Mental health is a bigger driver of cost than chronic illnesses

It has long been understood that mental and physical health go hand in hand, and that to treat one, you must fully understand the other. Thus it stands to reason when looking from a population health perspective that segmentation should address both physical and mental health.

However, there has been a constant complaint that the level of data available for mental health is poor. In our view, this is a wrong impression.

When combined with data sets from the rest of the NHS, mental health data can provide rich insights which can be used to understand patient needs, organise care, and inform payment

The quality of data that is available in mental health has dramatically improved thanks to introduction of cluster costing. When combined with data sets from the rest of the NHS, mental health data can provide rich insights which can be used to understand patient needs, organise care, and inform payment.

In fact, the top thing to secure parity of esteem between mental health and physical health would be to make robust analysis of mental health data routine, comparable to what is done with acute care. This would ensure that people with mental illness get the resources they need.

Over that past year we have worked with one of the best integrated data sets in the country: the Kent Integrated Dataset. The KID is a patient-level linked dataset covering almost all the 1.8m people in Kent and Medway, combining data from primary care, secondary care, mental health and social care.

Each record is pseudonymised to ensure the data is unidentifiable. These data allow the understanding of a person’s mental and physical health conditions, and their activity within the heath and social care system, and attributed costs. There is so much data, it is impossible to understand it without advanced analysis.

Population segmentation

To understand the needs of the population with mental illness we set out to segment the population into groups that make sense to clinicians, can be addressed as discrete groups with relatively common needs, and are analytically robust.

We explicitly sought to understand the impact of comorbidity, and to do so we grouped people based on physical health (Mostly healthy, 1 long term condition, 2 long term conditions, 3+ long term conditions), and mental health status (Mostly mentally healthy, depression or anxiety, severe and enduring mental illness, dementia). Based on the population and spend in each segment we could then calculate the spend per head.

In addition to this segmentation [see graphic 1], it is possible to use the integrated dataset to understand spend and activity, by segment, in much more detail [see graphic 2], eg bed days, accident and emergency attendances, number of mental health contacts, number and type of social care contacts. This is the first time, as far as we know, that an analysis like this has been done.

The results show that mental health is an even bigger driver of cost than age or chronic disease. A person with three physical long term conditions requires 10 times the spend of someone who is mostly physically and mentally healthy. But a person with severe and enduring mental illness requires 40 times the spend of someone who is mostly physically and mentally healthy.

People with dementia and SEMI account for less than 2 per cent of the population but one seventh of spend. More broadly, people over 16 with a mental health condition account for a sixth of the population, but almost two fifths of system spend. These figures account for the spend on people with mental health illness across all services – as opposed to simply those services branded “mental health”.

Profound implication

The implication of this is quite profound. It shows that by focusing the “parity of esteem” debate on matching growth rates in acute at about the level of spending (closer to 10 per cent than 38 per cent) the health service is dramatically underplaying the massive impact that mental illness has.

We can also incentivise the system to provide earlier preventative care

Dementia and SEMI are very complex areas, with dramatically higher spend than other segments of the population. We need to understand how to best design care models to meet the needs of these segments.

As we understand the consumption of resource and the activity and spend behind it, we can think about the pattern of activity today and how it should differ in the future. This includes providing new forms of care, for example, emphasising on crisis more, and supporting recovery for those suffering SEMI in the community. We should also be able to demonstrate that investment in early interventions both improves care and lowers overall costs.

The segmentation can also support a much better payment model, based on demand and need, which provides the resources required to deliver best practice package of care with an agreed, costed, amount per head for each segment of the population. By identifying patients with high demand (and, therefore, costs) we can target care to improve outcomes. We can also incentivise the system to provide earlier preventative care.

The NHS is blessed to have a population approach to health, with tax funded healthcare free at point of delivery and probably the best data environment in the OECD. We can use these to deliver better care for those suffering from poor mental health.