Treatment patterns, social determinants of health, and functional outcomes for older adults in supportive oncology.

Abstract

e18575 Background: Research suggests that disparities in all-cause cancer survival times include a dynamic interplay of social determinants of health (SDOH) and treatment decisions. The social environment and associated vulnerability to interpersonal factors, access to social resources and supports, socialization and socioecological experiences affect health outcomes. The purpose of this study was to examine the nexus of SDOH, treatment decisions, and functional outcomes for older adults in supportive oncology at a large academic medical center. Methods: After obtaining IRB approval, we conducted a retrospective chart review of supportive oncology older adults from 2014-2017 (n = 4495) using data restricted to ICD10 codes for solid tumors. We conducted a retrospective analysis to investigate SDOH and treatment decisions for the older adult cancer cohort. Demographic variables included patient age, race, sex, medical insurance, primary diagnosis, attending physician specialty, and comorbidities. Survival time was defined as time from admission to event of death or discharge. We used zip code median income as a surrogate for individual level social gradients. The data were analyzed using the Statistical Analysis Software (SAS) application version 9.4. We used logistics regression and Cox proportional hazards models to answer research questions. Results: The sample (n = 4495) of adult cancer patients was 50% female, 51% white, and mostly (81%) high SES position. The mean age was 66.24 years (SD = 15.23). Treatment patterns included 596 (14%) chemotherapy, 402 (9%) radiation, 415 (9%) both chemotherapy and radiation; and 3015 (68%) reported not having any treatment. Median survival time was 12.88 days. Over 95% of sample reported multiple comorbidities while 76% reported enrollment in government insurance. The crude association between treatment decisions and social position was 1.64 (CI 1.38,1.95) compared to the bias adjusted association which ranged from 2.60 to 4.54. However, the hazard of mortality between those with no cancer treatment was HR = 1.03 (0.96, 1.11) times the corresponding hazard among those receiving treatment. Conclusions: Although supportive oncology clinics may decrease symptom-burden, they may add to fragmentation of care if multiple personnel are involved. Our findings provided evidence that high socially positioned older adults were 4.54 times more likely to receive treatment compared to low socially positioned older adults although there was no meaningful difference in survival times. Clinicians need to create interventions to reduce treatment disparities. Further research is warranted to foster policy models to improve health equity and promote positive functional outcomes for older adults with cancer.