Differences in Cancer Stage at Diagnosis Between Patients Presenting at Rural vs. Urban Facilities: A Rapid Systematic Review and Meta-analysis

Abstract

Background: Rural populations experience disproportionately higher cancer mortality in the United States. One hypothesized mechanism is a later stage at diagnosis linked to reduced screening uptake, longer travel times, and constrained diagnostic capacity. Quantifying the rural–urban gap in the diagnostic stage can guide equitable screening and referral strategies.

Objective: To quantify rural–urban differences in stage at cancer diagnosis through a rapid systematic review and meta-analysis, estimating pooled adjusted odds ratios (aORs) for late stage (III–IV or distant) where comparable data permit. Secondary aims were to characterize heterogeneity by cancer site (screen-amenable vs. non-screen-amenable), rurality definition (residence vs. facility; RUCC/RUCA/Census), socioeconomic context (e.g., poverty, insurance), and geography; to perform sensitivity analyses (model specification, risk-of-bias strata); and to contextualize findings for Arkansas using state registry indicators to inform targeted screening and diagnostic access initiatives.

Methods: We conducted a rapid systematic search of PubMed/PMC and major journals (2000-September 29, 2025) for observational studies comparing rural and urban groups (by residence or facility context) that reported cancer stage at diagnosis. The preferred effect measure was the adjusted odds ratio (aOR) for late-stage disease (III–IV or distant). Two reviewers extracted study characteristics, rurality definitions (e.g., RUCC/RUCA), covariate adjustments, and effect estimates; risk of bias was qualitatively assessed. A random-effects meta-analysis was planned; however, due to heterogeneous definitions and incomplete reporting of variances, we present a quantitative range of comparable aORs with a forest plot of studies providing extractable 95% CIs, and otherwise provide a narrative synthesis.

Results: Of 312 records identified, 67 full texts were assessed; 9 studies were included in the qualitative synthesis, and 5 contributed extractable aORs. Across multi-cancer registries and site-specific analyses, rural patients had modestly higher odds of late-stage diagnosis, especially for screen-amenable cancers (breast, colorectal), with typical adjusted effects ~1.1–1.3. Arkansas registry data align with this pattern: rural counties had higher rates of late-stage colorectal cancer (AAIR: 21.9 per 100,000; 95% CI, 19.4–24.4). Heterogeneity reflected varying rurality metrics, staging bins, and covariate sets.

Conclusions: Rural patients tend to be diagnosed at slightly later stages, a small but meaningful disparity at the population scale. Interventions that increase screening uptake, accelerate diagnostic work-ups, and reduce financial/transport barriers with county-level targeting in rural populations, such as in Arkansas, are likely to narrow rural–urban gaps.