1. Scope of the policy
The policy needs to set out more clearly that its focus is data collected through routine NHS and social care, where an individual has not provided informed consent for the use of the data beyond direct care. This will be enabled through a network of Secure Data Environments which will stop the ‘physical’ transfer (dissemination) of this type of data to external organisations.
The policy should be explicit that it does not affect data access where the individual has provided informed consent for its collection and use (for example, through joining a research cohort or clinical trial). However, there is a grey area in relation to consent which needs further consideration to mitigate potential unintended consequences. Work needs to be undertaken to define what is meant by a ‘consented cohort’ or ‘consented trial’, to seek public input on this, and to identify a policy statement that can accommodates consented studies without risking substantial bias and research inequity that would impact minority and marginalised groups most heavily.
2. Linkage to non-NHS data
There are valuable classes of research that rely on linking individual level health data to individual level data from non-NHS sources. If NHS data is restricted to NHS operated SDEs alone, linkage may not be possible. For example, administrative data covered by the Digital Economy Act (DEA) can only be accessed in DEA Accredited Processing Environments. In our view it will be necessary to allow the operation of some accredited SDEs, perhaps operated with joint oversight, and the policy should provide for this, building on examples from other UK nations and ways of working during the pandemic response.
3. Harmonised data access processes
Whilst the proposed harmonisation is critical to reduce unnecessary administrative steps and promote consistent approaches to scrutiny of data access requests, we urge a more ambitious approach of ‘one project, one review’ to minimise delays and complexity. This is vital for large scale projects needing to work across more than one SDE, including UK-wide studies. This would be facilitated by an approach similar to the process of applying to a Research Ethics Committee, including public and patient involvement in the decision making. ‘One project, one review’ would result in efficient, faster data access for appropriate research in the public good and reduce complexity in a way that promotes better transparency and public understanding.
4. Exceptions and the role of co-design and public involvement in decision making
Maintaining both public and research community support is critical for the success of this policy and its implementation. It is therefore critical that language discussing exceptions is precise with guarantees of transparency and engagement, particularly related to any revisions, to ensure alignment with what the public deem acceptable. To facilitate this, we suggest adding a commitment to permanent structure(s) along similar lines to the Genomics England participant panel or to Citizen Juries so that debate about scope and what are acceptable uses and exceptions, and what are not, can take place over time, and the overarching decision making can be put into the hands of representative groups of citizens.
