Once the problem scoping has been undertaken, existing data can be collected and analysed to further understand and provide evidence of the nature, extent, and severity of the problem. This page explores different approaches to understanding data and how they can contribute to improvement efforts.
A variety of approaches are useful in selecting what to work on, and with whom. As well as using data, it is important to gather qualitative information from patients and staff who use services, perhaps using surveys, interviews, or focus groups. Where do patients and staff see problems with and best practice in the care they receive and provide?
Many of these approaches benefit from having baseline data from the service you are considering improving. Finding appropriate methods to display the data to allow those working on the improvements to interact with it is key. However beautifully analysed data is, it will not support improvement if it is not reviewed and understood.
Run charts (or time series charts)
Run charts are simple yet useful tools in quality improvement. They are line graphs that plot measures over time, often with a median. Run charts help determine if improvement has occurred and if it has been sustained. They provide valuable information on the performance and value of changes made to a process. While run charts are not as sensitive at detecting special cause variation as SPC charts, they are effective in assessing the effectiveness of change over time.
Statistical process control (SPC) charts
SPC is an accessible statistical approach to resolving problems and finding solutions. It can be used throughout the lifecycle of a project to assess performance and progress. SPC charts are a type of run chart that show performance over time with upper and lower limits to indicate levels of variation. By analysing SPC charts, you can identify special cause variation, which indicates the need for improvement in the system.
Pareto charts
A Pareto chart is a bar chart that shows how often different categories of events/incidents take place. By conducting Pareto analysis, major problems can be identified and tackled first, eliminating the biggest cause of the problem. The chart follows the 80/20 principle, where 80% of the effects come from 20% of the causes. This allows for a focus on improvement projects where they are needed most and will have the biggest impact.
Frequency charts
Frequency plots, such as histograms, bar charts, and dot plots, are useful in investigating the spread of continuous data and identifying patterns. Histograms display groupings of a continuous measure, while bar charts show categories and their corresponding frequencies. Constructing frequency plots with a sufficient number of data points can provide valuable insights into the data.
Scatter plots
Scatter plots depict the relationship between two variables in pairs of observations. They show how changes in one variable may affect another. Positive correlation indicates that the variables increase together, while negative correlation suggests that they decrease together. Scatter plots can help identify cause-and-effect relationships for further investigation. It is important to analyse separate scatter plots for different time periods with random and non-random variations to understand the differences.
