Imagine trying to find the right key for a lock by guessing and checking, which is how many arthritis patients get their medicine. This project uses a smart computer system to group patients by their specific biological and behavioral patterns. It helps doctors pick the right treatment the first time instead of relying on trial-and-error.
Treatment for difficult-to-treat rheumatoid arthritis currently relies on trial-and-error, leading to ineffective therapies and high societal costs. There is no data-driven way to identify which patients will not respond to standard drugs.
A federated learning infrastructure and computational models to stratify RA patients into phenotypes. A clinical decision aid is also being developed for pilot testing.
If you are a drug developer dealing with high failure rates in clinical trials for RA — this project developed computational models that stratify patients into phenotypes. This allows for more precise patient selection, targeting the 20% of patients who are typically unresponsive to standard therapies.
If you are a software provider dealing with a lack of data-driven tools for rheumatologists — this project developed a clinical decision aid. This tool transforms real-world clinical data into actionable treatment strategies to reduce the socio-economic burden of RA.
If you are a clinic manager dealing with high costs and poor outcomes for chronic joint disease — this project developed a federated learning infrastructure. This allows you to identify patients at risk of progressing to difficult-to-treat status and apply preventive strategies early.
Based on available project data, no specific pricing or implementation costs are provided as the project is currently in the research and validation phase.
Yes, the project is building a European Learning Health Care System using a federated learning infrastructure across 7 countries to ensure scalability and data privacy.
Based on available project data, specific licensing terms are not mentioned, but the consortium includes an SME (MDW) and several universities, suggesting a collaborative development model.
The project uses a privacy-proof federated learning infrastructure and adheres to FAIR principles for data handling to ensure security across different clinical nodes.
The project period runs from 2023-05-01 to 2029-04-30, with a pilot study planned to assess the effectiveness of the decision aid.
The consortium is heavily weighted toward academic and clinical expertise, with 5 universities and 2 research institutes across 7 countries. However, the inclusion of one specialized SME (Medical Data Works) and the EULAR-PARE patient network indicates a strong push toward practical, real-world application and data integration. The 10% industry ratio is low, but the strategic partnership with a federated learning expert suggests the technical infrastructure is a priority for commercial viability.
Contact Universitaire Medisch Centrum Utrecht (UMCU) regarding the STRATA-FIT consortium.
Contact SciTransfer to explore licensing opportunities for the D2T RA stratification models.
