1e control charts - ihi home pageapp.ihi.org/.../presentation_1e_-_control_charts.pdf · ·...

1E – Control Charts

Advanced Measurement for

Improvement Seminar

March 20-21, 2017

Two Types of Variation

Common Cause

Is inherent in the design of the

process

Reflects the “business as

usual” state of the process

Is due to regular, natural or

ordinary causes

Affects all the outcomes of a

process

Results in a “stable”

distribution that is predictable

Also known as random or

unassignable variation

Special Cause

Due to irregular or unnatural

causes that are not inherent in

the design of the process

Reflects a “different mode” of

the process

Affects some, but not

necessarily all aspects of the

process

Results in an “unstable”

process that is not predictable

Also known as non-random or

assignable variation

A Stable Process

A predictable (stable) process has only common causes in

play.

!

A Stable Process

© R. Scoville & IHI • 5

Successive samples from a stable process differ only by chance

0

2

4

6

8

10

12

07.

5 1522

.5 3037

.5 4552

.5 6067

.5 7582

.5 9097

.5

0

2

4

6

8

10

12

14

16

07.

5 1522

.5 3037

.5 4552

.5 6067

.5 7582

.5 9097

.5

0

2

4

6

8

10

12

14

07.

5 1522

.5 3037

.5 4552

.5 6067

.5 7582

.5 9097

.5

0

2

4

6

8

10

12

14

16

18

07.

5 1522

.5 3037

.5 4552

.5 6067

.5 7582

.5 9097

.5

0

2

4

6

8

10

12

14

07.

5 1522

.5 3037

.5 4552

.5 6067

.5 7582

.5 9097

.5

0

2

4

6

8

10

12

14

16

07.

5 1522

.5 3037

.5 4552

.5 6067

.5 7582

.5 9097

.5

Time

What Common Cause Variation Looks Like

Points equally likely above or below center line

No trends or shifts or other patterns

0

10

20

30

40

50

60

70

80

90

100

1/1/2

008

1/3/2

008

1/5/2

008

1/7/2

008

1/9/2

008

1/11/2

008

1/13/2

008

1/15/2

008

1/17/2

008

1/19/2

008

1/21/2

008

1/23/2

008

1/25/2

008

1/27/2

008

1/29/2

008

1/31/2

008

2/2/2

008

2/4/2

008

2/6/2

008

2/8/2

008

A Stable Process is Predictable

Thus you can confidently

Counsel patients about what to

expect

Plan for the future

Inform management

Use PDSA testing to improve

it!

!

Stable “OK”

A process may be operating

in a stable, predictable

fashion but still produce

unacceptable results!

Special Causes

Unintentional

When the system

is out of control

F e b r u a r yA p r il

1

2

3

4

5

6

7

8

9

1 0

1 1

1 2

1 3

1 4

1 5

1 6

1 7

1 8

1 9

2 0

2 1

2 2

2 3

2 4

2 5

2 6

2 7

2 8

2 9

3 0

3 1

3 2

1 6 P a t ie n t s in F e b r u a r y a n d 1 6 P a t ie n t s in A p r il

Min

ut

es

2 . 5

5 . 0

7 . 5

1 0 . 0

1 2 . 5

1 5 . 0

1 7 . 5

2 0 . 0

2 2 . 5

2 5 . 0

2 7 . 5

3 0 . 0

A

B

C

C

B

A

U C L = 1 5 . 3

C L = 1 0 . 7

L C L = 6 . 1

X m R C h a r tWaiting times

Intentional

When we’re trying

to change the

system

An unstable system!

Where Do Special Causes Come From?

Inherent instability in the process

Lack of standardization – a chaotic process

Changes in personnel, equipment, management, etc.

Unusual extrinsic events

Catastrophes, breakdowns, accidents, personnel issues

Entropy

Equipment wear, desensitization, habit, emerging culture

Intentional changes – part of an improvement initiative

Unintended Special Causes

An unstable process is subject to special causes. These

represent fluctuations in underlying processes.

Process A(the one we think we’re measuring)

Process B

Process C

Time

Removing Special Causes

Standardize the process by

imposing a design & selectively

eliminating special causes.

Now your process changes can

have testable, repeatable impact.

Special causes present:CHAOS!

Special causes removed

!

Stabilize, Then Improve

Once the process is stable,

your changes can have a

predictable, repeatable impact.

HERDING

CATS MOVIE

HERE

If you can’t predict the

future behavior of the

process, you’re

improvements won’t stick!

Managing with Data

Are special

causes

present?

Are outcomes

acceptable?

Do nothing

Use PDSA to

improve the

process

Standardize:

• Investigate &

eliminate special

causes,

• Or redesign the

process

No

Yes

Yes No

Changing the Process

Intervention

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

A successful process change is an intentional special cause!

Improvement

Once special causes are eliminated and the process is

stable, we can test changes to improve process capability

Reduce variationMove the mean

Tools for Assessing Process

Line Charts

No decision rules

‘Ocular’ tests only

Run Charts

Decision rules based on ‘50/50’ principle

Cannot reveal special cause

Minimum of 6 points to detect an improvement shift

Control Charts

Most sensitive tools for detecting special causes including

successful changes

Shewhart Control Chart

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Week

Percent of Patients with Pressure Ulcers

3-sigma control

limitsMean

Subgroup

Normal Distribution (Xbar-S Charts)

95.46%

99.73%

68.26%

-3σ +3σ

-2σ +2σ

-1σ +1σ

Mean

For normally distributed data, +/- 3 sigma limits

include 99.7% of the common-cause values.

What’s a Subgroup?

A point on a Shewhart control chart

A set of observations taken from the process at a point in

time (area of opportunity)

Process should be stable inside the subgroup (maximize

variation between subgroups; minimize variation within

subgroups)

A subgroup is a ‘snapshot’ of the process at a particular

place and time

Subgroups

X X X

X XX

XX

Subgroup

Value

Subgroup

Value

Subgroup

Value

Observation

Area of

Opportunity X

Subgroup values combines 1

or more individual

observations:

• Count

• Percent

• Rate

• Average

• Single continuous value

Subgroup Example

Measure: time to process a new hospital admission

We suspect that admission staff on different shifts use

different procedures for processing patients.

We should choose a subgroup to minimize variation

within the subgroups

Subgroup = average time for 5 admissions randomly

selected within a shift on Unit X

What are possible subgroups for admission

process time?

Shewhart Control Chart

Upper control limit

Lower control limit

Center: “best guess” value

±3σ limits

include 99.7%

of common

cause values!

A Shewhart chart is a ‘special cause detector’

– a statistical display that helps you infer the

presence of special causes in a process,

beyond a reasonable doubt.

Special Cause

Upper control limit

Lower control limit

Center: “best guess” value

±3σ limits

include 99.7%

of common

cause values!

A single point outside the control limits is likely

NOT generated by a stable process, but by

some others set of causes.

A single point outside the control limits

Six consecutive points increasing (trend up) ordecreasing (trend down)

Two our of three consecutive points near a controllimit (outer one-third)

Eight or more consecutive points above or belowthe centerline

Fifteen consecutive points close to the centerline(inner one-third)

Rules for

Detecting

Special Cause

Tests for Special Cause

Outside of limits: A data point that falls outside the limits on the chart, either above the upper limit or below the lower limit.

Shift: Eight or more consecutive POINTS either all above or all below the mean. Skip values on the mean and continue counting points. Values on the mean DO NOT make or break a shift.

Trend: Six points all going up or all going down. If the value of two or more successive points is the same, ignore one of the points when counting; like values Do Not make or break a trend.

Two Out of Three: Two out of three consecutive points in the outer third of the chart. The two out of three consecutive points can be on the same side, or on either side of the center line.

15 points Hugging the Centerline: 15 consecutive points close to (within inner third of limits) centerline.

Mammography Screening

A primary care plan sends postcards each month to

remind women age 50 and older to get mammograms.

Measure: Percent of women in a sample of 50 who

obtain documented mammograms within 3 months of

receiving postcard reminder.

The team used a P-Chart to plot their data.

What Should the Process Manager Do?

Percent With Mammograms

0%

5%

10%

15%

20%

25%

30%

35%

40%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Month

What Should the Process Manager Do?

Percent With Mammograms

0%

5%

10%

15%

20%

25%

30%

35%

40%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 20 21 22 23 24

Month

Improvement Strategy: Special Cause Variation

When process exhibits unintended special cause, something

not typically part of the process design is affecting the

process:

Identify when and where the special cause occurred.

Learn from the special cause.

Take action based on the special cause – standardize.

Irrelevant special cause: remove from consideration

Undesirable special cause: remove it and make it difficult for it to

occur again.

Desirable special cause: make it a permanent part of the health

care process.

Adapted from Provost, L. P. and S. K. Murray (2011). The Health Care Data Guide: Learning from Data for Improvement. San Francisco, Josey-Bass.

Improvement Strategy: Common Cause Variation

When a measure stable (only common cause variation) we

are seeing only variation inherent in the current process

design.

Since the process is performing as well as it can, to

make it perform better will require process redesign.

Identify aspects of the process to change.

Use process analysis tools to understand the process;

use change concepts or staff suggestions to identify

changes

Test changes using Plan-Do-Study-Act (PDSA) cycles.

Implement successful changes using the PDSA cycle.

HCDG Page 108

32

Adapted from Provost, L. P. and S. K. Murray (2011). The Health Care Data Guide: Learning from Data for Improvement. San Francisco, Josey-Bass.

Proceed with Caution

There are many types of control charts, which are

appropriate for different types of data.

Calculation methods are specific to the type of chart, but

interpretation is the same for most chart types.

You cannot create a valid control chart using a simple

standard deviation calculation.

Geek Alert!ATTENTION!

The following material is intended

for geek audiences primarily.

To avoid disorientation,

statophobes and other normal

individuals should limit their

consumption of these details…

The X bar S Chart

For continuous data

Each dot on an 𝑋 chart is the average of multiple

measurement within the subgroup

A pair of charts: 𝑋 and S

𝑋 plots subgroup averages

S plots standard deviation within each subgroup

Subgroup size can be equal or unequal

Common cause variation is modeled by the normal

distribution

Xbar and S chart P36

Subgroup value is

average of

observations within

the subgroup

Limits depend on

variation within

subgroups

(that’s why they are

wavy)

S chart plots

subgroup standard

deviations


Date

Setup

Case 1

Setup

Case 2

Setup

Case 3

Setup

Case 4

Setup

Case 5

Setup

Case 6 Count Average Std. Dev.

1/1/2016 20 30 36 21 33 5 28.0 7.2

1/2/2016 26 70 51 3 49.0 22.1

1/3/2016 25 11 22 3 19.3 7.4

1/4/2016 49 36 36 21 4 35.5 11.4

1/5/2016 53 23 25 3 33.7 16.8

1/6/2016 16 21 12 3 16.3 4.5

1/7/2016 36 60 44 27 51 17 6 39.2 15.8

1/8/2016 45 7 14 3 22.0 20.2

1/9/2016 8 32 12 3 17.3 12.9

1/10/2016 26 17 12 3 18.3 7.1

1/11/2016 18 14 56 22 26 29 6 27.5 15.0

1/12/2016 11 16 28 3 18.3 8.7

1/13/2016 10 27 33 19 4 22.3 10.0

1/14/2016 21 18 49 3 29.3 17.1

Average 26.9 12.6

• Xbar chart plots

subgroup

averages

• Center line is

average of

subgroup

averages


Date

Setup

Case 1

Setup

Case 2

Setup

Case 3

Setup

Case 4

Setup

Case 5

Setup


1/1/2016 20 30 36 21 33 5 28.0 7.2

1/2/2016 26 70 51 3 49.0 22.1

1/3/2016 25 11 22 3 19.3 7.4

1/4/2016 49 36 36 21 4 35.5 11.4

1/5/2016 53 23 25 3 33.7 16.8

1/6/2016 16 21 12 3 16.3 4.5

1/7/2016 36 60 44 27 51 17 6 39.2 15.8

1/8/2016 45 7 14 3 22.0 20.2

1/9/2016 8 32 12 3 17.3 12.9

1/10/2016 26 17 12 3 18.3 7.1

1/11/2016 18 14 56 22 26 29 6 27.5 15.0

1/12/2016 11 16 28 3 18.3 8.7

1/13/2016 10 27 33 19 4 22.3 10.0

1/14/2016 21 18 49 3 29.3 17.1

Average 26.9 12.6

• S chart plots

subgroup

standard

deviations

• Center line is

average of

subgroup

standard

deviations

X bar and S Formulas

Pages 160 and appendix page 196

39

𝑈𝐶𝐿𝑥 = 𝑋 + (𝐴3 × 𝑆𝑏𝑎𝑟) 𝐿𝐶𝐿𝑥 = 𝑋 − (𝐴3 × 𝑆𝑏𝑎𝑟)

𝑈𝐶𝐿𝑠 = 𝑆 + (𝐵4 × 𝑆𝑏𝑎𝑟) 𝐿𝐶𝐿𝑠 = 𝑆 + (𝐵3 × 𝑆𝑏𝑎𝑟)

Xbar chart:

S chart:

𝑋is average

of subgroup

averages

Sbaris average

of subgroup

averages

Factors A3, B3, B4

from a

statistical

table

X bar and S Formulas

Pages 160 and appendix page 196

40

Factors A3, B3 and B4 come from a statistical table:

Note that A3 becomes smaller as sample size increases

𝑈𝐶𝐿𝑥 = 𝑋 + (𝐴3 × 𝑆𝑏𝑎𝑟) 𝐿𝐶𝐿𝑥 = 𝑋 − (𝐴3 × 𝑆𝑏𝑎𝑟)

𝑈𝐶𝐿𝑠 = 𝑆 + (𝐵4 × 𝑆𝑏𝑎𝑟) 𝐿𝐶𝐿𝑠 = 𝑆 + (𝐵3 × 𝑆𝑏𝑎𝑟)

Xbar chart:

S chart:

Xbar and S chart

P41

Date

Setup

Case 1

Setup

Case 2

Setup

Case 3

Setup

Case 4

Setup

Case 5

Setup


1/1/2016 20 30 36 21 33 5 28.0 7.2

1/2/2016 26 70 51 3 49.0 22.1

1/3/2016 25 11 22 3 19.3 7.4

1/4/2016 49 36 36 21 4 35.5 11.4

1/5/2016 53 23 25 3 33.7 16.8

1/6/2016 16 21 12 3 16.3 4.5

1/7/2016 36 60 44 27 51 17 6 39.2 15.8

1/8/2016 45 7 14 3 22.0 20.2

1/9/2016 8 32 12 3 17.3 12.9

1/10/2016 26 17 12 3 18.3 7.1

1/11/2016 18 14 56 22 26 29 6 27.5 15.0

1/12/2016 11 16 28 3 18.3 8.7

1/13/2016 10 27 33 19 4 22.3 10.0

1/14/2016 21 18 49 3 29.3 17.1

Average 26.9 12.6

UCLx LCLx UCLS LCLS

44.72 9.00 26.10 0.00

51.22 2.51 32.10 0.00

51.22 2.51 32.10 0.00

47.22 6.50 28.35 0.00

51.22 2.51 32.10 0.00

51.22 2.51 32.10 0.00

42.98 10.75 24.61 0.37

51.22 2.51 32.10 0.00

51.22 2.51 32.10 0.00

51.22 2.51 32.10 0.00

42.98 10.75 24.61 0.37

51.22 2.51 32.10 0.00

47.22 6.50 28.35 0.00

51.22 2.51 32.10 0.00

Width of limits

determined by average

subgroup variances and

subgroup size

Between-

subgroup

variation reveals

stability (or not)

in the process

Narrower limits are more sensitive to special

causes among the subgroups. How?

• Bigger subgroups

• Less within-subgroup variation

X bar and S Charts

Testing a Change43

Date

Setup

Case 1

Setup

Case 2

Setup

Case 3

Setup

Case 4

Setup

Case 5

Setup

Case 6

1/1/2016 20 30 36 21 33

1/2/2016 26 70 51

1/3/2016 25 11 22

1/4/2016 49 36 36 21

1/5/2016 53 23 25

1/6/2016 16 21 12

1/7/2016 36 60 44 27 51 17

1/8/2016 45 7 14

1/9/2016 8 32 12

1/10/2016 26 17 12

1/11/2016 18 14 56 22 26 29

1/12/2016 11 16 28

1/13/2016 10 27 33 19

1/14/2016 21 18 49

1/15/2016 15 12 7 18 6

1/16/2016 11 9 14 8 12

1/17/2016 12 13 16 11 14 9

New subgroups plotted

using baseline average

(26.9)

Change introduced here

Was the change effective?

When the Process Has Changed…P45

The chart reflects the new process (with

temporary limits). It will take additional data to

determine if the post-change process is stable.

Two Mistakes to Avoid

Tampering (Type I Error or False Positive)

Responding to a data point as if it were a special cause when, in

fact, the system is stable

Failure to Detect (Type II Error or False Negative)

Ignoring a data point that indicates a special cause when, in fact,

the system of causes has changed

What Do Type I and II Errors Mean for Setup Time?

Change introduced here

Consequences P49

ACTUAL SITUATION

ACTIONWhen NO special cause is occurring in the system

When a special cause is occurring in the system

Respond to individual

events (special cause

approach)

TYPE 1 ERROR: Tampering

Cost: • Tampering makes

variation worse• Fail to address

underlying process structure

Treat events as part of

system; work on

changing the system

(common cause

approach)

TYPE 2 ERROR: “Failure to Detect”

Cost:• Fail to remedy special

cause• Wasted effort to

redesign process when not needed

• Adds variation to process.

Why 3 Sigma Limits?

The limits have a basis in statistical theory

They have proven in practice to distinguish common and

special causes

3-sigma limits approximately minimize the risk of Type 1

(‘tampering’) and Type 2 (‘failure to detect/)

3-sigma limits protect the morale of workers by defining

the magnitude of variation built into the process: IT’S

THE SYSTEM: NO ONE’S FAULT

Provost, L. P. and S. K. Murray (2010). The Data Guide - Learning from data to improve health care. Austin TX, Associates in Process Improvement - www.pipproducts.com.

Why 3 Sigma Limits: Balancing the Risk

High

Low

Ris

k

+/-

3SL

+/-

1SL

+/-

2SL

+/-

4SL

+/-

5SL

+/-

6SL

Tampering(Type I Error)

Failure to Detect(Type II Error)

Source: R. Lloyd

3-sigma limits

Choosing and Using Control Charts

Which chart to use?

Data types; subgroups; rates

Practice with chart interpretation

Type of Data

Discrete

Count Classification

Equal area of opportunity

Unequal area of opportunity

Percent or proportion

Continuous

1 observation

per subgroup

C-Chart U-Chart P-Chart Individuals X-bar & S

>1 observation

per subgroup

Discrete and Continuous Data

Discrete – count or classify things in categories

Continuous - take on any fractional value on a

continuous scale


Continuous

Time ThroughputMoney

Discrete

Classification Count

Scale

More Discrete and Continuous Data

In making this determination, consider the nature of the

observation; i.e. disregard aggregations such as counts,

percentages, averages, etc.

Temperature Continuous

Height or weight Continuous

LDL Continuous

Cycle time for patient visits Continuous

Complete in less than 1 hr Discrete

# Patient falls in Unit 12 last week Discrete

Average A1C for practice’s DM patients Continuous

% of women receiving m’grams in 90 days Discrete


In making this determination, consider the nature of the

observation; i.e. disregard aggregations such as counts,

percentages, averages, etc.

Minutes waiting time Continuous

Number of needle sticks/1000 employee days Discrete

LDL Continuous

Length of Stay Continuous

Number errors per 5 insurance claims Discrete

Number of deaths per month Discrete

Percent mortality/month Discrete

Number of errors per 100 med orders Discrete

From Continuous to Discrete

You can always convert a

continuous measure into a

discrete measure by ‘slicing’ it

into categories.

You cannot go the other way:

Information is lost!

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

Severely underweight

Underweight

Normal

Overweight

Obese Class I

Obese Class II

Obese Class III

BMI

RULE OF THUMB: Store “unprocessed” data values in your database –you can always assign to categories later:• BMI• Blood pressure• Birth date

Classification (Percentages, Proportions)

35 apples20 red15 green

100 drug orders6 have ≥1 errors94 have no errors

7 babies2 boys5 girls

50 stat labs48 results w/in 30 min2 >15 min

57% red 71% girls

6% defective

96% OK

Counts P61

Stains on

carpet

Needle

sticks last

month

Falls on 5A

in 2015

Errors on a

drug order

Patients with

MRSA infection

Spots on a dog

Fixing Unequal Areas of Opportunity (Rates) P62

Stains per

square

meter

Errors per 100

drug orders last

month

Catheter infections

per 1000 catheter

daysFalls per 1000

patient days

Type of Data

Discrete

Count Classification

Equal area of opportunity

Unequal area of opportunity

Percent or proportion

Continuous

1 observation

per subgroup

C-Chart U-Chart P-Chart Individuals X-bar & S

>1 observation

per subgroup

Chart Calculations

Each chart type has its own construction formulas &

procedures.

See Mohammed & Worthington (2008)* for overview of

calculation of basic chart types. See The Data Guide for

details.

Once constructed, all charts are interpreted in the same

way.

*Mohammed, M. A., P. Worthington, et al. (2008). "Plotting basic control charts: tutorial notes for

healthcare practitioners." Qual Saf Health Care 17(2): 137-145.

P Charts (for Proportions)

Underlying observations are binary classifications. For example, C-section or no Completed within 1 hour

Infected or not Risk assessed at current visit

Observations are randomly sampled from the process in each subgroup

The subgroups are independent

Each plotted data point is the percent (between 0 and 100%) of the classified observations in the subgroup

Common cause variation is modeled by the binomial distribution

When Do We Recalculate Limits?

You’re still gathering data to find a stable baseline: you

have “trial” limits with less than 20 subgroups

You have identified special causes and want to assess

stability with those subgroups removed

When improvements have been made and the

improvements result in special causes on the chart

When you have reason to believe that the process is

now operating in a new mode, and you want to assess

its stability

Operating Margins P67

You are a member of the board of trustees at two hospitals (Hospital A and

B). You have information that leads you to suspect that Hospital A is well-

managed and in solid financial condition, whereas Hospital B has shown

indications of being poorly managed and in unstable financial condition. In

September 2014, you received the first-quarter financial report for 2014

from hospital A. You are puzzled because the net operating margins

(NOMs) for 2013 and 2014 identical to the NOMs from hospital B, which

you suspect is financially unstable.

In an effort to resolve your puzzlement, you ask both hospital A and B to

provide you with a control chart displaying their NOM by month beginning

with January 2013. Hospital A sends you the actual monthly figures, as

well as the following control chart.

What kind of data are financial ratios?

What type of control chart should they select?

Individuals Charts

Continuous data. Each data point is a single observation

Cost for each hip replacement

Satisfaction score for each client

Total dollars income each month

Test score for each student

Total patients seen each week

Exceptions to the rules

Huge aggregated variables data

average length of stay for our 12,000 patients admitted each month

Averaged data for which we cannot get the numerator and

denominator

I chart is fallback for this….

Skewed data can create problems with limits

Individuals Charts

Week

Average

Time to

Answer

Moving

Range

Oct-14 5.01

Nov-14 5.52 0.51

Dec-14 4.6 -0.92

Jan-15 3.42 -1.18

Feb-15 3.53 0.11

Mar-15 3.84 0.31

Apr-15 4.51 0.67

May-15 6.89 2.38

Jun-15 6.31 -0.58

Jul-15 10.57 4.26

Aug-15 6.75 -3.82

Sep-15 4.04 -2.71

Oct-15 2.39 -1.65

Nov-15 2.53 0.14

Dec-15 3.26 0.73

Jan-16 4.09 0.83

Feb-16 4.81 0.72

Mar-16 5.37 0.56

Apr-16 4.82 -0.55

May-16 7.19 2.37

Jun-16 2.35 -4.84

Jul-16 1.94 -0.41

Aug-16 2.07 0.13

Sep-16 1.84 -0.23

Oct-16 1.22 -0.62

Variance for limits estimated from

differences between successive

values (Moving Range)

Subgroup value is a

single continuous obs.:

subgroup size = 1

(limits are straight lines - they are based on average MR)

Moving range chart is often omitted.

Operating Margins: Hospital A

Is this a stable system?

What would you predict will be the ratio for month 22?

What action would you recommend?

P70

Operating Margins: Hospital B

Is this a stable system?

What would you predict will be the ratio for month 22?

What action would you recommend?

P71

Flash Sterilization P72

“Flash sterilization” refers to steam sterilization cycles where unwrapped medical

instruments are subjected to an abbreviated steam exposure time and then used

promptly after cycle completion without being stored.

The Centers for Disease Control and Prevention (CDC), the Joint Commission, and

AORN all state that flash sterilization should be kept to a minimum and should not be

used as an alternative to purchasing additional instruments, to save time, or for

convenience. AORN documents also state specifically that “flash sterilization may be

associated with increased risk of infection to patients because of pressure on

personnel to eliminate one or more steps in the cleaning and sterilization process.”

Inventory is a widely blamed culprit for the use of flash because it is said that flash

sterilization may be overused to “compensate for insufficient inventory of instruments.”

The Sunny Days Surgical Center has reviewed its logs for data on

its use of flash sterilization.

U Charts

Subgroup is a rate, based on a count and a defined area

of opportunity

Assumes subgroups counts are independent

The area of opportunity is well defined. It may vary in

size

The numerator is frequently larger than the denominator

The data are modeled by the Poisson distribution

73

C Charts

Subgroup is the number of observations in the area of

opportunity

Assumes the subgroup counts are independent

The area of opportunity is well defined and is nearly

constant for each subgroup.

The data are modeled by the Poisson distribution

74

Flash Sterilizations

What kind of data are these?

Is the area of opportunity the same

each week?

What is the appropriate type of control

chart?

P75

Sunny Days recognizes that flash sterilizations

are putting patients at risk.

They acquire additional instruments, and

deploy them beginning in week 20

Week Flashes Total Surgeries

1 35 84

2 43 80

3 38 79

4 39 81

5 42 83

6 47 85

7 46 79

8 39 75

9 37 83

10 43 80

11 25 80

12 46 80

13 49 87

14 43 72

15 50 86

16 52 79

17 44 80

18 47 89

19 51 81

20 41 72


Sunny Days recognizes that flash sterilizations are putting patients at risk.

They acquire additional instruments, and deploy them beginning in week 20

How should they plot the new data?

P76


After 10 weeks, the team reviews their data.

Was the change successful?

What should they do with the control chart?

P77

UCL

LCL

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1 2 3 4 5 6 7 8 9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

Flash Sterlilizationu-chart

Sterilizations per surgery

New equipment


The team established a new baseline for weeks 21 – 30, and continued to plot new data.

Has the team maintained their improvement?

What is your assessment?

P78

Scaling U-Charts

79

ADE rate per dose: hard to read!

These graphs convey identical information:

ADE rate per 1,000 doses:(# of doses / 1000)

Inpatient Medication Process

Review: What Measure? What Chart? Why?

Errors per 100 orders

U-chart

Average timeXbar/S

Individuals

Percent orders with incorrect dose

P-chart

Percent with incorrect drug

P-chart

Count of orders Individuals

Average timePercent within limit

Xbar/SP-chart

Measuring Infrequent Events

Is this team improving its infection rate? What is the

problem? How could you fix it?

Holy Family HospitalRate of occurrence of hospital-acquired MRSA infections

If a P or U chart has more than 25% zeros, switch to a

time- or cases-between measure.


Rate of occurrentce of MRSA BSI and HAP per 1000 patient days

0.00

0.50

1.00

1.50

2.00

2.50

Sept Oct Nov Dec Jan

When events are infrequent, it’s hard to see whether our work is having the intended result.


Plotting the number of days since the prior event shows that infections are becoming steadily less frequent.

15 days since last event (today 1/31/2008)

Note – frequency of

cases should be

relatively constant!


Time between: Number of days between events

Use T-Chart

Assumes volume is relatively constant

Data are just dates of occurrence

Cases between: Number of processed items (e.g. cases,

patients) between events

Use G-Chart

Standardizes volume

Requires more complex data extraction

G-Chart

for

Events-

Between

Rare Events: Joint Revisions*

Have we

improved?

*Simulated Data

T-Chart plots days

between events.

ReferencesBenneyan, J. C. (2001). "Number-between g-type statistical quality control charts for

monitoring adverse events." Health Care Manag Sci 4(4): 305-18.

Benneyan, J. (2008). "Design, use and performance of statistical control charts for clinical

process improvement." International Journal of Six Sigma 4(3): 219-239.

Langley, G. J., K. M. Nolan, et al. (2009). The improvement guide : a practical approach

to enhancing organizational performance. San Francisco, Jossey-Bass.

Moen, R. D., T. W. Nolan, et al. (1999). Quality improvement through planned

experimentation. New York, McGraw Hill.

Mohammed, M. A., P. Worthington, et al. (2008). "Plotting basic control charts: tutorial

notes for healthcare practitioners." Qual Saf Health Care 17(2): 137-145

Perla, R. J., L. P. Provost, et al. (2011). "The run chart: a simple analytical tool for

learning from variation in healthcare processes." BMJ Qual Saf 20(1): 46-51.

Provost, L. P. and S. K. Murray (2010). The Data Guide - Learning from data to improve

health care. Austin TX, Associates in Process Improvement - www.pipproducts.com.

http://www.pipproducts.com/

1e control charts - ihi home pageapp.ihi.org/.../presentation_1e_-_control_charts.pdf · ·...

Documents