Lihui Liang, Jiancheng Huang, Qingbai Jiang
IntroductionWe evaluated whether county-level urbanicity at diagnosis was associated with overall survival (OS) across 19 solid tumor groups in SEER.MethodsThis retrospective, population-based cohort study included adults aged ≥18 years with first primary invasive solid tumors diagnosed during 2004-2023. County-level urbanicity at diagnosis was defined using the SEER-linked, year-specific Rural-Urban Continuum Code (RUCC) as nonmetropolitan (NM; RUCC 4-9), small/medium metropolitan (SMM; RUCC 2-3; metropolitan areas with <1 million residents), or large metropolitan (LM; RUCC 1; metropolitan areas with ≥1 million residents). Cancer-specific overlap-weighted Cox models estimated mortality hazard ratios (HRs), complemented by 36-month OS, restricted mean survival time (RMST), and sensitivity analyses.ResultsThe cohort included 5,486,535 patients: 664,872 in NM, 1,571,092 in SMM, and 3,250,571 in LM areas. In the primary adjusted model, SMM and LM residence were associated with significantly lower all-cause mortality than NM residence in 13 and 14 of 19 tumor groups, respectively, after false discovery rate (FDR) correction. For lung cancer, the HRs were 0.97 (95% CI, 0.96-0.98) for SMM versus NM and 0.93 (95% CI, 0.92-0.95) for LM versus NM; neither contrast was significant for esophageal cancer. Direct SMM-versus-LM comparisons differed significantly in 8 tumor groups, all favoring LM. Absolute differences were modest; the largest 36-month OS difference was 3.59 percentage points (stomach cancer, LM vs NM), and the largest 36-month RMST difference was 1.26 months (other digestive system/peritoneal tumors, LM vs NM).ConclusionsCompared with NM residence, SMM and LM residence were associated with more favorable OS in multiple solid tumor groups, with associations observed more broadly for LM residence. These observational findings do not establish causal effects of residence.