Event-centred data versus patient-centred data

Changing the way we describe health and social care activity

Here's a diagram (which we'll discuss in a bit more detail later) that was drawn for me as part of an exercise in which health and social care managers were asked to sketch how they saw "the whole system" in their minds' eyes:

The health and social care whole system represented as a pyramid.

There are barriers to us getting data across to health and social care professionals. When we number-crunchers try to describe their world using data our efforts are often met by glazed-over expressions.

One of the barriers we're up against is a cultural barrier I've talked about previously. It's to do with a mismatch between two different views of the world. Health and social care professionals see the world one patient or one service user at a time; data analysts on the other hand see the world as aggregations or summaries of patients or service users.

One answer to this clash of cultures is to try and create data displays in which you can almost "see" individual people. In the On resonance piece, for example, I talked about red and green dots as a way of doing this. But there is more to it than this.

There's something else that gets in the way of our mission to describe the world meaningfully. This additional barrier is also cultural, and it's to do with the things that we're counting or summarising. As things stand, our counts, our summaries, our aggregations of activity, they tend to be of events as opposed to people. It's admissions or attendances or consultations or visits. And they're not linked to individuals.

But what we actually need to do is count people (and if we call them 'people' as opposed to 'patients' or 'service users' then we surmount the language barrier of how different organisations refer to them), and how many admissions did each person generate, and how many attendances for that same person, and—before that (or simultaneous to that)—how may GP consultations or community nurse visits or social work assessments for that person, and so on.

Here's the thing that got me thinking about all this. It's a diagram (shown at the top of this article, below the first paragraph) that was drawn as part of a Whole Systems Mapping project I'm currently engaged on in Scotland, and this diagram was one that I got when I asked a room of social care managers (and others) to sketch what "the whole system" looked like from their perspective.

This representation of the health and social care system as a pyramid— with expensive institutional healthcare at the apex and the person in their own home (supported by family and neighbourhood) at the base of the pyramid—is practically begging to be populated with person-centred data. In my mind, I now want to know about individual pathways: how does an individual go from being self-supporting in their own homes, perhaps making some use of Third Sector services (this is the second tier up, the bit of the pyramid with the assortment of disconnected circles, squares and triangles: a self-help group, or perhaps a Church-based community group), then moving onto GP consultations in the next level up, some community nurse visits (both of these would be on the healthcare left-hand side of the third tier up, and if the person was receiving homecare, that'd be the section to the right), and then details of interventions or events in the top tier: outpatient clinic attendances, A&E attendances, hospital admissions.

But the point is that this diagram is directing me towards looking at all this health and social care activity from an "individual patient journey" perspective. I'm suddenly a lot less interested in knowing—separately and in isolation—how many A&E attendances in total, how many district nurse contacts in total, how many homecare visits in total, how many GP consultations in total. (Yes, these event counts are important in some contexts, but they lack resonance unless they are placed in the context of a specific patient or service user journey.)

So that's the challenge: patient-centred data exhibits as opposed to event-centred data exhibits.

[13 January 2017]