Beyond development: Global disparities in gastrointestinal cancer outcomes across 176 countries.
BACKGROUND Mortality-to-incidence ratio (MIR) - defined as the age-standardized mortality rate divided by the age-standardized incidence rate - has been proposed as an ecological proxy for cancer treatment access at the population level. We characterized MIR for four gastrointestinal (GI) cancers across 176 countries and quantified the Surgical Access Gap (SAG), defined as the residual between observed and human development index (HDI)-adjusted expected MIR. METHODS We conducted a cross-sectional ecological analysis using GLOBOCAN 2022 data for colorectal, gastric, esophageal, and pancreatic cancers across 185 countries. MIR was calculated as the age-standardized mortality rate (ASMR) divided by the age-standardized incidence rate (ASIR). HDI-adjusted expected MIR was estimated using restricted cubic spline (RCS) regression (3 knots). SAG was defined as observed minus expected MIR. Spearman correlations between MIR and HDI and UHC Service Coverage Index were calculated. Exploratory cluster analysis was performed using k-means (k = 4, z-score standardized variables). WHO Mortality Database data (1994-2022, 140 countries) provided independent mortality trend characterization. GBD 2023 cross-source agreement analysis confirmed GLOBOCAN 2022 MIR estimates for colorectal (Spearman ρ=0.811) and gastric cancer (Spearman ρ=0.510); agreement was weaker for esophageal and pancreatic cancers.Results of 185 eligible countries, 176 had calculable MIR values. HDI explained 76.4% of colorectal cancer MIR variance (R²=0.764). HDI was strongly inversely associated with MIR for colorectal (ρ=-0.878) and gastric cancer (ρ=-0.839; all p < 0.001). Japan demonstrated consistently negative SAG across all four cancer types (colorectal SAG=-0.075; gastric SAG=-0.348), indicating outcomes substantially better than predicted by HDI. The United Arab Emirates, Qatar, and Saudi Arabia demonstrated positive SAG across all four cancer types, indicating outcomes worse than expected. Exploratory cluster analysis identified four country groupings; Japan, the Republic of Korea, Belgium, and Guyana formed the smallest cluster (n = 4) with the most favourable SAG profiles, although Guyana represented a borderline assignment. CONCLUSIONS MIR-based analysis across 176 countries identified substantial global heterogeneity in GI cancer outcomes beyond what HDI predicts. Japan's consistently negative SAG and the paradoxically positive SAG of several high-income countries suggest that factors not captured by composite development indices may be associated with population-level GI cancer outcomes. These hypothesis-generating findings may inform future health system research priorities.