Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
9/10/2013
1
Basic Tools for
Data Analysis
Michael D. Chance MSM, MBA, MSQM, CPHQ, CQIA, CSSGB
Page 1
xxx00.#####.ppt 9/10/2013 10:24:34 AM
Objectives
1. Define data.
2. Explain some barriers to successfully
using data.
3. Explain the purpose and use of select
quality data tools
4. Explain common misconceptions and
limitations that arise from reporting
“averages” or from relying on a single
tool.
Page 2
xxx00.#####.ppt 9/10/2013 10:24:34 AM
What is Data?
Data is:
Factual information used as a basis for reasoning, discussion, or calculation.
9/10/2013
2
Page 3
xxx00.#####.ppt 9/10/2013 10:24:34 AM
What Things are Considered Data?
• Medications Given (Dosages, Routes, etc.)
• Vital Signs / Symptoms
• Patient Characteristics (Age, Sex, Race, etc)
• Costs
• Therapies
• Staff Turnover, Working Hours, Ratios
• Times (Admission, Discharge, etc.)
• Anything we can document to measure performance.
Page 4
xxx00.#####.ppt 9/10/2013 10:24:34 AM
“We measure performance in
healthcare for two basic purposes.
We measure first as a basis for
making judgments and decisions…
Second, we measure as the basis
for future improvements”
Dennis O’Leary Former President,
JCAHO
Page 5
xxx00.#####.ppt 9/10/2013 10:24:34 AM
Barriers To Putting Data Into Action
• Don’t even know where to get data / info
• Paralysis by analysis
• No one is interested in it
• Incorrect interpretation of data
• Too complex to understand
• Defensiveness
9/10/2013
3
Page 6
xxx00.#####.ppt 9/10/2013 10:24:34 AM
Stages of Coping with Data
•Stage I: “The data are wrong….”
•Stage II: “The data are right, but it’s not
a problem…”
•Stage III: “The data are right, it’s a
problem, but it’s not my problem…”
•Stage IV: “The data are right, it’s a
problem, it’s my problem…”
Page 7
xxx00.#####.ppt 9/10/2013 10:24:34 AM
How Do We Make Sense of Data?
Page 8
xxx00.#####.ppt 9/10/2013 10:24:34 AM
Quality Improvement Tools
“If the only tool you have is
a hammer, you will see
every problem as a nail.”
Abraham Maslow, 1966
9/10/2013
4
Page 9
xxx00.#####.ppt 9/10/2013 10:24:34 AM
Purpose of QI Tools
•Describe and improve processes
•Evaluate process or output
variation
•Assist with decision-making
•Analyze data in a variety of ways
•Display information
Page 10
xxx00.#####.ppt 9/10/2013 10:24:34 AM
QI Tools Help Answer 5 Questions
1. Where am I at?
2. Where do I want to be?
3. How do I get there?
4. Am I still on the right path?
5. How well did I do?
Page 11
xxx00.#####.ppt 9/10/2013 10:24:34 AM
Partial List of QI Tools
Affinity diagram Arrow diagram Balanced scorecard Benchmarking Box and whisker plot Brainstorming Cause-and-effect/Ishikawa/fishbone diagram Cause analysis tools Check sheet Control chart Critical incident Data collection and analysis tools Decision matrix Design of experiments (DOE) Evaluation and decision-making tools Failure mode effects analysis (FMEA) Fishbone/Ishikawa/cause-and-effect diagram Five S (5S) Five whys and five hows Flowchart Force field analysis Gage repeatability Gantt chart Histogram House of quality Idea creation tools Impact effort matrix
Kano model Matrix diagram Mistake-proofing Multivoting Nine windows Nominal group technique Pareto chart Plan-do-check-act (PDCA) cycle or plan-do-study-act (PDSA) cycle Problem concentration diagram Process analysis tools Process decision program chart (PDPC) Project planning and implementation tools Quality function deployment (QFD) Relations diagram Scatter diagram Seven basic quality tools Seven new management and planning tools SIPOC+CM diagram SMART matrix Spaghetti diagram Stratification Success and effect diagram Survey Tree diagram Value stream mapping Voice of the customer table (VOCT)
9/10/2013
5
Page 12
xxx00.#####.ppt 9/10/2013 10:24:35 AM
Purpose of QI Tools
However,
you don’t
have to use
EVERY tool
for every
problem.
Page 13
xxx00.#####.ppt 9/10/2013 10:24:35 AM
Histogram
Pareto Chart
Scatter Diagram
Run Chart
Control Chart
Basic Decision Making Toolbox
Page 14
xxx00.#####.ppt 9/10/2013 10:24:35 AM
Basic Decision Making Tools
•Bar Charts
9/10/2013
6
Page 15
xxx00.#####.ppt 9/10/2013 10:24:35 AM
Histograms
• A bar graph that shows the distribution
of CONTINUOUS data
• A snapshot of data taken from a process
• Summarize large data sets graphically
• Compare process results to specification
• Communicate information to the team
• Assist in decision making
Page 16
xxx00.#####.ppt 9/10/2013 10:24:35 AM
Histogram Creation
Range No. of Occurrences
1 – 10
11 – 20
21 – 30
31 – 40
41 – 50
51 – 60
61 – 70
71 – 80
81 – 90
91+
Check Sheet:
Page 17
xxx00.#####.ppt 9/10/2013 10:24:35 AM
Histogram Creation
Histogram:
9/10/2013
7
Page 18
xxx00.#####.ppt 9/10/2013 10:24:35 AM
Histogram Analysis
Absorption Time
Fre
qu
en
cy
403632282420
20
15
10
5
0
Histogram of Absorption Time
30
5
X
Variable N Mean St Dev Minimum Median Maximum
Absorption Time 100 30.009 5.002 13.759 30.694 42.076
Mean (or Average):
Standard Deviation:
Descriptive Statistics:
Page 19
xxx00.#####.ppt 9/10/2013 10:24:35 AM
Histogram Example
Asthma Related Study
• Of the 835 children who performed the
free running test, 2 experienced
considerable dyspnea, cough, and
wheezing that prevented them from
completing the test. There was a 10%
decrease in PEFR in 285 (34%)
subjects, a 15% decrease in 177
(21.2%), and a 20% decrease in 69
(8.2%).
• Mean decreases in PEFR during the
free running test are shown in the
histogram in Figure 2. The distribution
was skewed to the left and most
values were between 0 and -20. The
interval with the largest concentration
of results (approximately 20% of the
total sample) was between -5 and -10.
Page 20
xxx00.#####.ppt 9/10/2013 10:24:35 AM
• Histograms are a
snapshot in time and
show “distribution”.
• They do NOT show
trends over time.
Histogram Analysis
0
5
10
15
20
25
0
5
10
15
20
25
0
5
10
15
20
25
0
1
2
3
4
5
6
7
9/10/2013
8
Page 21
xxx00.#####.ppt 9/10/2013 10:24:35 AM
Basic Decision Making Tools
•Pareto Chart
Page 22
xxx00.#####.ppt 9/10/2013 10:24:35 AM
What is a Pareto Chart?
•Bar chart arranged in
descending order of height
from left to right
•Bars on left relatively more important than those on right
•Separates the "vital few"
from the "useful many"
(Pareto Principle)
•80/20 Rule
-80% of the gain from 20% of
the categories
Budget Allocation
80
3827
14 115
0
20
40
60
80
100
Co
de
20
Co
de
10
Co
de
50
Co
de
30
Co
de
40
Co
de
60
Thousands
Page 23
xxx00.#####.ppt 9/10/2013 10:24:35 AM
Why use a Pareto Chart?
• Breaks big problems into smaller pieces
• Displays causes or problems in priority order
• Identifies most significant factors
• Shows where to focus efforts
• Allows better use of limited resources
9/10/2013
9
Page 24
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Case Study
Overall Rating
75%
80%
85%
90%
95%
100%
1-24 2-7 2-21 3-6 3-20 4-3 4-15 4-30 5-15 5-30 YTD
Goal Satisfaction
Where do we start?
Page 25
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Example - Pareto Chart
Delay: 27
Communication: 15
Procedural / Competency Issues: 15
Staff Issues: 13
Environment – Cold: 12
Documentation/Admin/Billing: 11
Appointments: 7
Entertainment: 4
Location: 4
Parking: 4
Gown Size: 3
Other: 7
N = 122
Page 26
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Example - Pareto Chart
27
15 15
1312
11
7
4 4 43
7
22%
34%
47%
57%
67%
76%
82%85%
88%92%
94%
100%
0
5
10
15
20
25
30
Dela
y
Com
m
Pro
c/Com
pSta
ff
Env
iron
Adm
inApp
t
Ent
erta
in
Loca
tion
Par
king
Gow
n
Oth
er
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Complaints Cumulative
N = 122
9/10/2013
10
Page 27
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Example - Pareto Chart
27
15 15
1312
11
7
4 4 43
7
22%
34%
47%
57%
67%
76%
82%85%
88%92%
94%
100%
0
5
10
15
20
25
30
Dela
y
Com
m
Pro
c/Com
pSta
ff
Env
iron
Adm
inApp
t
Ent
erta
in
Loca
tion
Par
king
Gow
n
Oth
er
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Complaints Cumulative
N = 122
Break point
Page 28
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Example - Pareto Chart
Page 29
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Basic Decision Making Tools
•Scatter Diagram
9/10/2013
11
Page 30
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Scatter Diagrams
A graph of paired data points plotted on a
table that helps identify the possible
relationship between the changes observed
in two different sets of variables.
Page 31
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Why use Scatter Diagrams?
• Supplies the data to confirm a hypothesis that two
variables are related.
• Provides both a visual and statistical means to test
the strength of a potential relationship.
• Provides a good follow-up to a Cause and Effect
Diagram to find out if there is more than just a
consensus connection between causes and the
effect.
Page 32
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Interpreting Scatter Diagrams
0
1
2
3
4
5
6
7
8
9
10
Pai
n F
ree
Tim
e
Amount of Drug Given
Does Amount of Drug Given Affect Pain-Free Time?
Correlation
0
0
1
2
3
4
5
6
7
8
9
10
Pat
ien
t S
atis
fact
ion
Rat
ing
Patient Waiting Time
Does Patient Waiting Time Affect Patient Satisfaction?
0
1
2
3
4
5
6
7
8
9
10
Min
ute
s O
ut
of
Res
trai
nt
Mg/Kg Drug Given
Does Amount of Drug Given Affect Minutes Patient Out
of Restraints?
0.00
5.00
10.00
15.00
20.00
0 50 100 150
Pre
ssu
re (
PSI
)
Temperature (F)
Pressure of Gas at Temperature
9/10/2013
12
Page 33
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Is there a correlation?
Interpreting Scatter Diagrams
97.998
98.198.298.398.498.598.698.798.898.9
9999.199.299.399.499.599.699.799.899.9100
0 10 20 30 40 50 60 70
Ou
tco
me
Input
Scatter Diagram
Outcome (y)
Linear (Outcome (y))
Page 34
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Beware of False Correlation
Interpreting Scatter Diagrams
97.998
98.198.298.398.498.598.698.798.898.9
9999.199.299.399.499.599.699.799.899.9100
0 10 20 30 40 50 60 70
Ou
tco
me
Input
Scatter Diagram
Outcome (y)
Linear (Outcome (y))
Page 35
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Example Scatter Diagrams
9/10/2013
13
Page 36
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Basic Decision Making Tools
•Run Chart
Page 37
xxx00.#####.ppt 9/10/2013 10:24:36 AM
What is a Run Chart?
A line graph of data points plotted in
chronological order that helps detect
special causes of variation.
Page 38
xxx00.#####.ppt 9/10/2013 10:24:36 AM
What is a Run Chart?
•A running record of process behavior over time.
•Requires no statistical calculations.
•Shows process behavior at a glance.
•Can detect some special causes.
•Time sequence is plotted on horizontal axis.
•Measure of interest is always plotted on the vertical axis.
•Center Line is the mean score (or median).
9/10/2013
14
Page 39
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Run Chart
Patient Restraint Rate
0.00%
0.10%
0.20%
0.30%
0.40%
0.50%
0.60%
0.70%
0.80%
0.90%
1.00%
Ja
n-0
2
Feb
-02
Ma
r-0
2
Ap
r-0
2
Ma
y-0
2
Ju
n-0
2
Ju
l-0
2
Au
g-0
2
Sep
-02
Oct-
02
No
v-0
2
Dec-0
2
Ja
n-0
3
Feb
-03
Ma
r-0
3
Ap
r-0
3
Ma
y-0
3
Ju
n-0
3
Ju
l-0
3
Au
g-0
3
Sep
-03
Oct-
03
No
v-0
3
Percen
tag
e R
est
ra
ined
Page 40
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Example Run Chart
Page 41
xxx00.#####.ppt 9/10/2013 10:24:36 AM
Basic Decision Making Tools
•Control Chart
9/10/2013
15
Page 42
xxx00.#####.ppt 9/10/2013 10:24:37 AM
What is a Control Chart?
A statistical tool used to distinguish
between process variation resulting
from common causes and variation
resulting from special causes.
Page 43
xxx00.#####.ppt 9/10/2013 10:24:37 AM
When people do not
understand variation
•see trends where there are no trends
•blame and give credit to others for things over
which they have little or no control
•build barriers, decrease morale, and create an
atmosphere of fear
•never be able to fully understand past
performance, make predictions about the future
and make significant improvements in processes
Page 44
xxx00.#####.ppt 9/10/2013 10:24:37 AM
Control Chart
•Elements of a Control Chart
Measurement
Scale
Horizontal Axis
x-axis
Time Units or Sequence
10
9
8
7
6
5
4
3
2
1
0
Centerline
1 5 10 15 20
Vertical Axis
y-axis Upper Control Limit
(UCL)
Lower Control Limit
(LCL)
9/10/2013
16
Page 45
xxx00.#####.ppt 9/10/2013 10:24:37 AM
Why use Control Charts?
• Monitor process variation over time
• Differentiate between special cause and
common cause variation
• Assess effectiveness of changes
• Establish the basis for determining process
capability
• Communicate process performance
Page 46
xxx00.#####.ppt 9/10/2013 10:24:37 AM
Control Chart vs Bar Chart
90
73 74
3543
51 48
05
1015
2025
3035
404550
5560
6570
7580
8590
95100
Sun Mon Tue Wed Thu Fri Sat
Consecutive Patients
Tim
e i
n M
inu
tes
Daily Avg
Goal
Weekly Avg
N=250 WOW!
Goal <60
Avg = 59
Page 47
xxx00.#####.ppt 9/10/2013 10:24:37 AM
Control Chart vs Bar Chart
0
25
50
75
100
125
150
175
200
225
250
275
Consecutive Patients
Tim
e i
n M
inu
tes
Door to MD
Avg
N=250
9/10/2013
17
Page 48
xxx00.#####.ppt 9/10/2013 10:24:37 AM
Control Chart vs Bar Chart
0
25
50
75
100
125
150
175
200
225
250
275
Consecutive Patients
Tim
e i
n M
inu
tes
Door to MD
Avg
Goal
UCL
LCL
N=250 Goal = <60
Mean = 59
Std Dev = 55
UCL = 224
LCL = 0
Page 49
xxx00.#####.ppt 9/10/2013 10:24:37 AM
Control Chart vs Bar Chart
0
25
50
75
100
125
150
175
200
225
250
275
Consecutive Patients
Tim
e i
n M
inu
tes
Door to MD
Avg
Goal
UCL
LCL
N=250 Goal = <60
Mean = 59
Std Dev = 12
UCL = 95
LCL = 23
Page 50
xxx00.#####.ppt 9/10/2013 10:24:37 AM
Example Control Chart
9/10/2013
18
Page 51
xxx00.#####.ppt 9/10/2013 10:24:37 AM
Improvement Strategies
After making a Run Chart or a Control chart, what’s next?
The type of variation determines your approach:
SPECIAL CAUSE VARIATION?
If negative, eliminate it.
If positive, emulate it.
But don’t change the process!
COMMON CAUSE VARIATION?
If process is functioning at an unacceptable level, change the process!
Don’t “tamper” with individual data points!
Page 52
xxx00.#####.ppt 9/10/2013 10:24:37 AM
Resources
The Memory Jogger 2
- Goal QPC
The Quality Toolbox
- Nancy Tague
Page 53
xxx00.#####.ppt 9/10/2013 10:24:37 AM
Basic Tools for Data Analysis
Michael D. Chance
MSM, MBA, MSQM, CQIA, CPHQ, CSSGB
Quality Improvement Specialist
Phone: 832-824-1308 Email: [email protected]