Lung Cancer Screening Improvement Collaborative
Mammography Positioning Improvement Collaborative
Prostate MR Image Quality Improvement Collaborative
Recommendations Follow-Up Improvement Collaborative

How to Select and Use Run and Control Charts

Purpose

Run chart and control charts are visual representations that help us to define normal variation within a process, identify changes in performance, and develop a compelling improvement narrative. This article will help you determine the appropriate chart for your data set, learn how to analyze your chart, and provide tips to improve your visualization.


Chart Elements

For a defined group of consecutive data points or mean grouping, the following chart elements may be visible depending on chart type.


P-Chart with legend


Mean or Average

The mean or average of your process data is represented by a solid horizontal blue line. The mean line will have data points that exist randomly above and below the center mean line.   The first mean grouping average is referred to as the baseline.  


Target

The target line is represented by a solid horizontal green line.  This line visualizes the target value you are trying to obtain by the end of your project.


Upper and Lower Control Limits

Control limits represent the normal variation expected in your process.  They are represented by grey dashed lines above and below the mean.   Based on control chart type the control limits can be static horizontal or dynamic.


Types of Charts

Run Chart

A run chart is a collection of data points plotted over time.  Run charts do not contain upper or lower control limits for defined mean groupings. 

Run Chart Image

P-Chart

A P-Chart is a collection of proportional data plotted over time.  The percentage is calculated by the number of times something happened divided by the number of times something could have happened.

The number of times something happened divided by the number of times something could have happened equals the percentage of time something happened

For each mean grouping, an upper and lower control limit is established – these limits display the amount of natural variation expected in the process. If the weekly sample size changes, weekly control limits within a mean grouping will fluctuate based on those different weekly sample sizes. The larger the sample size, the less the variability and the narrower the control limits[.


I-Chart

An I-Chart is a collection of distinct events plotted over time.  For each mean grouping, an upper and lower control limit is established – these limits display the amount of natural variation expected in the process. The control limits do not fluctuate within a mean grouping range.


Analysis

When analyzing your chart, use the three rules below to identify indicators of non-random variation. If none of these rules are applicable for your data set, then you are seeing random variation within the process.


  • Shift: Seven or more consecutive data points that lie either above or below the mean. Data points that fall on the mean do not contribute to or break a shift.  When a shift is observed, it is due to a corresponding change in your process. A new mean grouping can be identified to determine the natural variation for this new process.

 

  • Trend: Five or more consecutive data points that are all increasing or decreasing.  When a trend is observed, it is not appropriate to define a new mean grouping since the process is still changing. Once a shift has occurred and the process has stabilized, then a new mean grouping can be defined.


  • Astronomical Point (Special Cause): A point that occurs outside of the control limit range.  If a data point lies on the control limit, it is still within normal variation.  When an astronomical point is observed, it means that something out of the ordinary occurred, and you should investigate the cause of this data point.


Annotation

It is important to annotate your charts, as this provides context for changes in the data. Annotations will enable you to identify which interventions impacted your performance and will help support your improvement story.


References

Larson, David (2022). Statistic Process Control. ImPower Video.

Provost, L.P. & Murray, S.K. (2011). The health care data guide: Learning from data for improvement. Jossey-Bass.

Did you find it helpful? Yes No

Send feedback
Sorry we couldn't be helpful. Help us improve this article with your feedback.