We have a crucial role to play

The Deputy Director of Business Intelligence at Imperial College Health Partners looks at how NHS data analysts are helping to transform health and care:

They may be an army of keyboard warriors tucked away in the vaults of NHS buildings from Penzance to Carlisle and everywhere in between – but our hugely talented data analysts are in pole position to change the game for patients now and in the years to come.

These analysts working in hospitals, CCGs, local councils and other public agencies are our version of Google Deep Mind and should be classed as such, equally as capable and the back bone to creating a health and care system fit for the 21st century.

You may have heard the latest buzz words around the NHS – ‘population health management’ or PHM for short – this rudimentary put means taking a data-driven planning and delivery approach to care to achieve maximum impact; helping system leaders best plan resources for now and the future.

This PHM approach is where our talented data analysts can make a huge impact on direct patient care; identifying and finding creative ways to tackle the different health risks facing us all, literally changing the service that people receive and improving their health as a consequence.

There is growing global evidence of the positive impact of PHM approaches – in Staten Island, USA, for example, they saw A&E attendances for asthma plummet after spending 10% of their budget as a grant to improve housing standards and purchasing vacuum cleaners for those in need in their population. Channelling resources in to cleaning the homes of those having repeat admissions, rather than offering higher doses or more drugs, had a profound effect on lowering admission rates.

As data analysts in the NHS we proudly come to work to use our skillset, like the whole of the NHS workforce, with the priority of ensuring that our work positively impacts and helps patients live healthier lives. I believe that PHM unlocks, starting with data analysts, us making this change happen for the whole population.

Data analysts understand the codes and algorithms needed to segment a population and help ask the right questions to obtain a useful result, our clinical colleagues can then apply their medical expertise to understand these patterns enabling solutions through these powerful collaborations.

The NHS is currently running a series of pilot sites testing the power and scope of what can be achieved using these collaborations of the expert knowledge of data analysts, public health specialist and highly skilled clinicians and this will be rolled out over the coming months.

The pilot has not been without challenges as linking local datasets is a tricky task, but where it has been achieved we have been able to gain new and extremely useful insights to help improve care for patients.

Let’s take the example of a groups of patients who all have the same diagnosis but who all use NHS services differently with a subgroup using services a lot more frequently and as a result use more resources. There has been excellent work already done by colleagues to further explore the needs of these patients and to then enable resources to better meet these needs.

Through PHM this work can now go further; through understanding populations wider or social determinants of health and then to target resources at gaps in care, or areas more amenable to change, could make a bigger impact and offer more value to the whole health system. This is now more possible due to integrated care systems, where council, public health, voluntary sector, GPs and hospitals are all working better together, making it far easier to collaborate on data and system change.

By linking datasets, we could explore information on these patients, and use the linked data to help us to discover more about them – for example are they all from a care home population? It could mean that the frail and elderly in a particular area are being admitted unnecessarily too frequently and therefore a new care model could be considered to address these issues.

In the past, information on where those elderly people lived may not have been linked to the hospital data available and therefore a wider picture revealing a trend would not have emerged for analysis. Local relationships, and data sharing, mean this is now possible.

The solutions are the next step – use of skype with the local hospital, upskilling of care home staff, extra helplines to provide reassurance. These are all models of care being provided at multiple places in the country which could be used if we knew that the problem in that area was excess referrals to A&E from care homes.

The fact is that we can always improve care pathways but for many individual treatments or interventions there may only be marginal gains, because the treatments work well and the pathway may already almost fully optimised.

But if we see the same patients going to hospital for the same exacerbations over and over again, the improved local organisational relationships we now have mean we can start to build a picture around that cohort of patients and address the wider determinants of health.

For example, people with COPD who are admitted to hospital frequently may find more drugs are not the answer. They may live in a house that’s damp and mouldy and all they need is support with their electricity bill in winter – it’s better for those people and would cost us less as a system to address the damp and mould issue rather than the costs of repeated admissions and drugs.

Data analysts hold the key to unlocking the wealth of possibilities and as the PHM approach comes to a town near you soon, I’d encourage you to raise your hand and get involved because it really will mean shaping patients’ futures for the better.

  • To register as part of our on line Population Health Management network and stay up to date with the latest information please email:
Andi Orlowski

Andi Orlowski is a health economist and leads the business intelligence function at Imperial College Health Partners which supports the North West London Sustainability and Transformation Partnership.

He is particularly interested in population health analytics especially the concept of impactibility modelling and finding those people most amenable to an intervention.

Andi is also a Senior Advisor for NHS England on Population Health Management in the Operations and Information Directorate. He works with STPs and ICSs across the country advising and working with them on population health management.

Leave a Reply

Your email address will not be published. Required fields are marked *

One comment

  1. Mohammed Absar says:

    This way of linking data from across the health system and focusing on cohorts of similar patients to design interventions is absolutely the right way to go. More drugs is definitely not always the right thing to do!

    We carried out a number of such projects at Tower Hamlets in east London working with GP, mental heath and social care colleagues from the council, with support from UCL to design focused care programmes and personalised budgets for some patients.

    More Information here: