Close up of scientist hand holding tube with leaf-259452-edited.jpeg

The True Cost of Bad Data

In the healthcare sector, acting fast saves lives.

Matthew Swindells is NHS England’s national director for operations and information, so it’s fair to assume he knows what he’s talking about. Speaking at the recent E-Health Week conference in London, he cut straight to the chase when he stated that, “Your predictive algorithm is worth nothing if it is based on crap data.” (1)

His point was that while artificial intelligence and predictive algorithms are set to transform healthcare and patient care over the coming decade, they’re a pipe dream until the quality of the data being used improves substantially.

Mr Swindells went on to describe the NHS as an organisation that wastes time, resources and money collecting multiple, often conflicting, versions of the same patient data. Yet while that sounds damning, he was just being truthful. The NHS is no more or less guilty of this than many businesses and organisations. The major differentiator between healthcare and retail, for example, is that conflicting data sources in the NHS could delay treatment and affect people’s lives. That’s why Mr Swindells was right in highlighting the need for the NHS to get better at “linking data and sharing data, and getting the data quality right.” As is reflected by the "Getting it Right First Time (GIRFT)" national programme.

We find it refreshing and encouraging that such self-critical concerns are being voiced openly. Mr Swindells should be applauded for noting that exotic technologies, automated systems and stunning on-screen visualisations are meaningless unless they are underpinned by an accurate, ‘single source of the truth’. Without this fundamental foundation, they’d just be the sweet butter icing on a stale, inedible cake.

Garbage in is always going to equal garbage out, no matter how well you try to dress it up. We know this because time after time, we have helped forward-thinking organisations to deliver outstanding projects, leading to efficiencies and improving healthcare services.

With healthcare trusts, such as Hull and East Yorkshire, NHS Highland and South Warwickshire, we have helped to reduce the time between referral and treatment by keeping every patient highly visible within the system. So rather than shuttling copies of data around the trusts from department to department, all stakeholders now have trusted, accurate data from a single version of the truth.

NHS Digital’s vision is that by 2020, “care professionals will have timely access to the information, data, analysis and decision-support systems that they need to deliver safe and effective care” (2). Yet the collection and eventual AI analysis of patient data across the entire NHS is a non-starter until the issue of data quality is first addressed at a trust-by-trust level.

Reliable data can no longer be a technical matter, left in the hands of the IT department. With so much at stake – literally lives in the case of the NHS – the sourcing and handling of data must become a strategy, with every level of a business or organisation – especially those at the top – being active participants and advocates.

Of course, the technical demands of a fully-integrated data system are a challenge for most businesses or, in this case, health trusts. Making the ‘pipe dream’ happen is possible though. What if each individual trust was able to configure shared, linked systems that meet their own individual operational requirements? Mr Swindells’ vision to get better at “linking data and sharing data, and getting the quality right” can be turned into a reality.

So can and will this happen to the NHS, or any other organisation facing similar issues? Yes and yes. In our forthcoming blog, ‘Single version of the truth – turning a pipe dream into reality’ we’re going to break it down into five points that can be actioned by your business or organisation.

Subscribe to our blog to have it delivered straight to your inbox.

Feel free to comment your views below.

(1) Time wasting over data arguments 'outrageous', says NHS England director:

(2) NHS Digital 2020 Vision: