compared between interpreter users and non-users via chi-square test. Logistic regression was used to assess association between adherence, interpreter use, and other sociodemographic factors (sex, age, ethnicity, race, area poverty rate, area education attainment rate). Area poverty rate and area education attainment rate (bachelor's degree or higher) were defined by mapping the patients’ zip codes to the 2020 American Community Survey data. Results: 556 patients were analyzed, including 78 interpreter users (75 Spanish, 1 Chinese, 1 Vietnamese, and 1 Khmer) and 322 non-users. Mean age was 63.4 years (SD ± 6.48). 83.8% had an initial LDCT with Lung-RADS category of 2, 44.8% were female, 51.8% were Hispanic, 85.1% white, and 52.2% lived in areas with poverty rates of more than 10%. Interpreter users showed a higher LCS adherence rate compared to non-users (80.77% vs. 67.36%, p < 0.01), particularly within the Lung-RADS category 2 group (85.71% vs 69.98%, p < 0.01). Logistic regression showed interpreter use was a positive predictor for adherence (OR = 2.380, p < 0.01), and area poverty rate was a negative predictor (OR = 0.475, p < 0.01). Conclusions: Contrary to previous studies linking LEP to poorer cancer screening outcomes, our findings reveal that LEP patients requiring medical interpreters had higher LCS adherence. This suggests medical interpretation may play a crucial role beyond language translation by enhancing patient education and facilitating better communication, thus improving adherence. These findings underscore the need to explore the broader role of medical interpretation in reducing disparities and improving cancer screening.