It’s time to recognise the benefits of medical data use
Using medical data may help improve public health services and help identify patterns of disease, which could lead to more effective practices for prevention. Our head of clinical ethics and data, Dr Jon Fistein, argues that we should say more about these benefits.
It is widely recognised that routinely-collected health data can be used to improve the healthcare provided to individuals and to communities. The US Institute of Medicine champions the ‘learning healthcare system’ in which routinely-collected data are used to drive better, more efficient medical practice and patient care. Indeed, many argue that such uses of data are the only ways to improve services, reduce waste and make health service provision sustainable.
Nevertheless, people remain concerned about potential impacts of data use on their privacy and confidentiality.
One possible reason for this is uncertainty about who might be able to access data about them, and whether those with access can be trusted to use the data appropriately. The modern NHS is not a single entity but a partnership between public, private and third-sector organisations working together to develop treatments and deliver care. A recent Wellcome Trust report found that the public recognise value in commercial and other ‘non-NHS’ organisations accessing health data in order to develop new medicines or ensure existing treatments are safe. However, they tended to underestimate how much access is necessary to achieve these aims.
People were also confused by technical terms such as ‘anonymisation’ and ‘identifiability’, tending to assume that data are either anonymous or not. In reality, anonymity is context-dependent. For example, the more data gathered about an individual, the more likely it is that someone might infer that person’s identity from their ‘profile’. In genomic research, the genome itself is potentially identifying. In rare disease research, people involved in the field may be able to identify individuals from relatively limited data. Advances in technology may increase the risks of reidentification by inference and it may become impossible to truly anonymise data without also destroying their value.
Perhaps it is time to stop focussing on the potential risks of reidentification and instead emphasise the actual benefits of data use. After all, many people seem willing to share information about themselves with commercial organisations, through supermarket loyalty cards, wearable health trackers and social media, possibly because they perceive direct benefits and a sense of choice over how these data are used.
One common theme that consistently emerges from research on public attitudes to data use is the need for those handling data to be trusted. How can researchers wanting to use health data demonstrate and maintain their trustworthiness?
Part of this is about explaining which data will be used for particular purposes, being clear about the benefits and risks, and being transparent about the many parties involved. We need a wider public discussion to decide when, how and by whom health data should be used for public benefit. Representative citizens’ juries, with the time to consider these complex issues in depth, may offer a route to defining a framework for data use, to realise the learning health system vision to which we all aspire.
This article has been adapted from the original published on The Guardian on 29 June 2016.
The MRC is supporting the Independent patient data taskforce that will build on the work of the Caldicott review to help to develop a framework for clear and transparent discussions with the public, patients and healthcare professionals about how data can be used to improve health.