Guest post by Paul Carlin from South Eastern Health and Social Care Trust
MIDAS is all about sharing, sharing information in a coherent way that assures quality and presents it in a form that is readily accessible and actionable, but for what purpose?
We assume a benefit, and rationally we can suppose that having the ability to link disparate heterogeneous silos of data would bring added value, particularly for longitudinal analysis between datasets. There is a viable argument that verification of previous evidence/ research would follow as a natural consequence of aggregated collated high quality data, as well as potentially offering greater insight. By default this could aid interdisciplinary working, enhance reproducibility and greatly add to the evidence base.
It also potentially enables more esoteric objectives for analysis? How for example does the geographical provision of bus stops drive physical activity in towns, and through modelling, how could changes perhaps help improve activity for the local population?
The requirement would be linking activity data, transport data, health outcome data and perhaps social/ economic data to examine the above aim. Policy could thus be explored and drafted to drive the public health agenda, but there is a downside.
Individuals would need to accede to this type of use, certainly under the current legislation where the needs of the many do not outweigh the needs of the one. You could argue that perhaps the exemptions within the GDPR, in respect of research could allow this type of project to go forward, but how viable would this be without individuals actively consenting to the use, and thus having the ability to limit the utility of such an approach?
The risks are thought to be substantive, as worries over security of data, transfer, collation and accessibility for linkage would all need to be firmly addressed before this could take place. One could ask what the real risk is, I suppose, in that so what if its identifiable data, what effect could this possibly have?
The simple answer would be we don’t know, because the power that the data has really is only defined within the context it’s used, and who knows how someone with nefarious aims would try and leverage value out of it?
This brings us to what I consider the real crux of the issues around how data is used, particularly within the MIDAS context, how we can simply access the inherent value of this data whilst assuring ourselves that we have addressed the risks associated with use.
I would contend that we should reframe this, in that we should suppose a scenario in which we think of population based decision making, as a value to society and as we are all members of society we default to a system that allows the use of this type of siloed data to be used. This does not ignore the individual, but supposes that the individual understands that data will be used within particular contexts (it happens currently when we look at security/ terrorism legislation), for this to happen then the public must be aware of this need and have the necessary information to be assured of the merit, but perhaps more importantly should have an understanding of their responsibilities as members of society. This will be part of our work going forward within MIDAS, looking at how we might articulate the need and understand the palatability or otherwise of this approach to the public. The risks remain, but there is a tangible understanding of the importance and relevance of use, and perhaps a system of governance and assurance of quality would help protect and minimise the risk?