Home » Lean Six Sigma Articles » Primitive Charts Hide Improvements

Primitive Charts Hide Improvements

Improve Your Data Analysis Skills Using Control Charts and Histograms

I was reading an old issue of the Joint Commission Journal on Quality and Patient Safety. In an article about door-to-balloon times the authors used a scatter plot showing each time and a target line (90 minutes or 1hr 30 minutes):

Door-to-Balloon Time Scatter Chart QI Macros

But what can we really tell from this chart? Not much.

Scatter Charts can be Unclear and Confusing

What if we used an XmR control chart to show performance over time?

Door-to-Balloon Time XmR Control Chart QI Macros

We can see that the process is stable and predictable with an average time of 74 minutes. There is a downward trend starting at point 49. Was an improvement implemented? Looking at the Range chart we see a run of points below the center line starting with point 56. Did an improvement reduce the variability?

We could have also shown the data as a histogram to dramatize the inability to meet the 90 minute target:

Door-to-Balloon Time Histogram QI Macros

Or we could have used the Capability Suite to show all this and more:

Door-to-Balloon Time Capability Suite QI Macros

I'd like you to consider that control charts and histograms would have told this story more clearly and established a baseline for improvement.

Use Control Charts to Show Before and After Improvement

We could add the improved DB times and demonstrate improvement with the same XmR control chart and histograms:

Door-to-Balloon Time Control Chart of Results QI Macros

This shows we've both reduced D2B time and variation. There's still a statistically significant downward trend in the middle that I would want to analyze to determine if we tried but failed to lock in a significant improvement.