Making data count

by Sam Riley

18 May 2018


Statistical process control offers a more powerful way to analyse data than traditional ‘red, amber, green’ approaches

Did you know that if you have two numbers, the second one will be higher, lower or the same as the previous number?  Of course you did – this is a simple fact that we all understand.

When you leave the office every day – do you always arrive home at precisely the same time?  I know that this is not the case for me and I imagine that the vast majority of people have the same experience.  In our everyday lives, we acknowledge that some things simply vary due to factors outside our control. 

For many people, though, in their working lives they forget that natural variation exists.  ‘So what’ you may think.  Well the problem that frequently occurs when you forget this is that you can over-react to data that is simply normal.  Whether it’s A&E performance this week being worse today than yesterday or the spend on a particular drug being higher this month than last.

In many people’s working lives they regularly look at two data points and if the second data point has moved in the wrong direction, panic ensues.  Investigations and action plans are demanded – and provided.  A reason needs to be provided for that second number being ‘worse’.  What if the answer is as simple as ‘the change in the data was due to natural variation’? 

How much time could we save if we had a way of looking at data which enabled us to understand when something unusual is going on that merits investigation?  Then we could spend our time focusing on those issues that need to be understood and where action needs to be taken to improve the situation.  We could stop ‘investigating’ all of those instances when changes in the data were simply normal.

Well, there is such a technique that helps us do precisely this.  It is called statistical process control (SPC).  This technique was originally developed in the manufacturing industry, but it lends itself extremely well to healthcare data including finance data.  SPC is a technique that is underpinned by science. There is a wealth of published research that shows better decisions are made when data is looked at over time and especially when SPC is used.

So, why don’t more people in the NHS use this technique?  No doubt there are a multitude of reasons.  People may not be aware of SPC as an approach.  Even if they are, they may not have access to the knowledge and tools to analyse the data in this way and make the correct interpretation.  They may have seen so many reports with red, amber, green (RAG) over the years that they can’t imagine that there is a different (and more powerful way) to analyse data.

With NHS Improvement’s recent publication of Making data count, I hope that more people will understand the pitfalls of two-point comparisons and RAG reports and learn about the value of looking at data over time and using SPC. 

This interactive guide gives people the knowledge and tools to be able to create and interpret SPC charts and encourages closer working between analysts and decision-makers.  The intention is that Making data count will support more effective decision making in the NHS and make a real difference to patients.

Read Making data count on the NHS Improvement website or follow #plotthedots on Twitter