Jane Norman

President of PKP, Inc is an internationally recognized consultant on leadershipand improvement who has been an executive at Caterpillar, Inc., ConAgraInc., Conrad Company and OCHIN while focusing on integrating improvementinto the business strategy. With over 30 years of experience, she hasconsulted with leaders and developed improvement professionals inmanufacturing, food, distribution, technology, software and health care intwelve countries. She is the co-owner of both Austin API, Inc. and ProfoundKnowledge Products, Inc. which develop workshops and virtual learningenvironments which support clients and their improvement teams. She is thecreator of the Accelerated Model for Improvement (Ami™) methods,developed from the API Model for Improvement. Ami™ workshop &materials are used worldwide to define and complete improvement projectswithin 100 days. She is a Certified Quality Engineer (CQE) with a BA in NaturalScience from St. Ambrose University in Davenport, Iowa and a MBA fromRollins College in Winter Park, Florida.

Born in Salem, Oregon, Jane grew up in the Midwest as the daughter ofteachers/administrators, attending schools in Missouri, Kentucky, Iowa andFlorida. She has been a chapter officer for the American Society for Quality,and was certified as a Quality Engineer (CQE). She was also an advisor inJunior Achievement. Married to her husband, Cliff, they have five adultdaughters and 6 grandchildren. They currently reside near Austin, Texas. Sheis a member of New England Women and Daughters of the AmericanRevolution. Her hobbies are the study of US & World history, genealogy,performing music, and her grandchildren. Jane is a co-author of the book,Transforming Health Care Leadership: A Systems Guide to Improve PatientCare, Decrease Costs and Improve Population Health.

Jane & Cliff are co-authors of two new books:

Transforming Health Care Leadership: A Systems Guide to Improve Patient Care, Decrease Costs, and Improve Population Health

Patient Safety at a Glance: For Medical Students, Junior Doctors and Nurses

WorkshopUsing Data to Learn About Variation While

Avoiding Traps

Breakout Session for this Afternoon

• Moving data from data bases to information and knowledge for action. – Introduction to Software owned by IHS to produce useful

displays of data –QI Charts.• Explore the use and application of run charts and

control charts in the health care setting. • Appreciate the power and usefulness of using analytic

methods for decision making and improvement.

3

The Bead Company (TBC)A Quality company!

We want you to join our team!

#1World Class

Beads!

Help Wanted: 8 Skilled Employees

Hiring for The Bead Company

• 4 Willing Workers• Educational requirements minimal, some coordination required• Willing to do their best• Job dependent on performance (100 “good beads” per day)

• 2 Inspectors• Must be able to count accurately and write legibly

• 1 Inspector General• Must be able to count, displays management potential and can report the

results loudly to the Recorder and to management• 1 Recorder

• Write legibly, good in addition and division, computer skills helpful

Job Requirements for the Willing Workers

• Must be willing to put forth best efforts

• Continuation of job is dependent on performance (No Green beads!)

• Conform to work standards-100 beads per day – Any color but Green!

7

8

9

Anscombe’s Four Data Sets - How Are the Data Sets Different?

Data Set 1 Data Set 2 Data Set 3 Data Set 4X Y X Y X Y X Y

10.00 8.04 10.00 9.14 10.00 7.46 8.00 6.588.00 6.95 8.00 8.14 8.00 6.77 8.00 5.7613.00 7.58 13.00 8.74 13.00 12.74 8.00 7.719.00 8.81 9.00 8.77 9.00 7.11 8.00 8.8411.00 8.33 11.00 9.26 11.00 7.81 8.00 8.4714.00 9.96 14.00 8.10 14.00 8.84 8.00 7.046.00 7.24 6.00 6.13 6.00 6.08 8.00 5.254.00 4.26 4.00 3.10 4.00 5.39 19.00 12.5012.00 10.84 12.00 9.13 12.00 8.15 8.00 5.567.00 4.82 7.00 7.26 7.00 6.42 8.00 7.915.00 5.68 5.00 4.74 5.00 5.73 8.00 6.89

Average 9.0 7.5 9.0 7.5 9.0 7.5 9.0 7.5

Statistical Summary of the Four Data Sets:Each data set has 11 data points for X and Y.Each data set has the same averages for X and Y.Each data set has The same correlation coefficient for X and Y: r = 0.86Each data set has the same least-squares regression equation: Y = 3.0 + .5X

with r2 = 0.667 and standard error = 1.24.

23-6

10

Scatterplots for Anscombe’s Four Sets of Data

23-7

11

Run Chart - a graphical record of a measure plotted

over time

Unit 1

0102030405060708090

100

date

Jan

Feb

Mar Apr

May Ju

n

Jul

Aug Sep

Oct

Nov

Dec

Change Made

Cyc

le T

ime

(min

.)

12

0

10

20

30

40

50

60

70

80

Avg BeforeChange

Avg AfterChange

Cyc

le T

ime

(min

.)

Improvement in Cycle Time

70

35

13

Importance of Time Order Graph Improvement in Cycle

100 Time

0102030405060708090

date

Jan

Feb

Mar Apr

May Ju

n

Jul

Aug Sep

Oct

Nov

Dec

Change Made

Cyc

le T

ime

(min

.) Unit 2 – same before and after plot

14

Cycle Time Results for Units 1, 2 and 3

70

35

0

10

20

30

40

50

60

70

80

Avg BeforeChange

Avg AfterChange

Cyc

le T

ime

(min

.)

0102030405060708090

100

date Jan

Feb

Mar

Apr

May Jun Jul

Aug

Sep Oct

Nov

Dec

Change Made

Cycle

Time

(min.

)

Unit 1

Unit 3

Unit 2

0102030405060708090

100

date Jan

Feb

Mar Ap

r

May Ju

n

Jul

Aug

Sep

Oct

Nov

Dec

Change Made

Cycle

Tim

e (m

in.)

0102030405060708090

100

date Jan

Feb

Mar Ap

r

May Ju

n

Jul

Aug

Sep

Oct

Nov

Dec

Change Made

Cycle

Tim

e (m

in.)

Interchange of Two Points…

• Any symmetric function of a set of numbers almost always throws away a large portion of the information in the data…A statistical test is a symmetric function of the data.

• In contrast, interchange of two points in a plot of points may make a big difference in the message that the data are trying to convey for prediction.

15Dr. W. Edwards Deming 7-14-90

Reference: Quality Improvement through Planned Experimentation

Hospital Family of Measures

18

Improvement of Quality

SPC-8-18

Small Multiples Schematic

Subcomponents

1

2

3

4

5

6

Overall

© 2002 Institute for Healthcare Improvement

Patient Satisfaction GraphsOverall Satisfaction

70

75

80

85

90

95

100

Jan Feb Mar Apr May June July Aug Sept Oct Nov

Avg

. Com

posi

te S

core

70

75

80

85

90

95

100

Jan Feb Mar Apr May June July Aug Sept Oct Nov

Avg

. Com

posit

e Sco

re

Respect Comfort

• 70

• 75

• 80

• 85

• 90

• 95

• 100

• Jan• Feb• Mar• Apr• May• June• July• Aug• Sept• Oct• Nov

•A

vg. C

ompo

site

Sco

re

© 2002 Institute for Healthcare Improvement

Small Multiples:

Overall System and 22 Districts

Good

Average Waiting Times:All Primary Care Clinics in VHA System

Stratification of Data!

DG 5-11

Subgrouping on ADE Rate

System-wide Aggregate ADE Rate

10

12

14

16

18

20

J/05 F M A M J J A S O N D J/06 F M A M J

AD

E/10

00 D

oses

ADE Rate by Day of Week

Common Cause Hospitals Subgrouped in 6 Month Increments

10

11

12

13

14

15

16

17

18

19

20

M-1

M-2

M-3

T-1

T-2

T-3

W-1

W-2

W-3

H-1

H 2

H-3

F-1

F-2

F-3

S-1

S-2

S-3

U-1

U-2

U-3

AD

Es/1

000

Dos

es

Night Shift?

ADE Rate Subgroupd by ShiftCommon Cause Hospitals by Qtr

10

11

12

13

14

15

16

17

18

19

20

D1 D2 D3 D4 D5 D6 E1 E2 E3 E4 E5 E6 N1 N2 N3 N4 N5 N6

AD

E/10

00 D

oses

28

Anscombe’s Four Data Sets - How Are the Data Sets Different?

23-6

Data Set 1 Data Set 2 Data Set 3 Data Set 4X Y X Y X Y X Y

10.00 8.04 10.00 9.14 10.00 7.46 8.00 6.588.00 6.95 8.00 8.14 8.00 6.77 8.00 5.7613.00 7.58 13.00 8.74 13.00 12.74 8.00 7.719.00 8.81 9.00 8.77 9.00 7.11 8.00 8.8411.00 8.33 11.00 9.26 11.00 7.81 8.00 8.4714.00 9.96 14.00 8.10 14.00 8.84 8.00 7.046.00 7.24 6.00 6.13 6.00 6.08 8.00 5.254.00 4.26 4.00 3.10 4.00 5.39 19.00 12.5012.00 10.84 12.00 9.13 12.00 8.15 8.00 5.567.00 4.82 7.00 7.26 7.00 6.42 8.00 7.915.00 5.68 5.00 4.74 5.00 5.73 8.00 6.89

Average 9.0 7.5 9.0 7.5 9.0 7.5 9.0 7.5

Statistical Summary of the Four Data Sets:Each data set has 11 data points for X and Y.Each data set has the same averages for X and Y.Each data set has The same correlation coefficient for X and Y: r = 0.86Each data set has the same least-squares regression equation: Y = 3.0 + .5X

with r2 = 0.667 and standard error = 1.24.

29

Scatterplots for Anscombe’s Four Sets of Data

23-7

Stages of Denial with DataReactions to Challenging Data

1.The data are wrong

2.The data are right, but it’s not a real

problem

3.The data are right, it’s a real problem,

but not my problem

4.The data are right, it’s a real problem,

and it’s my problem, but I don’t need to

do anything about it

5.The data are right, it’s a real problem,

and it’s my problem and I will fix it

System 1; Quick, Intuitive System 2: Slow Analytic