Deciding which Control Chart to Use Statistical Process Control (SPC) Charts

In contrast to the run chart there are numerous ways to construct control charts. The
decision of which to use often hinges on the type of data you have collected. ‘A one page book’
There are two categories of data:
‘Count what is countable, measure what is measurable. What is not measurable, make
measurable'."
Proportion or Attributes Data Galileo Galilei
Continuous or Variables Data This one page book covers some of the basic theory regarding the use of Control Charts
Counts of event that can be aggregated
This is data that can take on differing
into discrete categories - acceptable for quality improvement. It builds on the topics summarised in the MCA ‘Run Charts’ one
values on a continuous scale, over
vs. non acceptable, infected vs. not page book and is best used after those concepts have been practiced and are
time—e.g. length of stay, total number
infected, late vs. on time, Attendance understood. Run charts are convenient and are easy to construct and understand but
of discharges, attendances etc
vs. non attendance they are not as sensitive in detecting special causes as control charts are. The table
summarizes some of the differences to consider between Run Charts and SPC Charts.

Continuous or Variable Data Run Charts vs. SPC Charts

At its most simplistic level the appropriate chart to use for continuous data is the
Run Charts SPC Charts
XmR chart. This chart dose not assume any underlying distribution and control limits are
based on the average moving range of point to point variation. There are other types of Simple & easy to understand and interpret More complex, not just one type to consider
charts for continuous data where each subgroup or unit of time on the x axis has more Easy to create with paper or in Excel Need a special template or special software
than one observation (X bar and S charts and X Bar and R charts).
Discrete or Attributes Data Less sensitive and can miss some special More sensitive and powerful tool—control limits
The p-chart (p—Percent or Proportion) is the most easily understood and most often used cause signals provide additional rigor.
control chart for attributes data. It plots the percentage of events over time. On this chart No predictive capability Control limits show the precision and can
the control limits vary with each data point and are calculated based on the denominator. accurately predict future performance.

Once again there are Best with 12-25 data points Best with 12-25 data points
other types of charts that
can be used for discrete The Anatomy of a Control Chart
data, such a the U Chart
which is useful for data
best measured as a ratio, Upper Control
or a C Chart used for Limit (UCL)
discrete data where the
denominator is fairly
A variable (y)
Mean
constant, and the G
Chart, which is useful for
rare events (time to
Lower Control
event analyses).
Limit (LCL)
Resources: Further information about using Control Charts
 Carey (2003) - Improving Healthcare with Control Charts
 Carey & Lloyd (1995) - Measuring Quality Improvement in
Healthcare Time (x)
1
4 ©2014 Trustees of Dartmouth College, Sheffield Microsystem Coaching Academy Note that the centre line is the Mean, as opposed to the Median on a Run Chart.
Basic SPC Theory Basic Tests for Detecting Special Cause Signals on a Control Chart

The statistical principles behind the development of control charts were first developed by 1. One Point outside of the Control limits (either Lower or Upper)
Walter Shewhart in the 1920’s. Shewhart realized that some variation (Common Cause)
was part of the normal (chance) variation of life, and some variation was due
to “assignable causes” or statistically significant variations (Special Cause).

In the schematic below, a normal “bell shaped” distribution is displayed. The standard
deviation (SD) is a measure of the dispersion of this distribution. SPC charts use an
adaptation of this concept in the calculation of control limits.

SPC chart control limits are not

calculated using standard deviations
because an underlying distribution is
not assumed. SPC control limits are UCL
approximately +/- 2.67 SD or 3 3s
Sigma
3 sigma deviations from the mean. Mean
33s
Sigma
LCL

Statistically there is a
0.3% probability
of having a data
point falling beyond
99.7% will be within 6 SDs these limits. 2. Eight or more successive values on the same side of the centre line
When a data point (Shift)
0.3% will be outside 6 SDs does exceed the
in a normal distribution UCL or LCL this is
an indication of a
special cause. We
can also predict that
if nothing in the
system changes
99.7% of future
points will be within
the two control limits
Mean – 3SD Mean Mean + 3SD
3. Seven or more values in a row continually increasing or decreasing
6 SD (Trend)
There are different formulae for calculating the control limits for the different types of 3
SPC charts. Most SPC software applications such as WinChart, SPC Excel, QI Macros,
Note that there are some other more detailed tests for detecting special cause signals but
Minitab, and others can do control limit calculations automatically.
for most healthcare processes the 3 tests above will usually be adequate.
(See Carey & Lloyd 1995, or other SPC texts/articles for more detail).