Insights into Cancer Modelling: Analysis of Some Cancer Parameters in Romania for Public Health

Abstract

Background: Cancer represents a major public health burden in Romania. Using GLOBOCAN 2020 estimates, this study provides an updated empirical examination of the relationships among cancer incidence, mortality, rank, cumulative risk, and selected age-standardised indicators, with the aim of providing an empirical quantitative description of the Romanian cancer profile and comparing selected characteristics with global patterns. Methods: GLOBOCAN 2020 estimates for Romania were systematically analysed across 35 cancer sites. The relationships between the number of new cases (Nc) and the number of deaths (Nd) were investigated using linear and log-transformed/power-law regressions. Additional empirical models were applied to describe the dependence of incidence and mortality on cancer rank through a derived severity parameter, and selected results were compared with World 2020 estimates. Results: Romania recorded an estimated 98,886 new cancer cases and 54,486 cancer-related deaths in 2020. Lung cancer ranked first for both incidence and mortality, whereas breast and prostate cancers showed high incidence with comparatively lower Nd/Nc ratios. A linear relationship constrained through the origin yielded Nd=0.5565Nc, while the log-transformed analysis produced a power-law relationship characterised by α=1.01291 and k=0.4372. Rank-based modelling showed a strong concentration of incidence and mortality among a limited number of leading cancer sites. Comparisons with World 2020 estimates highlighted differences in the relative distribution of major cancers and in selected empirical parameters. Conclusion: The proposed empirical framework provides a quantitative description of the incidence–mortality structure of cancer in Romania in 2020 and generates interpretable, empirically grounded parameters that may facilitate comparisons across cancer sites, countries, and future GLOBOCAN datasets. Because the analysis is based on estimated cross-sectional data, the observed relationships should be interpreted as empirical associations rather than as direct evidence of biological or clinical causation. Further validation using longitudinal cancer-registry data and uncertainty analysis would strengthen the predictive and comparative value of the proposed models.