Innovation in primary care and primary care networks – supporting discussions on new models of care

Understanding need

Good service design begins with understanding need, not activity. Yet understanding need, while a simple statement, is wrapped in misunderstanding, assumption and complexity.

Practices, PCNs and the wider primary care system need to be able to answer the following question:

• • What capability do we have to meet need, and where might we have gaps?

To understand this, and move the conversation away from a circular discussion on, for example, the share of a physiotherapist. We need to begin with understanding need.

If we are seeking innovation, one of the key aims is to avoid quantifying need in a manner that fits existing capabilities (see my blog on understanding the work of healthcare – Downham 2020). If we don’t, we risk perpetuating the same primary care model we have had for 50 years – something we should be seeking to avoid. So we have to ask different questions of our current data systems, and also use new systems to fill in the gaps.

The following is an example of a modest piece of work designed to stimulate discussions and innovated service models. It does so by using two dominant domains of need and then segmenting them. It is not designed to provide the ultimate scientific truth, but rather it is designed to provide just enough challenge and insight to base improvement decisions on. The domains are:

1. Level of Acuity
2. Social Context Pressure – click here (Downham & Easton 2020) for the implications of social context on primary care

Level of acuity (medical complexity):

The following chart details a clinician led review of 73 high attending* patients in a general practice in an urban area. These patients alone accounted for 1218 primary care interventions (F2F or telephone consultations) in the previous year – a relatively modest figure compared to some practices. It shows that 24% of patients were low acuity –yet these are amongst the highest attending group in the practice.

Only 6.8% were seen as clinically complex or high acuity such as having multiple morbidities or possible acute exacerbation. The inference here is that a large proportion of high attenders do not have high clinical acuity. So what might be driving their attendance?

*The audit excluded palliative, gold standard, care home and house bound patients.

The interplay between Acuity and Social Context Pressure:

It is the interplay between these two domains that define the complexity of a patient. For example a practice may reasonably be able to meet the needs of someone with medium or even high levels of acuity if the patient’s social context is stable. Conversely, a practice with a traditional model will struggle to help a patient with even low levels of acuity if the patient is under high levels of social context pressure.

The table below details the same sample of high attenders but with the two domains combined, producing a hugely useful segmentation. Note that a different practice might conclude different segments and proportions.

Each segment details the proportion % of the high attending group (n=73) that falls within it.

So what might this table tell us?*** 

The table shows four main segments – A to D. For the Lean / QI geeks out there (me included) these could be described as value streams. Their needs should be met in different ways.

A) Stable in context (≈ 46% of the high attenders): This is the second largest segment – a segment most suited to a traditional model of general practice. This group spans the full range of medical acuity, but crucially the patients are of stable social context. This work is Linear in nature (see this blog for an explanation of Linear work). It also includes high acuity patients that are stable in context – i.e. they have a network around them. The key ingredients in terms of practice capability to meet these needs would be:

• Regular longer appointments.
• Active management and reduction of unwarranted clinical variation.
• High continuity.
• Group consultations.
• In house mental health provision (for example CBT capability).
• Holistic consultations that are able to hold the relational, biographical and caring aspects of practice in addition to the bio-medical (Pratt 1995).

B) Patients of low and medium acuity and under pressure in social context (≈ 45%): In this case the majority are female**, over half are <=40 years old and 2/3rd of the primary conditions are depression or anxiety. This is where a predominantly bio-medical response will not break the cycle with patients. Practices need more comprehensive capability to help people under social pressure. It is complex problem that requires a networked response (such as a primary care network) with community services, social care and most importantly the community. Practices need the following capability:

• • Use of patient activations scores.
  • Social prescribing.
  • High quality MDT working (Malby 2019).
  • Complex care nursing (Malby 2019)
  • Developing a model of highly skilled generalists, rather than adding on new professionals, services and assessment (reducing handoffs and fragmentation).
  • Integrated and holistic assessments.
  • Asset based approaches (LSBU 2020) and developing links and assets with the community. This starts with hosting coffee mornings and identifying community champions.
  • An asset based approach that can evolve into a much more comprehensive network response to helping those struggling with their life context. This may include:
    • Network response that includes social care, the community, 3rd sector and schools.
    • Development of mental health capability (such as CBT and in house MH nurses).

**Sometimes, with this kind of analysis it is who is not showing up that is also very interesting. In this case why are males not coming in? Do they not suffer from anxiety or depression? Are they going elsewhere or just not accessing services?

C) The highly complex (≈ 8%): This group is the most complex. Note these are not necessarily people who are the most clinically complex, as it is predominantly social context (under pressure or turbulent life context) which is driving the complexity. In fact, in this example, none of these patients are high acuity. Key elements include:

• • Robust and integrated response (such as from an ICS) that features primary care, social care and the third sector.
  • Response is de-fragmented (radically simplified) and has high levels of relational continuity.
  • Work to ‘flip’ the clinical specialist model where possible. So the specialist consults into the patient and generalist (for example the GP who has the relationship with the patient) rather than the generalist referring the patient into a specialist setting.

D) WHO ARE THESE?: A small but interesting group. These patients are low acuity and are stable in social context. What is driving their attendance? What don’t we know about them? Have we inadvertently created dependence? In this case a little further fact-finding is required – urgently.

To conclude:  

If the same segmentation is focused on a smaller sample, the top 20, the pattern intensifies. It is the social context pressure that intensifies, rather than the cohort increasing in acuity.

The benefit of beginning to take a simplified segmentation approach is that different parts of the system don’t mix up each other’s work. It is clear what each part of the system should be concentrating on. Without clarity, the natural tendency for the system is to act down. With clarity, we can focus the on the work that each level is best placed to do.

Practices are amazing things. They can support people who are high acuity if the patient is stable in life context (they have a network around them). Some patients do, but many don’t as this analysis shows. It pays to invest early in models that help with life context pressure. If not, these patients will be attending A&E on a regular basis.

It is also clear that to break the cycle for all of these segments, many practices need a change of model. The current, traditional, overwhelmingly bio-medical response will not change things. Which is why the current fixation with access (more appointments) is so concerning; greater access does not necessarily mean better outcomes if the model of care remains the same. This is about new responses to helping people.

These segments can be adjusted or drilled into to suit priority patient groups. For example, focusing just on Type 2 diabetics or COPD patients – this can help start the ball rolling given the funding often tied up with helping these specific groups.

The next step is to use this segmentation to identify gaps. For example a traditional model (with limited social capability) may find there are gaps in capability to meet need in all but segment A. So a gap in capability for over 50% of their high attenders – which is why, perhaps, they attend so much. A practice and wider system can then plan to address those gaps in their model.

While this is looking at high attenders, it gives us important learning for other cohorts of patients. The same gaps in capability will feature for other patients – even if they are not high attending (yet). Changes in model will benefit those patients who are under social pressure and under the radar – contributing primary care priorities such as anticipatory health, tackling health inequalities and COPD prevention.