Big Data and Statistics
The emergence of new digital technologies is allowing huge quantities of information on health exposures and outcomes to be captured, stored, and analysed using advanced statistical techniques. ‘Big Data’ will provide new ways to conduct research and offers the potential of a dramatic increase in the scale and efficiency of clinical studies. We are capitalising on this opportunity through our involvement in Oxford’s new Big Data Institute, opened in 2017, which will enhance the close collaboration between clinical epidemiologists, statisticians, computer scientists, engineers and ethicists necessary to develop this field of science. The Deputy Director of the Big Data Institute, and co-theme leader Professor Martin Landray, is leading a pioneering programme of work within PHRU that aims to harness Big Data methodology to facilitate large-scale epidemiological studies and improve the efficiency of randomised trials by identifying barriers to their efficient conduct.
Our work in this area is supported by a computing team, overseen by Dr Alan Young, that produces and maintains systems designed to meet the specific requirements of our new and on-going studies, and to ensure that they are sufficiently adaptable to provide a platform for future work. Our systems focus on aspects that are key to study quality: recruitment of large numbers of eligible participants; data collection and study conduct; study management, coordination and oversight; laboratory management including analysis and storage; ascertainment, confirmation and classification of outcome; and analysis, reporting and dissemination.
Our bespoke IT systems comply with relevant security and data privacy policies and regulations (including EU Data Protection Directive 95/46/EC, and US Health Insurance Portability and Accountability Act (HIPAA) 1996) as well as, where appropriate, standards for clinical trial systems (including US Code of Federal Regulations Chapter 21 Part 11 “Electronic Records and Signatures” and EU Clinical Trials Directive).