Mar 16: UMC Utrecht coordinates study on the use of AI in management of rheumatoid arthritis
Investigators at the department of Rheumatology & Clinical Immunology have received a large European grant to coordinate a study that investigates the use of computational models as a decision aid in management of difficult-to-treat rheumatoid arthritis. The total budget of this 6-year study is Euro 6.1 M of which Euro 1.3 M has been allocated to UMC Utrecht. The study coordinator and principal investigator are Paco Welsing and Jaap van Laar respectively. The consortium partners are located in Austria, Sweden, Portugal, Germany and the Netherlands. Other partners are Medical Data Works, a spin-off of University of Maastricht and a network of national patient organisations (EULAR-PARE).
Difficult-to-treat rheumatoid arthritis (D2T RA) is an area of high unmet medical need with major socio-economic consequences for patients and society. Contributing factors have been identified including co-morbidities, drug-related, biological and behavioral factors. However, identifying these patients with specific underlying and overlapping problems, or patients at risk, is a big challenge in routine clinical practice. Currently, treatment decisions are random and not sufficiently patient tailored nor data-driven. The STRATA-FIT (Stratification of Rheumatoid Arthritis: CompuTational models to personalise mAnagement strategies for difFIcult-to-Treat disease) consortium sets out to develop and validate computational models to identify and stratify D2T RA patients into clinically relevant phenotypes using real world clinical data. In addition, the investigators will measure biomarkers of inflammation to further characterize these phenotypes. Subsequently, they will execute a pilot study with a clinical decision aid based on their models to assess the effectiveness of personalized treatment strategies. In parallel they will develop a computational model to identify early RA patients at risk of developing D2T RA. By doing so they cannot only provide better treatment for patients with D2T RA but also work towards its prevention in early RA patients. STRATA-FIT will establish a unique European Learning Healthcare System, using a privacy-proof, state-of-the-art federated learning infrastructure in which patients with, or at risk of D2T RA are identified, stratified and treated in a personalized manner.
STRATA-FIT builds on previous work by consortium partners, who initiated and led the European Task Force on developing points to consider for managing D2T RA. Specifically a hackathon organized by, and using routine practice data from UMC Utrecht resulting in promising, preliminary prediction models for D2T RA. It brings together clinical experts, patient research partners and clinical-, biological-, data- and computer-scientists to tackle this major clinical challenge. When successful, STRATA-FIT will lead to more (cost-) effective D2T RA care and will greatly improve the quality of life of D2T RA patients while lowering the burden of D2T RA on Europe’s health care systems and society.