The second question of the Model for Improvement asks 'How will we know that change is an improvement?' Using data to understand your problem and identify whether your improvement efforts have made a difference is fundamental to the QI approach.

 

Types of measures

There are four types of measures we can use to understand whether the changes we introduce are helping us achieve our improvement aim:

• Outcome measures: These tell us whether we're achieving our overall aim
• Process measures: These tell us whether the specific process changes we have implemented are working as planned 
• Balancing measures: These monitor whether the changes we're introducing may be having a negative impact elsewhere in the system. For example, if we're reducing length of stay rates, are readmission rates increasing?
• Financial measures: These are used to monitor the possible financial impact of our changes

 

Presenting data over time

In QI, data for improvement is usually presented over time. This is most commonly using a run or statistical process control (SPC) chart.

Viewing data in this way helps us tell a story - it allows us to see whether a system has changed over time following the implementation of our change ideas. It also enables us to collect information about an issue by understanding trends or cycles in our baseline data. 

 

Understanding variation

In order to build stable systems, we must be able to understand expected variation within a system. The nature of healthcare means there will always be some variation, but our aim is to reduce unwarranted variation. By understanding variation, we can focus our attention in the right areas and distinguish 'signal' from 'noise'.

• 'Common cause' variation naturally occurs within a process, due to common factors that do not change over time. These will appear as data points within the control limits plotted.
• 'Special cause' variation occurs outside of the expected natural variation of a process, and warrants further investigation.

 

Statistical process control

Run charts and control charts use statistical analysis to calculate whether variation within a process is within expected limits. Data are plotted on a run chart, with a central line for the average, an upper line for the upper control limit, and a lower line for the lower control limit. This helps identify whether our data is behaving in line with common cause variation, or whether any changes are statistically significant and indicate special cause variation. 

 

 

 

 

 

 

Some key types of special cause variation to look out for when using SPC charts:

 

 

An outlier (fig. a) These are data points outside of the control limits.

 

 

 

A shift (fig. b) A number of consecutive data points above or below the average, usually 7 or more consecutive points. This indicates a statistically significant sustained shift in performance. A trend A number of consecutive data points in an upwards or downwards trend, usually 7 or more.

 

 

If you want to learn more about data for improvement

• Join our module 3: Measurement for improvement or SPC Masterclass training (link to training page)
• NHS data guide link