November 10, 2021

What’s so good about trusted research environments (TREs) anyway?

John Marsh, one of five Public Contributors making sure public interest remains at the heart of DARE UK as a member of the Programme Board, tells us why he got involved and what he thinks the programme can achieve.

Public contributor John MarshI worked in consumer data and had no interest in health data until I was diagnosed with bowel cancer through the screening programme in 2008.

Following that, I started to get interested in health research, and particularly in data-driven research to support prevention and early diagnosis of cancer. I got involved as a patient representative in various projects in that area and realised that there are lots of barriers and delays between good, legitimate research ideas and actually ultimately improving health outcomes.

Specifically, I noticed: 

  • The relevant, useful data are often spread among several different organisations. Each organisation rightly has very strict rules surrounding research access to that data. Typically, those rules are broadly similar with the same privacy and security objectives, but different in some details, which means researchers can spend a considerable period adjusting their proposals to satisfy each set of rules. Applying one set of strict rules on access and use would help address that issue. 
  • With the traditional model, once approval for the research has been granted, the data is then released to the researchers and they must load it into their own IT infrastructure (within their university or other institution) for processing. In addition to building the required infrastructure to handle often unfamiliar data, general IT departments often face the additional and necessary burden of applying proportionate security controls and  standards to handle sensitive data as required by the data owners.  
  • Finally, if multiple unlinked data sources are involved, the researchers must set about linking them together, and problems arise because of different data standards and quality. They may also be duplicating work that has already been done by another research organisation for a different research project. 

These issues mean that researchers may spend months at the beginning of their project working hard to overcome these barriers before they can get on and do what they are best at – the research – and we all know that in health especially, ‘delays cost lives’.

Trusted research environments (TREs)

As a patient representative, I became very interested in trying to help overcome these barriers. Two to three years ago, through Cancer Research UK, I came across the concept of a trusted research environment (TRE) and was immediately struck that they could overcome many of the barriers. 

I became an advocate for well-run TREs, where a highly professional organisation sets up a top security infrastructure that holds diverse, pre-linked data that researchers can apply to access. A single set of comprehensive rules can be set up for access to the data so that researchers make a single application for access which will cover the requirements of all the original data sources. If their application is successful, they don’t have to set up a new IT infrastructure; they are given controlled access to the pre-linked data in the TRE. In other words, the data is research-ready. 

“If different trusted research environments (TREs) across the UK do not work together in a more joined-up way, this will lead to ‘data islands’ with different rules for access and resultant problems for researchers.”

Since learning about TREs, I’ve been involved with three as a patient representative and have followed the progress of several others. There is a risk that – as more data becomes available for research over time – if different TREs across the UK do not work together in a more joined-up way, this will lead to ‘data islands’ with different rules for access and resultant problems for researchers.

I think it’s time for one or two organisations to learn from what’s gone on and try to build larger-scale, well run TREs with diverse, pre-linked data so that the benefits of the concept can be better realised. From what I’ve learnt about the UK trusted and connected Data and Analytics Research Environments programme (DARE UK), I think it has a good chance of achieving that, and that’s why I’m glad to have the opportunity to try to help as a member of the Programme Board. 

You can find out more about John and our other Public Contributors on the Involving the Public page.