Lee Ann Johnson

PhD, RN

East Carolina University

Assistant Professor

NLST (Learn more about this study)

NLST-581

Oct 4, 2019

A Comparative Analysis of Lung Cancer Treatment in a Local and National Sample with Lung Cancer

Background: For individuals with lung cancer, research suggests sociodemographic factors, clinical characteristics, and treatments received may influence disparities and mortality outcomes. The population in eastern North Carolina (ENC) is more diverse than many parts of the country. Published research examining outcomes in lung cancer often does not reflect our regional diversity. For example, in our 29-county catchment area, minorities comprise 63% of the population in 7 counties and 43% in 18 counties. Regional poverty rates are high, with most counties reporting poverty rates between 35% to 50%. 19 counties are federally designated as rural (noncore and micropolitan). Factors that may influence disparities, particularly mortality, in the local population with lung cancer include screening prior to diagnosis, sociodemographic and clinical factors, and treatments received. To better tailor interventions to our local underserved populations with lung cancer, comparisons between factors and outcomes in the local population from ENC need to be compared to nationally representative samples.

Methods: We will use a cross-sectional design to address the aims of this study. Demographic variables, lung cancer treatment received (surgery, chemotherapy, and radiation), and mortality data will be collected. A database for a population from ENC who has been diagnosed with lung cancer will be developed through a review of electronic health records at the Leo Jenkins Cancer Center and will include: (a) demographic variables, (b) lung cancer treatment received (surgery, chemotherapy, and radiation), and © mortality data. Variables for the database of local persons with lung cancer will be matched to those collected and defined in the NLST data dictionary. Demographic variables, lung cancer treatment received (surgery, chemotherapy, and radiation), and mortality data will be requested from the NLST database.

Statistical Analysis: Aims 1, 2, and 3 all involve comparisons between two groups, Whites and Blacks in Aims 1 and 2, and NLST and ENC for Aim 3. All three aims will be assessed using the same statistical methods. For continuous outcomes, e.g. age, independent samples t-tests will be used to compare groups. For categorical outcomes, e.g., treatment received, chi-square tests of independence will be used to compare groups. Mortality will be assessed with continuous time survival models using group as a predictor of survival. Effect sizes, e.g. standardized mean differences or risk ratios will be computed for all analyses.