We have worked with healthcare data for over 25 years, including patient-level research data originating in hospitals and specialist clinics. We also produce our own patient record system which has been used in the NHS for the past 20 years.
Our experience of using OMOP as a data source has been gained through a number of high profile analytics projects, including:
- Large Medical College: We used OMOP as a source for a large project with a major US medical college. The project used OMOP as the data repository and included diagnostic, procedural, lab and medication data. We are currently expanding the data sources used as well as exploring supplementing the clinical data with financial data.
- Academic Health Science Centres: Using OMOP to unify multiple medical data sources into a single database, we provide population health analytics on around 0.25M patients and 1.5M encounters.
- Social Determinants of Health: We are using an OMOP database as the source for a project that combined social determinant data with clinical data.
Other database experience and projects
- SUS (NHS Secondary Users Service dataset).
- Disease Registries management and brokering for a major cancer research institution (primarily oncology registries but also diabetes and COPD).
- Horizon 2020 (2016-18): We led the 'Atmohealth' Horizon 2020 project. Partnering with institutions such as Imperial College, Karolinska, The National Centre for Mental Health, Wales and Limerick University, Ireland, we developed a pan-national collaborative analytics platform for population health.
- Large-scale import of prescribing data from primary care to support the identification of ‘microbial resistance hotspots’ associated with the prescribing of antibiotics.
- Enterprise implementation of our clinical research suite, including clinical registry platform and analytics platform at a major US Comprehensive Cancer Center.
- Partnered with and imported data from many EMRs such as EPIC and Cerner.
Relevant links
- Product Video: An overview of one of the healthcare specific products that Imosphere has built (https://www.youtube.com/watch?v=8hTCyccBw6M)
- Statistical and Analytic Expertise: Paper outlining building a model to manage risk in mental health services https://imosphere.co.uk/cdn/files/resources/assessing-and-managing-risk-in-mental-health-services.pdf
- Sample Project: In partnership with Clearsense, we worked to bring multiple data sources together into a single OMOP data source (https://home.mmc.edu/research/office-for-research/research-centers/meharry-data-science-center/)
- Published paper: A research project investigating data provenance in healthcare analytics - Xu, Shen & Rogers, Toby & Fairweather, Elliot & Glenn, Anthony & Curran, James & Curcin, Vasa. (2018). Application of Data Provenance in Healthcare Analytics Software: Information Visualisation of User Activities. (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5961786/)
We have significant experience of developing data collection/assessment tools that have been used on both a national and international basis. Our Overview Assessment was accredited by the Department of Health for use as a standardised needs assessment tool across health and social care in 2002 and was used by around 70% of councils and the NHS. (Clifford (2005) The Single Assessment Process: Simplifying Complexity. National association of Primary Care Review 2005, Issue 2, pp80-81).
We have also developed a suite of predictive algorithms and banding methods to support the allocation of resource in health and social care and special needs education. These are delivered via a centralised algorithm web service and are used by 60 councils and CCGs in the UK. They have processed over £7bn of public funds. (Clifford, Saunders and Gibbon (2013) Modelling the relationship between needs and costs: how accurate resource allocation can deliver personal budgets and personalisation. Research, Policy and Planning (2013) 30(2), 107-120).