Why and how data will save the NHS

Yesterday (1 July) the London Datastore published this blog post written by our CEO, Jonathan Lewis on their website. This is an important message so we have reproduced it here to make it possible for more of our stakeholders, partners and patients to read it:


Today Bromley Healthcare, the south London community healthcare provider I have led for the last four years, has become the first to publish clinical data on the London Datastore. We have done this because we believe that data will save the NHS.

I realise this is a hard concept for the many in the NHS to come to terms with because ‘data’ does not sound like a ‘clinical’ solution, and because historically the NHS has focused on changing structure to achieve better results.

Of course an MRI scan is in part a very sophisticated data gathering device, but the complete acceptance that good data is key to clinical decisions has not led to a corresponding recognition that it is equally important to informing changes to the system.

In many ways it is the systems of care as much as the individual care ‘moments’ that need sorting out. Now anyone from the NHS reading may well disagree with me, but if you look at the time, money and effort spent on reforming the NHS a lot more goes into structure than information. The NHS has a number of well-known challenges, all of which will benefit enormously from the better use of data.

Variable standards of care

Variable standards of care are comprehensively detailed for specific conditions such as diabetes in the NHS’s own publication, NHS Atlas of Variation in Healthcare Series. There is huge variation in care every time a clinician treats a patient and at a system or ‘pathway’ level in the multiple encounters that a patient might have with clinicians over a course of their treatment.

The complexity of an ageing population

An ageing population puts more pressure on a system that has no more money. Quite apart from the pressure of more clinically complex patients, the ageing population adds a much larger system management challenge – organising complex care around the individual.

This supply chain management skill – especially in terms of coordination between the different care providers – is not yet something the NHS is expert at. Just ask anyone who has had to sort out complex care for an ageing relative.

The NHS has tended to respond to this challenge by merging parts of the system (which will never work – but that is another blog) at the expense of building its supply chain management skills – a requirement that cries out for good data.

Mass system failure

We are starting to see the system fail en masse against key central targets including solvency and accident and emergency targets. This is a direct consequence of these increasing pressures and gap in skills and data.

What is the one thing that will make this better quickly? better data and a set of clinical, case and supply chain management systems that are fed by this data.

Before we go on, let me give you one example of the massive scope for data to improve both the quality and cost of care.

A leg ulcer is a wound on the leg that does not heal in a month. Many leg ulcers are not treated properly and become chronic, costing an approximately £10,000 per year to treat. A well-treated leg ulcer can be fixed in about six weeks for a cost of around £1500. It would not be an exaggeration to say that leg ulcers ruin lives – they can be smelly, require the patient to wear intrusive bandages and often discourage people from leaving their homes. I have heard of couples splitting up after 40 years of marriage because of leg ulcers, and in many cases this is totally unnecessary.

There are tens of thousands of these leg ulcers around the country that are being badly managed. How do you cure them? Consistent high quality care based upon good information. If we look abroad, there is a Spanish hospital group – Ribera Salud that through sophisticated data systems claims to have improved quality and reduced costs by 25%. This is the level of improvement sophisticated data can produce.

How do we use data to reduce variability in care?

The core question of course is how do we do this? How do we use data to reduce variability in care? The first thing you have to do is collect the data.

What I believe we have done at Bromley Healthcare that is unique, is introduce a system that records our clinical activity at an extremely detailed level. Each relevant part of each procedure, for every single one of our clinicians is recorded.

People are our key resource

This is based upon the premise that the quality of the service you can provide is only as good as the quality of data you record. Our key resource is our people. A very detailed understanding of what each and everyone does and what each and everyone achieves should enable us to understand what works best and enable us to standardise, and monitor, how individual patients and conditions should be managed.

This approach is producing a huge amount of data. Today we have started to publish some of this data on the London Datastore. The first data sets show anonymised appointment information from April 2014 to March 2015 for our major services.

The data includes patient ages, gender, outcomes, number of appointments within the time period, reason for discharge, number of patients seen by each healthcare professional and the location of those appointments. However, this is quite high level data.

Big data partner needed

What we need now is a big data partner to help us analyse the very detailed data about how exactly patients get treated, and what the impact this treatment has on their outcomes.

This very detailed recording of activity and outcomes by individual should allow the development of a whole new generation of clinical management tools. I am not sure this next generation set of tools exists yet.

Certainly some of the big managed care companies have very sophisticated systems for ensuring that the patient gets the right care at the right time, for instance that a nurse comes when needed. What I don’t know is whether anyone has a quality assurance system that ensures that every nurse who comes delivers a nursing intervention to the same high standard.

Answering the call for transparency in the NHS

Quite apart from recording this data, being transparent by publishing this data has an inherent value. In February 2013 the Francis Report – published after the inquiry into the failings at Mid-Staffordshire NHS Foundation Trust – called for more transparency in the NHS. We want to get the ball rolling on this.

The NHS is paid for by taxpayers – they should be able to see with complete clarity what they are getting for their money, and the information that public gets should be raw, albeit anonymised, data sets so that it can be analysed.

Better, more consistent, lower price services

What do I expect to get from this dive into data? Clinical management systems that provide much better and more consistent services at a much lower price to the tax payer. It strikes me that these clinical management systems might qualify as a disruptive technology, according to Clayton Christensen’s definition, because it will make us far more competitive within the existing payment mechanisms of the NHS system.

To use a crude analogy, antibiotics don’t seem to work that well any more. Likewise the generic organisation medicine of the NHS – the structural fix and top down savings plans are not reviving the patient from this torpor. Information and a set of new systems built upon it will be the new wonder drug for the NHS.

The Five Year View (NHS England’s view of future models of care) is a powerful disruptive campaign putting change more into the hands of people with no interest in the status quo, but it will only achieve its full effect if it is comprehensively provided with first rate information, and a new generation of analytical systems based upon it.

This blog was first published on the London Datastore.


  2 July 2015