By Prof. Petros Patias
.
OHDSI Europe Symposium 2026 took place in Rotterdam from 18 to 20 April, bringing together 375 participants working across observational health data science. The programme included 130 posters and 15 software demonstrations, creating a busy and highly collaborative environment.
The PREPARE-REHAB team was strongly represented, contributing three posters and one software demonstration. Together, these focused on a shared challenge: how to make complex clinical data usable for research and, ultimately, for better rehabilitation outcomes.
One strand of work looked at how different rehabilitation datasets can be brought into a common format. In “Experiences with OMOPping Rehabilitation Data: A Unified ETL Strategy for Nine Distinct Clinical Datasets”, the team presented a practical approach to aligning diverse data sources, making it possible to analyse them together rather than in isolation.
Another contribution focused on the work required before any modelling can happen. “An Iterative Annotation Pipeline for Building Clinical Datasets and Training Information Extraction Models” set out a structured way of annotating clinical data, improving quality step by step and supporting more reliable downstream analysis.
PREPARE also shared progress on prediction models in “PREPARE for Precision: Patient-level prediction models for nonsurgical thumb base osteoarthritis treatment”. This work explores how patient-level data can be used to support more tailored treatment decisions in clinical practice.
Alongside the posters, the team presented a software demonstration: “A Semi-Automated Pipeline for Extracting and Standardizing Clinical Concepts from Unstructured Medical Records”. The tool focuses on turning free-text clinical notes into structured data, helping to make information that is often difficult to use more accessible for analysis.
These contributions reflect the collaborative nature of the project, with partners across multiple institutions involved. Team members presenting at the symposium included Petros Patias, Ruud Selles, Erik Calcina, Esther Janssen, Krijn Polder, and Lisa Hoogendam, among others.
OHDSI Europe 2026 was an important opportunity to share progress, test ideas, and connect with others working on similar challenges. For PREPARE, it was also a chance to show how different pieces of work, from data standardisation to modelling and tools, fit together within a broader effort to strengthen data-driven rehabilitation.
.
Posted by: Nayara Güércio (Trilateral Research)