Our Latest Newsletter (22/11/2021)
In this ‘new-normal’ world, the EQUIP Team, EQUIP coaches and Thinking Space team want to wish you old-fashioned pleasures, happy memories and all the joys over those holidays.
Our gift for you is this useful tool, often overused and difficult to interpret: the funnel plot.
Enjoy the readings and for any questions please get in touch with Adele (firstname.lastname@example.org)
What is a funnel plot?
These are graphs that visualise variation within a system and identifies statistical outliers within it.
The y-axis represents the activity measure (either a proportion or rate per 1000 patients) while the x-axis represents the denominator.
These are represented as follows:
· The x-axis can represent the potential maximum frequency of the activity being measured when the funnel plot is measuring proportion of activity
· e.g. the x-axis showing the number of booked appointments while the y-axis displays DNA percentage rate
· The x-axis can also represent the population when the funnel plot is measuring rate per 1000
· e.g. The x-axis showing the practice list size while the y-axis displays referral rate per 1000
· Dots on the funnel plot represent GP practices.
· The middle straight line presents the average for all practices
· Lines either side of the mean represent the confidence limits (as seen in figure 1)
Why should I use a funnel plot?
Funnel plots allow the user to interpret the variation within a system.
If you wanted to compare vaccination rates between a group of practices, you may be tempted to simply order the practice by their vaccination rates from highest to lowest and look to focus on improving the practices on the lower end.
While this is feasible, the problem with this method is that It does not account for:
·The account for the number of patients legible for vaccinations in that practice
·The differences in practices vaccination rates being due to the normal variation that exists within systems
How to interpret a funnel plot?
1) Identify your practice on the funnel plot.
2) Does your practice sit outside of the +/-99.9% control limit?
a. If your practice does sit outside of this, then the practice is showing special cause variation and is a statistical outlier when compared to the other practices.
i. Special cause variation: variation that is caused by special circumstances or events that are out of the ordinary.
b. If your practice sits within the +/- 99.9% control limit, them your practice is showing common cause variation and is not an outlier.
i. Common cause variation: variation caused by ‘normal’ events
3) The smaller the x-axis value, the wider the confidence limit because our confidence in the rate of activity displayed being an accurate reflection on how well the practice is doing decreases when we have a smaller sample
EQUIP – The WHY we should embrace QI
Click edit button to change this code.