Access: Reducing barriers to data access for research in the public interest—lessons from covid-19

BMJ publishes its blogs regularly and provides a decent sounding board for practitioners in the UK. I rarely link to them (despite their local applicability and writing about how policy decisions impact them) but this one has a broad general applicability.

Opening up the data streams for a rapid accrual+ analysis seems like a pipedream. However, it can be difficult to provision issues related to data localisation and the “intellectual property” generated thereof. Approvals to use data in a blanket fashion, are going to be very hard as the governments realise its importance. They have proposed some ideas and I am reproducing them in its entirety here:

1.Reduce costs of administrative data access to researchers through core government funding for data processing, linkage and curation (avoiding cost-recovery models). This would enable more researchers to address questions in the public interest. This is already possible in some sectors, as demonstrated by the Department for Education for England and Wales, and in Sweden, where two thirds of MONA data system costs are centrally funded.

2.Simplify approval processes for de-identified data access through standardised guidance on necessary approvals proportionate to identification risk. Approval processes should be streamlined across organisations, including for demonstration of public benefit. [23] Reduce data release delays through increased capacity and more specialised data providers. Independent, accredited data providers should be created, with expert processing and disseminating capacity, knowledge of how data are used in research, and understanding of how best to prepare and deliver datasets to researchers (emulating the successful Secure Anonymised Information Linkage (SAIL) Databank in Wales).

3. Innovations that have allowed more timely data release during covid-19, such as the OpenSAFELY collaborative or more frequent releases of GP and hospital data, should continue and be made available to researchers to allow timely research on many topics. [26] Timely data release should not compromise quality, and organisations providing data should adhere to transparent and efficient response times. [15]

4. Enable more efficient data use through remote systems that comply with data protection requirements. [27] E-infrastructure must be improved to enable rapid data extraction and analysis. 

Reducing barriers to data access for research in the public interest—lessons from covid-19 – The BMJ