From Ideas to Action: Practical Tips for Making Improvements Dr Donald Campbell Clinical Epidemiology & Health Service EvaluationUnit.

From Ideas to Action:Practical Tips for Making Improvements

Dr Donald Campbell

Clinical Epidemiology & Health Service EvaluationUnit

Improvements

the will to make the change

the ideas to make the change

the execution of the ideas

Commitment to measurement and reporting

Topics

Aims, Measures, Changes

The Model for Improvement and the PDSA Cycle

Collecting and Displaying Data

The Breakthrough Collaborative

Brings together multiple sites with common aims

Working to spread and adapt existing knowledge

not developing new knowledge

Fundamental Questions for Improvement

What is the aim?

What will be measured to know the aim has been achieved?

What are the changes?

Aims focused on delays

Patient Satisfaction

Clinical

Operational

Measures

The key measures should operationalize the aim– LOS for admitted, discharge, and fast track

– Clinical improvements

– Patient satisfaction scores

Collect data on sub-components of the system judiciously

- ie, only if it is necessary

Changes

Based on your aims, identify the key changes in need of development/improvement in your system

Solicit input from others when you return to your organization

Median Time to Antibiotics AdministrationAnne Arundel Medical Center

120

140

160

180

200

220

4th Q-98 1st Q-99 2nd Q-99 3rd Q-99

Tim

e in

Min

utes

Time To AnalgesiaRoyal Melbourne Hospital

0102030405060708090

100

Week

Tim

e in

Min

utes Excessive workload

ED Median Total Length of Stay St Elsewhere’s Medical Center

110

120

130

140

150

160

170

12-A

pr

26-A

pr

10-M

ay

24-M

ay

07-J

un

21-J

un

05-J

ul

19-J

ul

02-A

ug

16-A

ug

30-A

ug

13-S

ep

27-S

ep

11-O

ct

25-O

ct

08-N

ov

Tim

e in

min

utes

Week

Median Time From Bed Requestedto Patient Upstairs

District General Hospital

30

35

40

45

50

55

60

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

Weeks

Tim

e in

Min

utes

Percent Left Without Seeing Doctor (1999)Royal King Arthur Hospital

0

2

4

6

8

10

Jan Feb Mar Apr May Jun Jul Aug Sep

Per

cent

Some Things to Consider When Making Improvements

Multiple PDSA Cycles (and time) are usually needed to adapt a change

Pay attention to detail Measurement - useful not perfect Promote the project Overcoming barriers to achieving success Hold the gains

Model for Improvement

Act Plan

Study Do

What is the aim?

What will be measured to know the aim has been achieved?

What are the changes?

Model for Improvement

Act Plan

Study Do

What are we trying to accomplish?

How will we know that a change is

improvement?

What change can we make that will

result in improvement?

Use of Data

Theories Ideas

Changes That Result in

Improvement

A P

S D

APS

D

A P

S DD S

P ADATA

1a

1b

1c

1d

160

180

200

220

240

260

280

300

320

LO

S (

min

ute

s)

Goal

1 2 3

Median LOS for Admitted Patients

Week

1. “quick-look” x-rays 3. Bed ahead

2. Work-up done on floor

Useful Measurement

Data directly related to aims Data collected in cycles to determine the

effect of a particular change Qualitative data to assist in refining a

change narrow bandwidth & stay on the money

Collecting Data

Use purposive sampling to conserve resources - Sample data daily for Fast Track, Main ED, Admitted - Summarize data weekly using the median to lessen the effect of outliers - To calculate Total LOS, use a weighted average of the medians

Integrate measurement into the daily routine

Collecting Data

Use sampling to conserve resources - Sample data daily for Fast Track, Main ED, Admitted - Summarize data weekly using the median to lessen the effect of outliers - To calculate Total LOS, use a weighted average of the medians

Integrate measurement into the daily routine

Operationalising Data Collection -some examples

Time to analgesia– pen/paper stuck on narcotics safe

– pain scale at triage

Fast track– identify on computer (or manually on assigned cubicle)

Ottawa ankle rules– aide memoire at triage and/or in cubicles

Length of Stay for Main ED Discharged Patientsn=1 per week

50

100

150

200

250

300

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29

Week

LO

S (

Min

.)

Avg=180, SD=50

Avg=135, SD=35

Median Length of Stay for Main ED Discharged Patients

n=14 per week

100

120

140

160

180

200

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29

Week

LO

S (

Min

.)

Median Length of Stay for Main ED Discharged Patients

n=28 per week

100

120

140

160

180

200

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29

Week

LO

S (

Min

.)

Median Length of Stay for Main ED Discharged Patients

n=300 per week

100

120

140

160

180

200

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29

Week

LO

S (

Min

.)

Example of a Data Collection Form

Date 0900 1400 2000 0100

5/14 Adm LOS

ED LOS

Fast Track

5/15 Adm LOS

ED LOS

Fast Track

Weekly Medians: Total LOS_________ ED LOS__________ Adm LOS_________ Fast Track_________

From the Wisconsin State Journal

Half Still Under Median

Despite the increase, union officials said about half the league’s players still earned less than the league-wide median of $75,000.

Understanding DemandEmergency Department Demand by Day

0

1

2

3

4

5

6

Saturd

ay

Sunday

Mon

day

Tuesday

Wed

nesday

Thursday

Friday

Average Number

of Patients

per Hour

Day

Understanding DemandEnhancing the understanding

Average Number

of Patients

per Hour

What about recording pts/hour by hour of the day?

What about pts/hour by day of the week?

0

1

2

3

4

5

6

Emergency Department Demand by Hour

0

1

2

3

4

5

6

7

8

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

Average Number

of Patients

per Hour

Hour of the Day

Good Visual Display of Data?

265

270

275

280

Before After

Attending MDS Do Work-up Outside ED

Goal for pneumonia:

– administer antibiotics to 95 % of patients w/in 60 mins of arrival to ED

Impact of early antibiotic administration:

– decreased mortality/decreased length of stay/reduced costs

56%

79%

65%

94%

0%10%20%30%40%50%60%70%80%90%

100%

March April May June

Pneumonia March-June 1998

Some Tips for Designing Good Graphics

Eliminate ink that does not add information Show the data Make good use of space

-Scale the graph so the data eventually encompasses most of the graphing area - Include information about multiple characteristics on the same graph or use multiple graphs on the same page

Integrate words with the data

Improving LOS for Admitted Patients from the ED

160

180

200

220

240

260

280

300

320

LO

S (

min

utes

)

Goal

Work-up done on floor

Bed ahead

Individual responsiblefor bed control

Quick-look x-rays

2/16/98 3/16 4/13 5/11 6/8 7/6

Week

The Key:Holding the Gains

Document processes and adhere to standards

Pay attention to orientation and training Assign ownership Use measurement and audits

The test: if you aren’t measuring you aren’t really trying!

Bibliography

Berwick D. A primer on leading the improvement of systems. BMJ 1996; 312: 619-622.

Berwick D, Nolan T. Physicians as leaders in improving health care: a new series in Annals of Internal Medicine. Ann Intern Med. 1998; 128:289-292.

Brock W, Nolan K, Nolan T. Pragmatic science: accelerating the improvement of critical care. New Horizons 1998; 6: 61-68.

Cook T, Campbell D. Quasi-Experimentation. (Boston: Houghton Mifflin 1979)

Langley J, Nolan K, Nolan T, Provost L.. The Improvement Guide. (San Francisco: Jossey-Bass 1996)

Lundberg G, Wennberg J. Editorial: a new proposal adn a call to action. JAMA 1997; 278: 1615-1616.

Rodgers E. Diffusion of Innovations. (New York: The Free Press 1995.

Tufte, E., The Visual Display of Quantitative Information, (Cheshire, CT:Graphics Press,1983)