Disparities across demographic, behavioral, and socioeconomic factors in lung cancer stage at diagnosis: An EPIC EHR study from the University of Maryland Medical System

Background: Lung cancer is the leading cause of cancer mortality in Maryland and the US with sex, socioeconomic factors, and insurance status disparities in stage at diagnosis. There is a rising burden among younger adults, never-smokers, women and racial/ethnic minorities, who often present with advanced disease and may not qualify for screening using current USPSTF guidelines. In this study, we evaluate the associations between age, smoking status, race, sex, and insurance status, and the stage at diagnosis using large real-world EHR datasets. Methods: We conducted a retrospective cross-sectional study of 6,446 adults diagnosed with lung cancer (ICD-10: C34) using EPIC electronic medical records from the University of Maryland Medical System between January 2016 and December 2024. We extracted age, sex, race, ethnicity, smoking status, insurance type, and stage at diagnosis classified as early (I-II) or advanced (III-IV) from the records. We used t tests for continuous variables and χ² tests for categorical variables. Results: Overall, 38% of patients were diagnosed with early-stage lung cancer and 62% with advanced disease. Advanced stage disease was commoner among younger patients, (mean age 68 vs 70 years, p<0.001), and those younger than 50 years (77%, p<0.001). Though there were more women (52%) in the study population, men were more likely to present with advanced disease (64% vs 59%, p<0.001). The study population was predominantly white (73%) but there was no association between race and stage (p=0.06). Current smokers were most likely to present with advanced disease (66%, p<0.001). Patients on Medicaid (78%) and uninsured individuals (65%) had a higher likelihood of advanced-stage diagnosis compared with those covered by Medicare (58%) or private insurance (66%). Conclusions: Younger adults, males, current smokers, and patients on Medicaid or having no insurance were more likely to present with advanced lung cancer. These findings highlight disparities in lung cancer diagnosis and underscore the need for further research.