Quality Improvement Zone

A funnel plot is a chart that shows the measure of interest on the vertical (y) axis and sample size on the horizontal (x) axis. A line showing the average of the measure, and ‘funnel limit’ lines are also shown.

Points above or below the funnel limits differ from the others by more than would be expected by chance.

A funnel plot is a useful way to show comparisons between different units. It helps to identify where there may be special cause variation.

It is helpful if you have data relating to different teams. For example hospitals, schools, NHS boards and Local Authorities. These examples are of different sizes. You want to avoid misleading rankings that don't account for this. The measure you plot would have a denominator (e.g. rate, proportion or average).

You might want to use it to identify which units you might need to focus improvement work on. Including which ones you might have something to learn from.

How to interpret it

Where a point is outside the funnel limits this indicates a difference that would be very unlikely if only common cause variation was present. This helps to focus inquiry. You may wish to look first into whether there are any reasons why the data might be different. For example are they using the same definitions? And then whether there is something else going on.

If a unit has better results than others, then there may be something to be learned from them that can lead to improvement across the system. If one is worse, then it could be a good place to focus on improvement.

How to create it

You can also create funnel plots in excel using most QI software or by doing the calculations yourself and using a scatter plot. Ask your local data expert if you’re not sure.

If you use QI charts software, you can sort your data by sample size and create an SPC chart appropriate to your data.