Data science research

We aim to utilise a wealth of NHS and research data to enable every patient to become a research participant, guide patient care decisions and usher in a new era of personalised medicine.

Theme leads: Professor Anneke Lucassen and Professor Dame Wendy Hall

Creating an ethical framework

We are using our expertise in clinical ethics and law to guide our use of patient data and develop a fully functional data repository. This will ensure that our collection, storage and use of NHS data sits within in a comprehensive ethical framework that is compliant with a changing legal environment.

This ethical framework will cover data entry into the repository, studies using the data, and how findings should be fed back to patients and family members. This will enable us to develop a data repository we can use to improve and personalise the healthcare delivered to NHS patients.

Data standards and usability

To be able to use the data we collate effectively, we need to use standardised language. We are therefore working to establish common vocabularies between the University of Southampton and the NHS, based on the Clinical Data Interchange Standards Consortium interoperable data standards. To ensure data security, we are storing the anonymised data in a central repository located within existing NHS systems.

Hardware and software

We are working closely with Microsoft Research, who are developing Microsoft Azure Cloud Compute, a new piece of software with the capacity to meet the security specifications required for NHS and government classified data.

In addition, Southampton is home to the most powerful academic supercomputer in England, IRIDIS 4, which we are using to process large data sets and images. Together, these will allow us to process and securely store large volumes of data from the Trust’s NHS electronic patient record and NIHR BRC research projects.

Data analytics to stratify NHS patients

We are combining our expertise in a wide range of areas to analyse the data in the repository, with the aim of discovering why some patients respond predictably to a treatment and others do not. By identifying specific traits in those who do not respond to treatments, we seek to delve deeper into the reasons why patients respond differently. This will help to guide treatment decisions and support a more personalised approach to healthcare.

Clinical reporting, feedback and evaluation

Working within our ethical framework, we will feed back results from studies which analyse the data to clinicians to inform and guide patient care. Through this approach, we will enable patients to receive the right treatment for them. We will also be able to track their progress to evaluate which treatments are effective for which patients, supporting research by our other NIHR BRC themes.     

Education and dissemination

We are working with Health Education England (Wessex) and the Southampton genomics programme to make data science open and transparent, and provide appropriate training for NHS health professionals and researchers. Our long-term objective is to scale and standardise our methods so that they can be used in clinical practice by other NIHR BRCs and NHS Trusts.