PROJECT SUMMARY/ABSTRACT Lung cancer is the leading cause of cancer death in the United States. Advances in lung cancer treatment have substantially improved survival. However, the benefit has not reached all racial/ethnic groups of patients equally. Compared with non-Hispanic European Americans (EA), non-Hispanic African Americans (AA) are disproportionately affected by lung cancer with higher incidence and inferior survival. Their poorer outcomes are largely driven by more advanced stages at diagnosis and underutilization of stage-appropriate treatment. Importantly, AA and EA patients with lung cancer can achieve comparable outcomes under similar treatment modalities. This strongly suggests that barriers to cancer care are central to outcome disparities, and strategies targeting specific barriers will be critical to reduce long-standing disparities. However, studies addressing the underlying mechanisms of lung cancer treatment disparities have focused on non-modifiable and much less modifiable factors. There is an urgent need to elucidate modifiable factors influencing lung cancer treatment in AA patients. As Penchansky proposed, healthcare access consists of five distinct dimensions, including affordability, accommodation, acceptability, availability, and accessibility. However, most of previous studies assessed access to lung cancer care based on insurance coverage and availability of providers, and other dimensions of access remain critically understudied. We found that insurance coverage and availability of cancer care collectively explained <50% of the excess risk of underutilization of guideline-concordant treatment in AA vs EA lung cancer patients. Thus, we hypothesize that excess risks of underutilization of guideline- concordant lung cancer care and mortality in AA vs EA patients are attributable to access barriers AA patients disproportionately experience. To test this novel hypothesis, we will develop an integrated database with AA and EA patients diagnosed with non-small cell lung cancer, primarily including data from the longitudinal SEER-Medicare database, national annual surveys of population-based samples of Medicare enrollees, a nationwide database of providers, and neighborhood contextual measures. Using advanced spatial statistical modeling to account for clustering within providers and neighborhoods, we will simultaneously assess five access dimensions in association with lung cancer care and outcomes (Aim 1), examine racial differences in access dimensions overall and by indicators of social disadvantage (Aim 2), and further quantify the independent and collective contributions of access dimensions to racial disparities in lung cancer care and outcomes (Aim 3). This will be the first population-based study to comprehensively assess the impacts of all five access dimensions on lung cancer treatment and their contributions to lung cancer disparities. The results will provide novel insights into which specific components of ac...