ABSTRACT Accurate cancer staging is essential for cancer research and treatment evaluation; however, real‐world data (RWD) often include incomplete or inconsistent UICC classifications. We present the development, optimization, demonstration, and release of the Munich UICC Staging Tool (MUST), an open‐source R script that automates UICC staging based on TNM classifications and tumor characteristics, improving data completeness and consistency. MUST incorporates official UICC staging rules (6th–8th editions). A key feature is the application of RWD‐specific rules to handle missing values and incomplete information. The tool is transparent, highly customizable, and adaptable to future TNM versions and diverse applications. MUST was validated using anonymized clinical tumor registry data from two Munich university hospitals (analyzed independently, n > 100,000), focusing on pancreatic and stomach cancers. Compared to documented UICC stages, MUST increased staging completeness from 60% to 72% in Clinic‐1 and from 20% to 74% in Clinc‐2. Agreement rates between documented and MUST‐generated UICC stages were 90% in Clinic‐1 and 95% in Clinic‐2. To assess reliability, a Confidence Level metric summarizes the proportion of original versus additional rules applied. 70% of MUST‐derived stages in Clinic‐1 and 48% in Clinic‐2 were classified as “Assured,” with the remaining cases relying on RWD‐specific rules. The results demonstrate that RWD‐specific rules substantially improve completeness, while the differences between clinics highlight variations in documentation and coding practices. MUST offers a reliable solution for UICC staging, addressing documentation gaps and improving data quality. Its adaptability makes it valuable for clinical documentation, registries, and large‐scale oncology studies.
Kühnel et al. (Sat,) studied this question.