An Introduction to Statistical Process Control Charts (SPC)
Steve Harrison Topics
• Variation – A Quick Recap • An introduction to SPC Charts • Interpretation • Quiz • Application in Improvement work
Variation Common Cause Variation
• Typically due to a large number of small sources of variation • Example: Variation in work commute due to traffic lights, pedestrian traffic, parking issues • Usually requires a deep understanding of the process to minimise the variation • Multiple factors
5 Special Cause Variation
• Are not part of the normal process. Arises from special circumstances • Example: Variation in work commute impacted by flat tyre, road closure, ice and snow. • Usually best uncovered when monitoring data in real time (or close to that) • Assignable cause
6 Special Cause - My trip to work
120
100 Upper process limit
80 Mean Lower process limit Min. 60
40
20
0 Consecutive trips Two Types of Variation
Common Cause: Special Cause: • chance cause • assignable cause • noise • signal
Statistically significant (not good or bad)
8 SPC Charts
9 SPC, Statistical Process Control or The Control Chart Elements 1. Chart/graph showing data, running record, time order sequence 2. A line showing the mean 3. 2 lines showing the upper and lower process ‘control’ limits Its best if you have 25 data points to set up a control chart, but 50 are better if available. Be careful of too many points… The Anatomy of an SPC or Control Chart
80 Upper 70 process control limit 60 50 Mean 40 Lower 30 process 20 control limit
10 0 F M A M J J A S O N D J F M A M J J A S O N D Measures of Central Tendency • Mean = Average – SPC Chart • Median = Central or Middle Value – Run Chart • Mode = Most frequently occurring value
12 Standard Deviation or σ
In statistics, standard deviation shows how much variation exists from the mean. A low standard deviation indicates that the data points tend to be very close to the mean; high standard deviation indicates that the data points are spread out over a large range of values. Standard Deviation and a normal distribution PRACTICAL INTERPRETATION OF THE STANDARD DEVIATION
99.7% will be within 3 s
Mean - 3s Mean Mean + 3s 3s AND THE CONTROL CHART UCL 3s Mean 3s LCL
6s Run Charts vs. SPC Charts Run Chart SPC • Simple • More Powerful • Easy to create in Excel or • Control lines show the on paper degree of variation • Less Sensitive • Need software • Only need 12-15 data • Better with 25+ data
pointsWard x – % of total TTOs completed by 12 noon points April 4 - May 15, 2012 80 70 60
50
40
30
20 Daily TTOs Completed by NoonbyCompleted TTOs Daily
10 % %
0
Apr Apr Apr Apr Apr
Apr Apr Apr
May May May
May May May
- - - - -
- - -
- - -
- - -
8 4 6
3 9 5
14 20 12 18 22
13 11 15
17 Special cause variation
90 80 70 60 50 40 30 20 10 0 F M A M J J A S O N D J F M A M J J A S O N D SPECIAL CAUSES - RULE 1
UCL Point above Upper Control Limit (UCL)
MEAN
LCL SPECIAL CAUSES - RULE 1
UCL
MEAN
Or point below Lower Control LCL Limit (LCL) SPECIAL CAUSES - RULE 2
UCL
MEAN
LCL Eight points above centre line SPECIAL CAUSES - RULE 2
UCL Or eight points below centre line
MEAN
LCL SPECIAL CAUSES - RULE 3
UCL Six points in a downward direction
MEAN
LCL SPECIAL CAUSES - RULE 3
UCL Or six points in an upward direction
MEAN
LCL Quiz – Does the chart show
A. Special Cause Variation? 67% B. Common Cause Variation? 33% C. Both of the above D. No Variation 0% 0%
No Variation
Both of the above
Special Cause Variation? Common Cause Variation? How many special cause signals are present on this chart?
A. 0 67% B. 1 C. 2 D. 3 33% E. 16
0% 0% 0%
0 1 2 3 16 How many special cause signals are present on this chart?
A. 0 100% B. 1 C. 2 D. 3 E. 16
0% 0% 0% 0%
0 1 2 3 16 How many special cause signals are present on this chart?
A. 0 100% B. 1 C. 2 D. 3 E. 16
0% 0% 0% 0%
0 1 2 3 16 What use is this?
• Evaluate and improve underlying process • Is the process stable? • Use data to make predictions and help planning • Recognise variation • Prove/disprove assumptions and (mis)conceptions • Help drive improvement – identify statistically significant change Example
Annotated SPC Charts
• One of the most powerful tools for improvement • Describe a process captured over time (as opposed to being a single sample) • Reveal any trends a process might be experiencing • When combined with careful annotation they track the impact of change
Annotated SPC Charts Application – Responding to Variation
33 Identify the cause: Process with if positive then can it be replicated or special cause standardized. If negative then cause variation needs to be eliminated
Reduce variation: Process with make the process even more reliable common cause Not satisfied with result: variation redesign process to get a better result
34 PRACTICAL Length Of Stay for Bowel Surgery Patients Quality Composite Ratio Outpatient attendances (Part 1) Outpatient attendances (Part 2) % discharged by noon THANKS!