HRRDLs for commissioning: a discussion towards Safe, Consensual and Transparent use of data in commissioning

Yesterday, medConfidential and others attended the HSCIC’s “Driving Positive Change” event, to briefly look back at the Partridge Review, and forward to future work of the HSCIC. The two major topics were communications of various types, and the proposed HSCIC “safe setting” where bona fide research could be conducted on data (currently subject to opt-out). Both of these things are welcome areas, and we seek to be closely involved in what happens next with the first public steps in the next week or so.

The Department for Health is running a consultation on “Accredited Safe Havens” for commissioning purposes, or, as they call it slightly less clearly, “Protecting personal health and care data”. The consultation gives NHS England companionship in terms of public engagement quality, and has led to a great number of puzzled looks by area experts. I’m currently attending a variety of meetings with a variety of organisations, and not only is no one really sure what the answers could be, few people agree on what the questions are intending to ask. This seems less than ideal.

Yet, as we are now in week 5 of a 7 week consultation, and no one really has a solid articulation of what the Department of Health are trying to do, I’ve put together this draft of a substantive paper on a way forward: “HRRDL’s for commissioning”. It’s based on previous work which has been adopted by HSCIC but after DH began their consultation drafting, which was as care.data was imploding around NHS England. If you think that the consultation as drafted takes no account of HSCIC’s progress since February, that’s because it mostly doesn’t. Comments by email are very welcome.

What is a safe setting? A safe setting is a physical venue where (usually remote) data can be accessed under tightly controlled and audited conditions. Restrictions are placed on who and what enters the room, what they do when in there, and what they can take out. This allows for research to be conducted on individual level records which have minimal protections (which, for health data, has other problems). They were previously discussed for legitimate research, along existing models. This paper takes the proposal further. We fully expect, and have no reason to disbelieve, that the optout codes for care data (and beyond) would be fully honoured. We intend that this proposal is fully compatible with the consent mechanisms that are in place, and that should be in place, and does not deny screening to those who have opted out of secondary uses. A safe setting can also restrict which individuals can see which data, which has implications for a granular approach to parity-of-esteem questions.

I don’t think that this is currently a final proposal so can evolve, (it’s dated so you can tell, and we’ll put a note here: when we do), and some may need more explanation, but if you’re interested in how we think commissioning data for invoice reconciliation and risk stratification (neither of which are direct care, so all come under the opt out process) could work in a way that is safe, consensual and transparent, I’d like to hear your comments below or to sam@medConfidential.org

Please note that making comments to us is not the same as responding to DH itself, which you can do online