A Pareto chart is a type of bar chart in which the factors that contribute to an overall effect are arranged in order from most frequent to least. This ordering helps identify the "vital few" — the factors that warrant the most attention.
It also includes a line showing the cumulative % (so you can see easily for example that the top 3 causes account for 80% of incidents).
According to the "Pareto Principle," in any group of things that contribute to a common effect, relatively few account for the majority of the effect.
Using a Pareto chart helps a team concentrate its improvement efforts on the factors that have the greatest impact. It also helps a team communicate the rationale for focusing on certain areas.
For example you might want to look at
They are a useful tool not only when identifying opportunities for improvement, but also in drilling down to understand the reasons for special causes on graphs showing counts, percentages and rates.
You may have data already collected, e.g. from incident reporting, or you might have to gather data specifically. You will need to arrange your data in a table showing number in each category of interest.
Pareto charts work best if you have more than 30 observations across the categories graphed (although if you have a lot of categories you may need more data). Smaller numbers can be misleading due to random chance.
The classification by categories is only a guide to where to focus attention. It does not necessarily tell you that something individual to that category is the cause of the different frequencies seen. In the example above, you may decide to start improvement efforts in ward 24L – although you would need to consider other things such as readiness of the staff there to be involved.
It is useful to know if underlying processes are stable (only random variation) across the time frame for the data in the Pareto chart, in order to correctly interpret it. Pareto charts don't explain what sort of variation is being observed.