Benedict’s comments are very thought provoking, and his concerns about the data subject in the ‘public benefit’ model are particularly important. To be clear, though, ‘public benefit’ may be a misleading term for our approach.
What we seek to do is to balance the interests of individual data subjects in privacy and autonomy with the collective interests of society in more and better health research, rather than to give either of those interests priority, as we argue the current ‘consent or anonymise’ regime does.
Still, the predominance of this regime over recent years may mean that moving to a more balanced approach where collective interests are accorded their proper weight might raise concerns about protecting, as Benedict puts it, the ‘special status’ of the data subject.
Benedict suggests that the data subject has a special status because it is they who will bear the costs if data is lost or misused, and yet the person who will decide whether data can or cannot be used in the first place is not the data subject – and therefore not the person who will pay the price of misuse.
In this way, the ‘third party’ who determines the legitimacy of data use therefore ‘speculates’ about what the opinion of the data subject might be, rather than ascertaining their actual wishes.
There are several points to respond to here. Firstly, Benedict notes that the chances of misuse of data are actually very small indeed. However, it is of course true to say that absolute data security can never be guaranteed.
But, given this impossibility, is it not reasonable to ask people to accept the very small risks of misuse – we might say they are ‘reasonable’ risks – as part of their contribution to the collective benefits of research? That is, is the bearing of very small risks not a part of reciprocity, so long as the contribution entailed by that reciprocity is a fair one?
There are two other, related points to be made about whether decisions should be made without the express consent of individuals
One thought here is as follows: we can reasonably presume that a strict interpretation of the ‘consent or anonymise’ regime would indeed seek to act on the ‘actual wishes’ of individuals – it would grant them strong rights over use of their data.
However, we can also reasonably presume that such an approach would significantly obstruct the course of research. So this approach could preclude the use of data that is valuable to precisely those rights-holders whose ‘actual wishes’ deny access to their data.
The problem here is one which arises from thinking only at the individual level in a system where individual decisions can be collectively irrational (as well as individual decisions being – by effect if not by intention – individually irrational, given the necessarily collective aspects of the problem in so far as a certain mass of data is necessary for useful inferences to be drawn).
Another point arising from the notion of the ‘actual wishes’ of individuals is that policy decisions must often be made not on the basis of actual agreement but rather on the basis of hypothetical agreement to justifiable social principles which in turn guide decision making.
This is the essence of the idea of the social contract as a system of rules and practices to which all relevant actors could sign up as reasonable, given that they recognise that other actors have legitimately different interests.
It is not, then – and indeed in practice cannot be – I would argue, so much about ascertaining the actual wishes of individuals as about ascertaining a set of justifiable principles to which all actors could agree, given the conditions of a pluralistic society where collective benefits need to be secured for the good of all.
Benedict’s idea of giving a representative sample of data subjects a vote is compatible with this and could perhaps be seen as a good democratic check on whether the principles and practices employed are indeed justifiable to the citizens who are parties to the social contract.
Dr Sarah Clark is a research associate at University College London. Please note that the views expressed in guest blogs on the Nuffield Trust website are the authors’ own.
Dr Clark is responding to Nuffield Trust Fellow in Health Policy Dr Benedict Rumbold’s blog: Accessing person-level data in health care: debating the key issues.
Clark S (2011) ‘Accessing health care data: the importance of collective reasoning’. Nuffield Trust comment, 16 August 2011. https://www.nuffieldtrust.org.uk/news-item/accessing-health-care-data-the-importance-of-collective-reasoning