Bringing People Together with Data to Solve Health Challenges

I feel compelled to write and champion the cause for arguably one of the most important public health crises that is upon us. Right here, right now, health systems are straining at the seams with a never-ending proliferation of morbidity related demand. Long term illness is sucking up resource like locusts feeding on a crop of lush green vegetables. The state of the nation’s health (for many countries) has moved from risk to reality. And we all keep talking about it. I would like to share my views and solutions from visiting, seeing and talking to health professionals across the globe. The call to action is strong and achievable. We just need to move fast.

Public Health England produced a Health Profile in 2018 that slaps you around the face with some frightening facts:

• The proportion of population 85+ is 2.7x greater than in 1971
• The 85+ are the greatest users of health and social services
• Life expectancy has slowed and there is a growing inequality with deprived areas showing the lowest life chances – 30% of UK living below the Minimum Income Standard
• As mortality from heart disease, strokes and major cancers fall, dementia and Alzheimer’s are rapidly on the rise
• Long term conditions such as diabetes continue to rise, 3.9 million in 2017 and a projected 4.1 million in 2020, rising to 4.9 million in 2035
• Low back and neck pain, as well as skin diseases are high demand issues
• Mental ill health has been increasing and 18.9% of the population present with common disorders
• Although smoking has reduced (great success), obesity is not falling. The NHS spends 6.1 billion treating overweight and diabetes related ill health

Absorb this: 70% of total health and social care spending boils down to long-term morbidity issues. Now factor in the trends seen, you can see how a creaking megalith is sliding into the grip of grim reaper himself.

And what about the impact?

Demand on our health and social care providers has risen significantly. Demand is outstripping the capacity to deal with it. Weeks to get a General Physician (GP) appointment and excessive waiting times in Emergency Department (ED) are normalized. This is just scratching the surface. Dealing with such high demand, often with low severity levels, has created a blizzard of risk which can prove hard to prioritize and intervene to spot escalating cases that really need acute care.

Talking of acute care, ED has become the entry point for everything including the kitchen sink. Folk are often using ED as the first point for minor issues which blocks the system for genuine users that really need it. ED is also having to cope with persistent and frequent visitors which are a consequence of the population health issues outlined in the bullets above. For example; elderly people with long-term morbidity issues often require complex treatment that can result in lengthy hospital stays. Too often you hear of the term ‘bed blocking’ where social care plans are not in place and the patient stays in the bed without being released at the earliest opportunity. Social care is straining at the seams too. Likewise, mental ill health patients require specialist beds and extra resource to deal – even from other agencies. It is not uncommon to see as many Police cars as ambulances in the A&E car park!

And what about our health professionals? There is a real and present issue for long-term stress related illness, retention of staff and overall wellbeing. The current picture is not sustainable.

And the solutions:

Bottom line is that we need to get upstream and reduce morbidity. How?

People and technology need to come together. Health and social care agencies need to come together. We need to use a data driven approach that underpins decision making and early intervention approaches. Problem solving ‘together’ needs to be a value. We need to truly understand our population health from a strategic to a tactical level (individual patient). We need to understand demand, risk and escalation of risk through patient pathways across agencies. We need to prioritize according to harm. We need prescriptive strategies for the cohorts created from the risk stratification process e.g. avoid hospital admissions for rising risk patients. We need to remove waste demand – how many GP and hospital no shows occur? We need to create a culture of treatment and monitoring at home for long-term low-level morbidity issues. There also needs to be a forensic approach to matching demand, capacity and capability 24/7.

Where do we start?

Leadership. Collective leadership. The integrated services approach needs a strength of commitment and tangible action that joins the health and social care system together. There are mechanisms in place e.g. Integrated Care Partnerships (ICPs), but politics, people, process and systems are letting it down. There are too many silos and chiefdoms in a single agency – let alone a whole system. It needs to be simple with exceptional leaders in place.

We need to ban the words “can”, “hope”, “could”, “might”, “difficult” - and any other word that does not actively support the critical mission the issue deserves.

We also need to share and accelerate good practice. On my travels I see some fantastic examples of where professionals are taking fight to the problem. However; very few professionals from other geographic areas have seen them. Why?

What exciting solutions are emerging?

Andy McGivern and Ria Powell from West Cheshire NHS Clinical Commissioning Group (CCG) are in the process of pulling together and working towards a Population Health Portal. They have approached it from a commissioning angle, and the priority was to max out on all data flows and build a full pathway of patient contact points. (They now have patient level data flowing for GP contacts, secondary care, mental health (by team\service), community care (team\service), ambulance, 111, prescribing info and in the next few weeks social care contacts. From this, they can understand:

• Risk stratification, including a one stop shop to include; latest test results (Primary care driven), prescribing data at patient level and other relevant conditions
• Frequent Fliers (all types not just secondary care focused)
• Flags for
  • Care homes\Cancer patients\Facility\Falls\LTC\Palliative care patients
• Geo-analytics to track all the cohorts\admissions
  • linked the above to include an overlay demographic page for wider determinant data (deprivation, household income, academic results, fuel poverty, mortality rates etc)
• A financial resource value for all contact points
• Segmentation for different hierarchies to serve the needs of CCG\ICP\Primary Care Network and GP Practice
• Referrals

They have built most of the above with a team of clinicians from local partners. It is aimed at supporting the way that clinicians work – collective leadership in action. The system has been adopted by GP colleagues, with a good coverage\usage rate

Not stopping there, West Cheshire have an ambition to include; Patient Registries (Currently working on EOL, Diabetes and a Gynae type one); Predictive Analytics; and getting to understand what areas\concepts we are missing from other partners.

And there is more great work to be shared:

Rob O’Neill and Matt Heys from Morecambe Bay NHS Trust have been developing and expanding their population health development. Clinical pathway dashboards are in place (stroke, respiratory, diabetes) so that users can see relevant population health information split by various disease specific filters (e.g. what's our population that are anti-coagulated look like).

One of the big things they have been doing recently in the population health space is a project on mental health. They are using data science techniques to predict/identify patients who may be at risk of conditions such as depression and also patients who are at risk of exacerbation (e.g. self-harm, ED attendance, attempted suicide). They have used Qlik Sense to build a dashboard to aid with Exploratory Data Analysis (from the data science point of view) but it's also a great tool for clinicians and care coordinators to be able to slice up their population and see how different variables affect mental health risk profiles.

This screenshot shows how many patients in their population have a diagnosis of depression and are currently on anti-depressants with various charts showing split by BMI, Gender, Age Band, Deprivation Index Decile, Postcode Sector, leading on to integrated analysis (primary and secondary care data linked together) to show how the risk changes depending on number of ED attendances in last year, open access plans (i.e. waiting lists) and inpatient admissions in last year. E.g. From this you can see that having just one ED attendance in the last year increases the likelihood of a patient being diagnosed with depression from 13.9% to 17.8% or that obese patients in the population are 23.7% likely to be diagnosed with depression.

Call to action:

The message is simple. Our leaders need to adopt an authentic collaborative leadership style (no politics) beyond authority and borders. There needs to be a forensic analytical approach to understanding the issues and opportunities. Actionable insight needs to be built into commissioning plans at a strategic level. At a tactical level, the data needs to drive a multi-agency early intervention approach – built into the tasking and delivery mechanisms. Things need to accelerate, and folks needs to get on with this right here, right now.