Challenges in big data

Real-world clinical data has become an invaluable source of information used in academia and industry to support evidence-based research, advise health-care delivery, and support new therapy development. Unfortunately, reproducibility, transparency, feasibility, and accuracy are placed at risk when these datasets are not parsed using expert-level tools. The vast amounts of data alone often present challenges to statistical approaches that can become unnecessarily slow, expensive and error-prone.

An innovative solution: Curator

Researchers at the University of Oxford have developed a new solution to these challenges with the creation of the innovative data science tool Curator. Curator performs automated data engineering, data mining and advanced curation on electronic health records to provide tailored clinical datasets. Data are efficiently, accurately, and consistently curated and delivered in a format that is ready for statistical analysis.

Thanks to Curator’s scalable design, the tool can be applied to a wide range of datasets including those provided by Clinical Practice Research Datalink (CPRD) and NHS Digital. Curator can be expanded as additional clinical datasets become available. To date, the software has supported more than 20 observational studies that span a broad spectrum of conditions and therapeutic investigations. Curator consistently saves time, reduces costs and improves both research quality and reproducibility.

The tool confers the following benefits:

• A vast clinical scope including common chronic conditions
• Significantly reduced time and costs per study
• Reduced human error by supporting reproducibility and facilitating comparisons within and between studies
• No direct competitors

Commercialisation

Oxford University Innovation is interested in speaking with companies or institutions that would like to test and adapt this tool to improve their processing and statistical analysis of real-world data.