Health Data: the problem of distinguishing between public and private sectors

The Wanderer above a sea of fog by Caspar Davi...

Why limit your view when you can see this far? The Wanderer above a sea of fog by Caspar David Friedrich, around 1818 (Photo credit: Wikipedia)

“The Open Data Era in Health and Social Care”, prepared by GovLab (NYU) has been released.

I have no issue with open data, and the more open the better. However, doctrine may interfere in respect of the way data are viewed in the UK.

The typical model is to focus on the NHS, as the main provider of healthcare services. Certainly, this makes good sense, on its own. But the NHS is not on its own. The title is a bit misleading, in that while Social Care in included, the English NHS this is not integrated, suffers from bureaucratic accounting rules that prohibit pooling of budgets (hence the problems with the Better Care Fund), coupled with means testing, a cash market, and a major role of charities in filling in service gaps. Countries with patient-copayments and transaction data manage to integrate health and social care around the patient because of the ability to avoid arbitary distinctions between provider types and their ownership. As a result of what is both a strength and weakness of the NHS, policymakers have had and continue to have considerable conceptual difficulty integrating public and private provision into a patient-centric and whole-system model of seamless care.

Healthcare is bigger than the NHS as people in the UK can buy private health/medical insurance, pay cash for private treatment or use private hospitals under NHS contracts. In addition, patients seek services from dentists, physiotherapists and pharmacists, and others, who in the main are outside the NHS in terms of practice patterns.

Let’s take medicines. Years ago the NHS explored electronic prescribing, a project initiative I was doing some policy work on. I had asked whether private prescriptions and dental prescriptions were to be included and was told, no, they were excluded as this was an NHS project. Of course, thinking such as this means that they were failing to look at the whole system of medicines prescribing. A patient for instance who is prescribed an antibiotic by a dentist (and they prescribe a lot of antibiotics) would discover not only that that information was not available to their GP, but the GP would likely not know that dental surgery had even taken place. And private/independent prescriptions were simply off the table!

The only way that Open Data Era thinking can prevail is when the English NHS and the Department of Health adopt whole systems thinking. The modern world is full of boundaries that are being breached by new technologies, that are challenging assumptions of the past that in the future will prove dysfunctional.

The NYU report (I am surprised at the lack of whole-system perspective — perhaps they didn’t know about the wider health system??) does not address the distinction between NHS and private/independent data (though they do make the point that Open Data might be used along with private or independently held data, but in the context of my remarks, this seems a fudge).

I won’t go into a detailed analysis of their logic model on page 45 of the report which crystalises their essential argument. Logic models are conceptual models that link various elements (inputs, outputs, outcomes) to desired impact in a coherent (logical) way. Needless to say, they start with NHS data. Examining the Activities/Outputs parts, would suggest that the full realisation of the stated benefits will not be possible. Limiting the data in, as the model does, means that achieving operational efficiency or resource allocation (impacts) will lack private sector comparators for instance. One output, Policies Created/Changed, is immediately compromised by the inability of the model to account for the role of the independent/private and not-for-profit sectors, which is about 10% of the total activity and expenditure. Indeed, their definition of ‘internal users’  (page 48) excludes non-NHS entitities, and they aren’t seen as ‘external users’ who might need to access NHS data. Furthermore, the approaches proposed to capture measurement limits the focus to state-mandated bodies (i.e. NHS), and therefore limits the ability of measurement to assess potentially new approaches to care that may be invented. So much for measuring innovation.

It would have been better to start  with the needs of data users and their objectives, in a whole system approach. This is the fundamental weakness in the logic model and limits the report considerabley. In the end, it makes me worry that the initiative will in the longer run fail to be as successful as it might be.As Einstein said: “No problem can be solved by the same kind of thinking that created it.”

Leave a Reply