Abstract
<title>Abstract</title> <p> <bold>Purpose.</bold> Socioeconomic survival disparities in glioblastoma (GBM) are established, but whether they operate through receipt of surgical resection has not been quantified. We decomposed the county income–based disparity into the portion transmitted through resection and the portion arising from interaction with it. <bold>Methods.</bold> Population-based Surveillance, Epidemiology, and End Results (SEER) 17-registry cohort: adults aged 18–64 with histologically confirmed GBM (ICD-O-3 9440/3) diagnosed 2007–2022. Exposure: county-level median household income, lowest (Q1) versus highest (Q4, reference) quartile. Mediator: receipt of surgical resection versus none/biopsy. Outcome: Overall Survival. We applied the VanderWeele four-way decomposition with a Cox estimator and an exposure–mediator interaction, adjusting for age, sex, diagnosis year, race/ethnicity, marital status, and rurality (1000 bootstrap replications). Sensitivity to unmeasured confounding used E-values. <bold>Results.</bold> Among 19,051 patients, median overall survival rose from 12 (Q1) to 15 months (Q4), whereas resection was near-identical (Q1 66.0%, Q4 68.9%). The lowest-income quartile carried a higher hazard of death (total-effect hazard ratio [HR] 1.23, 95% CI 1.17–1.30). The controlled direct effect persisted (HR 1.19, 95% CI 1.13–1.25). An income–surgery interaction accounted for 26.3% of the total effect (proportion eliminated 28.0%, 95% CI 10.9–44.6%). Surgery mediated almost none of it (proportion mediated 2.6%, 95% CI − 2.7 to 7.6%; pure indirect HR 1.00). E-values were 1.58 (total) and 1.50 (direct). <bold>Conclusion.</bold> The income-based survival disparity in GBM is substantial, but it does not flow through receipt of surgery. It acts on surgery instead. Equalizing who receives surgery would not close the gap. Moderate unmeasured confounding could explain the direct effect. </p>