After a delay of several months and extended discussions with the ICO NHS England has now sent out its patient information leaflets for care.data and the extractions will start in the next few weeks. Whilst the NUG continues to support the aims of this project we continue to have serious reservations about the manner in which it is being implemented.
At a time when there is growing emphasis on transparency in public bodies there is a perception that NHS England is being less than transparent about the use of identifiable data in pursuit if its, undoubtedly laudable, aims. We are particularly concerned at the misreporting in the national press, which has suggested that the extract from the GP record will be anonymised – it is not. The NUG was also disappointed that, despite being asked to do so, NHS England and the ICO have declined publicly to correct this misapprehension. These concerns have even made it into the pages of the prestigious journal Nature.
The safe practice of medicine relies on a bond of trust in the doctor-patient relationship, and the manner of the implementation of care.data puts this at risk. In general patients are very happy to share their information for the greater good; but not all, and not when they have not been asked first. If patients feel that they have been misled into identifiable data being shared by their GP with the HSCIC it may have a significant impact on that bond of trust. A leaflet landing sandwiched between the local gardener touting for new business and the latest takeaway offering free delivery can hardly be deemed adequate information, especially when it doesn’t even mention care.data!
There is no doubt that the issues around sharing of information are complex and, as the technology for accessing and using this data improves, are becoming more so. As surgery has become more complex and involved we expect surgeons to develop means of explaining this in order to obtain informed consent. The same MUST apply to the excision of one's data as applies to the excision of one's organs. This will almost certainly need a different model for managing consent, otherwise we will end up with an increasing forest of consent Read codes. Linking consent with online access to patient records may be one way forward.
The “doctor knows best” attitude emanating from NHS England at present is reminiscent of the worst paternalistic medical attitudes of previous generations; the informaticians must catch up with clinical practice in terms of engagement and shared decision making with patients over the sharing of their personal information. The issues are complex, and will get more so, with a wide range of local and national programmes making use of a range of data: de-identified, pseudonymised and patient-identifiable. In my opinion there is a lack of clarity in the NHS England materials about the possible release of identifiable information by NHS England, and the option to opt out of this element (the 9Nu4 opt out). In addition to the information on the NHS England website a number of practices and individuals have posted guides on the web to this maze: in particular I would commend Dr Neil Bhattia’s site. Dr Mike Walton has shared a useful presentation for practices prepared for his CCG and there is a useful resource for patients provided by MedConfidential.
Practices should note that the latest statement from the Information Commissioner’s Office makes clear that, whilst the NHS England information and campaign contribute to informing patients, they are not of themselves sufficient:
“The posters, leaflets and additional guidance provided by NHS England on their website are all useful tools to help you do this, but it is your responsibility to ensure that you are satisfied that, as far as practically possible, all patients are aware of these changes.”
The resources previously developed by the NUG still seem to us appropriate and remain available.
In the mean time in our practice we have created a protocol alert to ensure that the issue is raised and discussed with all patients. This is linked to a template using the national opt-out codes, and the EMIS code for express consent, pending agreement with the coding centre on an appropriate national code.