The relative impact of patient and institutional rurality on lung cancer disparities.
e20052 Background: We quantified variation in stage specific, guideline concordant treatment and examined the interaction with rurality and overall survival (OS). Methods: We used tumor registry data for non-small cell lung cancer (NSCLC) patients at 5 institutions in the Mississippi Delta from 2011-2017, including patient demographics, clinical stage, treatment, and OS. We defined rurality by Rural-Urban Commuting Area codes, hospital and patient zip codes; based stage-stratified treatment on National Comprehensive Cancer Network guidelines; used Chi-squared and ANOVA F-tests to assess differences across institutions and logistic regression to assess associations between appropriate care, patient- and institution-level rurality. We used Log-rank tests to examine differences in OS and Cox proportional hazard regression to calculate hazard ratios (HR). Results: 6,259 patients were identified across 2 rural (n = 1255, 20%) and 3 metropolitan (metro) institutions (n = 5004, 80%). There were significant demographic and clinical differences between institutions: proportion of African-Americans (range: 6-37%, p < 0.001), uninsured (3-18%, p < 0.001), patient rurality (17-99%, p < 0.001), ‘no treatment’ rates (17-31%, p < 0.001). Metro patients or those treated at metro institutions were more likely to receive guideline-concordant treatment (odds ratio: 1.34, 95% CI [1.20 - 1.49]; 1.45 [1.28 - 1.65], respectively) than their rural counterparts and had improved OS (HR: 0.89; 95% CI [0.84 - 0.95]; 0.68 [0.63 - 0.72], respectively). They were also less likely to receive ‘no treatment’ (0.62 [0.55 - 0.71], p < 0.001; 0.51 [0.49 - 0.66], p < 0.001, respectively). Among patients with proper care, there were no patient-level rurality based OS differences (p = 0.2203) but those treated at metro institutions had better OS (p < 0.001). When stage-stratified, only advanced-stage patients treated at metro institutions had better survival (p < 0.001), no other differences in OS were detected for early or late-stage patients. Conclusions: Institution-level rurality had greater influence than patient-level rurality on receipt of guideline concordant care and OS. Appropriate care eliminated patient-level rurality OS disparities.