How Big Should Big Data Be?
The main issue with Care.data is not a technical one of course. It is the abject failure of the project to communicate with users that the benefits of having some of their most personal information collected together outweighs the potential security risks and wider sense of unease. Yet, even if the communications had been better handled, would enough people feel convinced?
A major part of the problem is that the patient data in Care.data is meant to be anonymised. The purpose of Care.data is not to transfer specific patient data between GP’s and hospitals, but to provide a wide anonymised database that can be mined for trends, events, connections. These aspects are exactly those that big data aims to delve into. What has been revealed is that the anonymised data in Care.data could be combined with other sources to enable someone to identify specific patient groups and possibly individuals. And it’s the latest analytics techniques that are enabling this. Put another way, if analytics was not so powerful, then this might not be such an issue.
As with many technologies, the issue with big data is sometimes not whether we could but whether we should.