The Hyve provided technical and clinical staff of University College London Hospitals and other National Health Service (NHS) partners a high-level introduction to the OHDSI suite. During the online workshop held in February 2022, we explained the value of OHDSI and introduced them to best practices and key processes.
University College London Hospitals (UCLH), part of the National Health Service in the UK, provides first-class acute and specialist care in six hospitals across Central London. Its mission is to deliver top-quality patient care and high-quality education and research. The London hospitals have begun their journey with the Observational Medical Outcomes Partnership Common Data Model (OMOP CDM) and Observational Health Data Sciences and Informatics (OHDSI), aiming to convert datasets of interest from the Health Informatics Collaborative (HIC) to the OMOP CDM and perform analyses with the OHDSI tools. The audience at UCLH consisted of a mixed audience of both clinical and technical staff, two groups with varying levels of experience with the OMOP CDM. This presented a challenge to our trainers since they had to take this difference in background knowledge into consideration when giving the training.
How we solved it
The Hyve worked with UCLH to streamline and facilitate the adoption process of the various open-source components of the OHDSI stack by providing hands-on training and tailored education.
We provided an interactive workshop that consisted of three online sessions to both the clinical and technical staff at UCLH. Two of our OHDSI experts, Sofia Bazakou and Anne van Winzum, led the training. Live demos and exercises were provided to help the participants absorb and retain the newly acquired knowledge effectively. We also encouraged the participants to ask questions during the workshop. The training provided the participants with quick access to resources and foundational knowledge on the following topics:
Ins and outs of the current OMOP Common Data Model (OMOP CDM v5.3) including a deep dive into the OMOP standardised vocabularies, structure and conventions, and the ontology extension.
An introduction to the OHDSI ETL and data quality tools: White Rabbit, Rabbit in a Hat, Usagi, Data Quality Dashboard, CDM Inspection Report, and Achilles.
The OHDSI analytic toolset, focussing on the main features of Atlas and how to use these when executing an observational health study.
For reference, the participants were provided with all the recordings of the sessions, the presentations, and the answers to the homework.
The learning goals of the workshop were to get familiar with the OMOP data structure and analytic tools in order to apply this knowledge in the context of, in particular, hearing health and myeloma.
The workshop provided UCLH staff with a better understanding of the OHDSI suite and the value of OHDSI. The knowledge gained during the workshop provided a much-needed foundation for the six London hospitals in moving forward with their OHDSI journey. For example, during the workshop, we discussed what would be the best way to store pure tone audiometry data in the OMOP CDM.
The workshop was received with great enthusiasm and engagement from the UCLH clinical and technical staff. We are looking forward to seeing the workshop participants attending OHDSI community meetings and actively contributing to the working groups. Their expertise can, for example, prove valuable for the Vocabulary subgroup and improve how hearing test data is stored in the OMOP CDM vocabularies.