Gaining Efficiency in the Integration of a Multimodal Oncology Real-World Data Source into an Enterprise Analytics Platform

ISPOR Europe 2026

Nov 08, 2026
Oncology
Presentation

Sven-Eric Schelhorn, Jose Munoz, Sabrina de Souza

Objectives
Real-World Data (RWD) acquisition delivers value when paired with efficient, reproducible data integration. Yet integrating multimodal oncology RWD into an enterprise analytics ecosystem remains a time-consuming, error-prone process that limits the timeliness and reproducibility of evidence generation. This project assesses the methodological and infrastructural challenges of ingesting a de-identified multimodal oncology RWD source into the Merck KGaA data ecosystem; and aims to determine whether an end-to-end ingestion framework can improve the timeliness, accuracy, and reproducibility of oncology evidence generation for research and development.

 

Methods
In collaboration with Tempus AI, Merck KGaA implemented an end-to-end integration pipeline of seven distinct workflows (one clinical, six molecular) using Genedata Profiler-based modules to route over 70 data tables into the Merck KGaA XOP data lake.

 

Results
The project encompassed data ingestion, transformation, quality assurance, and the delivery of analytics-ready views for multimodal (MM) and clinical genomic (CG) cohorts within the Merck KGaA data lake accessible to researchers. This initiative emphasized data provenance and governance alignment, and reproducible analytics to support subsequent translational and clinical investigations.

 

Conclusions
The scientific value of RWD is realized when integration is fast and robust enough to make it accessible to researchers. An end-to-end, modular ingestion framework enhances data quality, reproducibility, and analytic readiness across clinical and molecular domains. UI-driven integration workflows facilitate maintainability and rapid adaptation to evolving data schemas and mappings. Rigorous provenance and quality assurance are essential for trust and interpretability of downstream analyses. Cross-organizational coordination between industry partners and data providers is critical for timely delivery and alignment with governance standards.