datma launches federated patient journey mapping to improve real-world evidence insights for life science companies

datma’s new AI-powered capability helps pharmaceutical and life science companies map treatment pathways, uncover care variation, and identify delays using privacy-preserving federated data analysis.

datma has launched a new patient journey mapping tool within its federated data platform, datma.FED, aimed at helping pharmaceutical teams generate actionable real-world evidence (RWE) without compromising data privacy. The feature enables analysis of care timelines from diagnosis through treatment and follow-up, helping users to identify where delays, therapy discontinuations, or variations in decision-making occur across clinical settings.

Part of the broader datma.FED platform, the new functionality aims to support secure, compliant collaboration between pharmaceutical companies, healthcare providers, and laboratories. Built to align with institutional governance and regulatory requirements, the platform will enable secondary use of real-world data (RWD) across decentralized networks.

"This launch is an important step toward helping pharma teams understand what's happening between diagnosis and treatment, where patients drop off, how providers make decisions, and why variation exists across care settings," said Robin Edison (Vice President of Product, datma).

The tool uses large language models (LLMs) to extract and structure information from both structured and unstructured data sources, such as clinical notes, lab results, and medication records. Events, such as treatments, diagnostics, or condition updates, are timestamped and organized into directed acyclic graphs (DAGs) that represent each patient’s care journey.

To ensure data consistency, the system applies a multi-step cleaning and normalization process. Event descriptions are standardized using curated vocabularies, timelines are normalized relative to each patient’s first recorded event, and daily treatments are aggregated and sorted chronologically.

All processing occurs within datma.FED’s federated architecture, ensuring that data remains securely within each participating organization and is never centralized. Patient journeys are computed locally and converted into anonymized numerical representations for aggregate analysis, enabling population-level insights while maintaining privacy. “This functionality gives organizations visibility without requiring raw patient-level data to be centralized or transferred,” Edison added.

The new capability supports a range of RWE use cases, including mapping diagnostic and treatment pathways, identifying delays or discontinuities, and analyzing variations that may impact patient outcomes or access to care.