Everyone today is talking about population health management (PHM) and at the heart of any credible approach to PHM is the concept of population segmentation. This is an analytics activity in which Sollis and our partners are very active at the moment. One of those partners is the Johns Hopkins University (JHU) with whom we have a ten year collaboration, the result of which is the most mature implementation of the Johns Hopkins ACG® System here in the UK.
Sollis and Johns Hopkins describe population segmentation as follows:
“…the purpose of segmentation is to enable specific groups, or cohorts of patients to be identified and explored in more detail, and ideally to define groups of similar patients who may be suitable for a particular intervention or new care programme.”
— Understanding Population Health, Sollis and Johns Hopkins University, 2017.
As I tour the country sharing Sollis population segmentation stories and experiences, I have had plenty of time to reflect on the topic and offer here some thoughts, which are intended to generate debate as much as insight.
1. Segment for cohorts but remember this is about the patient.
Population segmentation is a methodology that supports patient centred care. Rather than focus on diagnosis and single condition diseases, we segment populations because we need to understand the holistic needs of individual patients within the cohorts identified. The focus is on the person and not the various institutions with which the patient may come into contact.
2. This is not a cost cutting exercise.
With all the noise surrounding Accountable Care Systems — now re-branded Integrated Care Systems — there is a danger that many will see population segmentation as being inextricably linked to the need to cut costs.
While population segmentation is certainly concerned with achieving a better understanding of cost, it is not the sole consideration. Population segmentation is principally concerned with how best to design services that deliver optimal health care for the populations identified. The primary consideration is the delivery of quality care. Quality trumps cost. If we accept that the current ‘method’ of organising services — based around institutions not patients — is inherently wasteful, then surely we all have a duty to explore alternative service delivery models that help reduce waste and ergo cost.
3. The healthcare system is not the only system out there.
There is much talk today about the importance of learning systems. I agree, learning systems are vital if we are to craft a sustainable healthcare system for the future. I would contend that some of that learning needs to come from outside the health and care system. Other industries have long used market segmentation techniques in order to satisfy the demands of the end customer. Population segmentation holds much promise for the development of targeted services designed around patient centred needs.
4. Segmentation is about everyone.
Population segmentation includes everyone in the population. While there may well be a legitimate need to prioritise resources where need is greatest, this is not to ignore other population segments where there exist legitimate claims for resource allocation. Population segmentation should not be about rationing. Population segmentation is inclusive and includes healthy populations, multi-morbid populations as well as those nearing end of life.
5. It’s a framework for planning.
At its heart population segmentation is a methodology for better planning and specifically the planning of targeted resources. Ad-hoc service planning based on hearsay and fairy tale should be consigned to history. We need to sweat data assets in order to generate insights for strategic planning purposes.
6. It’s a framework and not a panacea.
Population segmentation is a methodology and not a panacea for all the ‘ills’ currently visited upon the healthcare system. Different population segmentation models exist — National Association for Primary Care (NAPC) and Bridges to Health being just two — but neither need be written in stone. Both exist as templates for testing different ways of segmenting. Different health and care economies may require tailored methodologies. If we are to design optimal health and care services for populations, then customisation will always be a requirement. NAPC and Bridges to Health provide excellent starting points.
7. The needs, wants and desires of patients and carers is paramount.
Never lose sight of the ultimate goal. The goal is to arrive at co-ordinated care that supports patient engagement and takes into account the needs, wants and desires of both patients and carers. Never lose sight of the fact that what matters to them is the most important consideration and that this extends to the outcomes by which we measure success.
8. You can’t do this without the data and advanced analytics.
It’s not possible. However DON’T PANIC. The healthcare IT industry is notorious for talking up the next big thing, be it Big Data, Artificial Intelligence (AI) or Blockchain, but significant results can be achieved through access to just secondary care and primary care data sets, the acquisition of which (IG not withstanding) should be relatively straightforward. Don’t try to boil the sea. Start small and build out.
Population segmentation activities are typically associated with health systems that take on risk for the health outcomes and costs of target population cohorts. Segmentation approaches allow such systems to identify patients who might be able to avoid health problems if given greater support. In this sense there is a tendency to link population segmentation approaches to the work of Accountable Care Systems which in the UK is a controversial topic.
However if you strip the politics away, at its heart segmentation is about providing insight into the specific health needs, wants and desires of populations and sub-populations. It therefore offers significant promise as a means of ensuring that services are tailored around the specific needs of clearly discernible and impactable patient groups. It is a key building block of designing Integrated Care Systems (ICS).
Population segmentation approaches are completely consistent with the aim of designing integrated and patient centred care, where the holistic needs of the patient are front and centre. For this reason alone such approaches are to be applauded.
Enter your email address and we’ll send you a link to the introductory guide, Population Segmentation.