QI Theory:
Quality Improvement in the Hospital
Goals for this Primer
• Understand fundamental concepts in
quality improvement
• Identify the environment and key steps for
a successful quality improvement project
• Become familiar with several quality
improvement tools and their use
Progress
Quality Improvement:
Bridging the Implementation Gap
How good is American healthcare?
Patient care
Time
Progress
Quality Improvement:
Bridging the Implementation Gap
We get it right 54% of the time.
-Brent James, MD, MStat
Executive Director, Intermountain Health Care
Patient care
Time
Progress
Quality Improvement:
Bridging the Implementation Gap
Scientific
understanding
Implementation
Gap
Patient care
Time
Hospitalists and Quality Improvement
• Complex process problems need multidisciplinary
solutions
• We are at the frontlines seeing system failures,
process errors, and performance gaps with our
own eyes -- which is our competitive advantage
• Improved quality delivers:
 better patient care…
 at lower costs…
 with potentially higher reimbursements (pay-forperformance)…
And it can make our jobs more interesting, fun, and
rewarding.
Section I:
Quality Improvement and Change
in the Hospital Atmosphere
Definition of Quality
• Meeting the needs and exceeding the
expectations of those we serve
• Delivering all and only the care that the patient
and family needs
“Definition” of Improvement
It is NOT…
 yelling at people to work harder, faster, or safer
 creating order sets or protocols and then failing to
monitor their use or effect
 traditional Quality Assurance
 research (but they can co-exist nicely)
Principle #1:
Improvement Requires Change
Every system is perfectly designed to achieve
exactly the results it gets
To improve the system, change the system…
Principle #2:
Less is More
You cannot destroy productivity
When changing the system, keep it simple
Illustrating Principle #2: Less Is More
Probability of Performing Perfectly
No.
Probability of Success, Each Element
Elements 0.95
0.99
0.999
0.999999
1
25
50
100
0.95
0.28
0.08
0.006
0.99
0.78
0.61
0.37
0.999
0.98
0.95
0.90
0.999999
0.998
0.995
0.99
Understanding Change in the Hospital
Atmosphere
• Change = not just doing something different, but
engineering something different
• at least one step in at least one process
• Hospital Atmosphere = hospitals tend to be viscous,
complex systems with default levels of performance
• change engineered to improve performance can be a foreign
concept - or even overtly resisted
Understanding Change in the Hospital
Atmosphere
•
•
•
•
•
•
A Common Strategy Which Commonly Fails:
Experts design a comprehensive protocol using EBM
over several months
Protocol is presented as a finished, stand alone
product
Customization of protocol is discouraged
Compliance depends on vigilance and hard work
Monitoring for success or failure is the exception to
the rule (with failures coming to light after patients
are harmed)
Flawed implementation leads to repetitive efforts
down the road
Understanding Change in the Hospital
Atmosphere
•
•
•
•
•
•
High-Reliability Strategies Commonly Succeed:
Build a “decision aide” or reminder into the system
Make the desired action the default action (not doing
the desired action requires opting out)
Build redundancy into responsibilities (e.g. if one
person in the chain overlooks it, someone else will
catch it)
Schedule steps to occur at known intervals or events
Standardize a process so that deviation feels weird
Take advantage of work habits or reliable patterns of
behavior
Build at least one - if not more - of these highreliability strategies into any changed process.
Understanding Change in the Hospital
Atmosphere
Change engineered to drive improvement depends on…
• Workplace Culture: personnel must be receptive to change
• Awareness: administrative and medical staffs must care
about performance and support its improvement through
change
• Evidence: local experts must identify which research to
translate into practice
• Experience: a skilled team must choose, implement, and
follow up changes to ensure:
1) improvement efforts are ongoing and yielding better
performance
2) productivity is preserved
An Atmosphere for Change
AWARENESS
EXPERIENCE
OF THE LOCAL PERFORMANCE GAP
WITH SIMILAR IMPROVEMENT
EFFORTS
Patient
Medical Staff
Administrative Support
Hospitalist Quality Officer
Multidisciplinary Team Members
Success Stories From Other Institutions
EVIDENCE
WORKPLACE CULTURE
TO TRANSLATE INTO PRACTICE
READY TO ACCEPT CHANGE
“Bedside” Teaching
Didactic Teaching Sessions
Local Expertise in Disease Literature
Task Load
Culture of Improvement
Culture of Negative Expectations
An Atmosphere for Change
AWARENESS
OF THE LOCAL PERFORMANCE GAP
Patient
Medical Staff
Hospital Administration
 Patient
At mercy and increasingly aware of
underperforming status quo
Now can access a new resource promoting
transparency in hospital performance:
www.hospitalcompare.hhs.gov
 Hospital Administration
 Medical Staff
Understands status quo is unacceptable
(IOM, Leapfrog, NQF, JCAHO)
Sees fiscal health tied to performance
against national benchmarks, ability to
reduce costs & LOS, improve margins,
and competitive reputation in the
community
Has professional responsibility to improve
Knows all too well where system fails
Recognizes that professional livelihood will
depend on paying attention to outcomes:
Pay-for-Performance
An Atmosphere for Change
Hospitalist Team Facilitator

Technical expert on Quality Improvement
theory and tools
Owns the team process, enforces ground
rules, helps judge feasibility
Teaches the team while doing
Successful Strategies Used By Others

Learn from mistakes of others
Adapt successes of others (tools and
methods): steal shamelessly
Get specific advice in ’Ask the Expert’
forums or other consortiums that collect
and share experience
EXPERIENCE
WITH SIMILAR IMPROVEMENT EFFORTS
Hospitalist Team Facilitator
Multidisciplinary Team Members
Successful Strategies Used By Others
Multidisciplinary Team Members
Chosen for hands-on, fundamental
knowledge of key processes
Inclusive, open, & consensus seeking
Impact not only the change(s) but the
implementation

An Atmosphere for Change
“Bedside” Teaching
Didactic Teaching Sessions
To an audience of residents or students
To build cadre of “experts” (and to help
meet ACGME requirements)
Download teaching pearls from SHM
resource rooms
To an audience of peers, administrators,
nurses, or support staff
To boost awareness, knowledge,
enthusiasm, and support
Download slide sets from SHM resource
rooms


EVIDENCE
TO TRANSLATE INTO PRACTICE
“Bedside” Teaching
Didactic Teaching Sessions
Local Expertise in Disease Literature
Local Expertise in Disease Literature
Decide what changes to make based on
the level of evidence
Establishes team’s credibility
Extends team’s authority when local subspecialists or experts participate in
selecting and implementing change

An Atmosphere for Change
Task Load
Culture of Improvement
Be sensitive about piling new tasks onto
over-tasked personnel
Use the input of personnel who will be
responsibile for implementing
Make it easy and desirable to do the right
thing

Extend it, one person and one project at a
time
Advertise successes
Use or adapt this online ‘cultural survey:’
http://www.patientsafetygroup.org/program/step1c.cfm

Culture of Negative Expectations
Overcome it, one person and one project
at a time
Attach pride to balance between
performance successes and failures
Consider using a ‘cultural survey’ to
identify problems and address them
through proper channels

WORKPLACE CULTURE
READY TO ACCEPT CHANGE
Task Load
Culture of Improvement vs.
Culture of Negative Expectations
Section II:
The Multidisciplinary Team
The Driving Force for Change
THE MULTIDISCIPLINARY TEAM
Leverages frontline expertise and experience.
Impacts not only the change/interventions,
but also the implementation
The Driving Force for Change:
The Multidisciplinary Team
A team is not the same as a committee…
Committee
• individuals bring representation
• productive capacity = single most able member
Team
• individuals bring fundamental knowledge
• productive capacity = synergistic (more than the sum of all
individual team members together)
The Driving Force for Change:
The Multidisciplinary Team
Features of a good team…
• Safe (no ad hominem attacks)
• Inclusive (values all potential contributors including
diverse views; not a clique)
• Open (considers all ideas fairly)
• Consensus seeking
The Driving Force for Change:
The Multidisciplinary Team
Consensus…
• definition: finding a solution acceptable enough
that all members can support it; no member
opposes it
• It is not:
 A unanimous vote (consensus may not represent
everyone’s first priorities)
 A majority vote (in a majority vote, only the majority
gets something they are happy with; people in the
minority may get something they don’t want at all,
which is not what consensus is all about)
 Everyone totally satisfied
The Driving Force for Change:
The Multidisciplinary Team
Three types of team members…
1) Team Leader
2) Team Facilitator
3) Process Owners (members with operational, hands-on
fundamental knowledge of the process)
The Driving Force for Change:
The Multidisciplinary Team
Team Leader…
• schedules and chairs team meetings
• sets the agenda (printed at each meeting)
• records team activities (working documents in
binder)
• reports to management (Steering Team)
• often a member of Steering Team
The Driving Force for Change:
The Multidisciplinary Team
Team Facilitator…
•
•
•
•
owns the team process (enforces ground rules)
technical expert on QI theory and tools
assists Team Leader
teaches while doing, within team
The Driving Force for Change:
The Multidisciplinary Team
Process Owners…
• chosen for fundamental knowledge
• will help implement
• should become leaders (so choose wisely)
The Driving Force for Change:
The Multidisciplinary Team
Team Ground Rules…
• All team members and opinions are equal
• Team members will speak freely and in turn
 We will listen attentively to others
 Each must be heard
 No one may dominate
• Problems will be discussed, analyzed, or attacked (not people)
• All agreements are kept unless renegotiated
• Once we agree, we will speak with "One Voice" (especially after leaving the meeting)
• Honesty before cohesiveness
• Consensus vs. democracy: each gets his say, not his way
• Silence equals agreement
• Members will attend regularly
• Meetings will start and end on time
A Brief Digression into Quality
Improvement Theory
Defining an Approach to Change
Will the team target ‘all’ patients in the
inpatient bell curve, or just a sub-group
considered ‘at-risk’ (depicted in the
outlying tail)? Is the quality of inpatient
care which is not in the tail somehow
‘acceptable?’
Before
Bell Curve:
Inpatient Population
Tail
worse
Defining an Approach to Change
If the team can identify and define an inpatient sub-group
‘at-risk,’ then improvement efforts could conceivably
focus just on these ‘at-risk’ patients - this is similar to
traditional Quality Assurance. Note that even if tail
events are eliminated, the quality of care for the rest of
the inpatient population (depicted by the unchanged
position and shape of the bell curve) does not improve at
all. While the mean does move toward better care, this is
due only to eliminating statistical outliers.
After
Before
Bell Curve:
Inpatient Population
Tail
worse
worse
Quality
Defining an Approach to Change
If the team identifies a performance gap applicable to a
wider patient population, the team may design changes in
processes with the potential for dramatic effect:
improvement and standardization in processes reduces
variation (narrows the curve) and raises quality of care for
all (shifts entire curve toward better care). This radical
change is what defines Quality Improvement.
After
Before
Bell Curve:
Inpatient Population
worse
Quality
Tail
better
worse
worse
Quality
Section III:
Tools for Engineering Change
Engineering Change
• Hospitals have two dynamic levels impacting
performance:
1) Processes
• tasks performed in series or in parallel, impacting patient care
and potentially patient outcomes
2) Personnel
• skilled people with hearts and minds, with variable levels of
attention, time, and expertise
Engineering Change:
What Variables Impact Quality Outcomes of Care?
Structure
Processes
Inputs
Steps
•Patients
•Equipment
•Supplies
•Training
•Environment
•Inventory Methods
•Coordination
•Physician orders
•Nursing Care
•Ancillary staff
•Housekeeping
•Transport
Outcomes of Care
Outputs
•Physiologic
parameters
•Functional status
•Satisfaction
•Cost
Engineering Change:
What Variables Impact Quality Outcomes of Care?
The two most dynamic levels impacting performance
Processes
Steps
•Inventory Methods
•Coordination
•Physician orders
•Nursing Care
•Ancillary staff
•Housekeeping Personnel
•Transport
Engineering Change
• Processes
 all those affecting relevant aspects of patient
care
• clinical decision making, order writing, admission
intake, medication delivery, direct patient care,
discharge planning, PCP communication,
discharge follow-up, etc
Engineering Change
• Personnel
 anybody who touches the patient or a relevant
process in the system
• departments, physicians, clerks, pharmacy,
nursing, RT, PT/OT/ST, care technicians,
phlebotomist, patient transport, administration
Engineering Change:
The Multidisicplinary Team Asks “What?”
• What?
 is the right thing to do?
 will make the system more effective?
Engineering Change:
The Multidisicplinary Team Asks
“Where?”
• Where?
 are the processes to improve?
• Brainstorming
• Multivoting & nominal group technique
• Affinity grouping
 do we start? (dissect and understand the processes)
•
•
•
•
•
•
•
Cause and effect diagrams (Ishikawa or ‘fishbone’ diagrams)
Tally sheets
Pareto charts
Flow (conceptual flow, decision flow) charts
Run charts
SPC charts
Scatter charts
Tools for Engineering Change:
Cause-and-Effect Diagram
• sometimes also called a ‘fishbone’ or Ishikawa diagram
• graphically displays list of possible factors, focused on
one topic or objective
• used to quickly organize and categorize ideas during a
brainstorming session, often as an interactive part of the
session itself (the added organization can help produce
balanced ideas during a brainstorming session)
Tools for Engineering Change:
Cause-and-Effect Diagram
Example: Adverse Drug Events (ADE)
Drug
Administration
Errors
Ordering
Errors
Nurse
Physician
Pharmacist
Physician
Pharmacy
Nurse/Clerk
Transcribing
Rate
Dilution
Spelling
Route
Time
Route
Scheduling
Nurse
Dose
Place outcome here
Dosage
Order Missed
Wrong
Drug
Age
Psychiatric
Gender
ADE
Unforeseen
Weight
Expected
Drug/Drug
Renal
Cognitive
Compliance
Electrolyte
Hepatic
Race
Patient
Errors
Past Allergic
Reaction
Absorption
Physiologic
Factors
Patient
Drug/Food
Drug/Lab
Pharmacokinetics
Pharmacodyamics
Pharmocologic
Factors
Pharmacist
Patient
Physician
Dietician
Tools for Engineering Change:
Pareto Chart
• graphical display of the relative weights or frequencies of competing
events, choices, or options
• a bar chart, sorted from greatest to smallest, that summarizes the
relative frequencies of events, choices, or options within a class
• often includes a cumulative total line
• used to focus within a broad category containing many choices, based
on factual or opinion-based information
• can combine factors that contribute to each item's practical
significance
Tools for Engineering Change:
Pareto Chart
100
Causes Contributing to Adverse Drug Events
90
80
Percent
Contributing
70
60
50
40
30
20
10
0
Causes
Causes
Tools for Engineering Change:
Sketching Processes or Flow
• Macro Process Maps
• Decision Flow Diagrams
The patient is
admitted to the
hospital
Tools for Engineering Change:
Macro Process Map
Example: Heart Failure Core Measures 2-3
The patient is
clinically identified
as having heart
failure
The ejection fraction
is evaluated
The patient is
prescribed an ACEI
in hospital
The patient is
prescribed an ACEI
at discharge
The patient is not
prescribed an ACEI
in hospital
The contraindication
for an ACEI is
documented in the
chart
The ejection fraction
< 40%
The ejection fraction
is documented in the
chart
The ejection fraction
> 39%
The patient is
excluded from the
target population
Tools for Engineering Change:
Decision Flow Diagram
Deep Post-Op
Wound Infection
UTI
Contributing layer dissected:
Prevention
Prevention
Pneumonia
Prevention
Patient
Preparation
Bacteremia
Other
Contributing layer dissected:
Prophylactic Antibiotics
Prophylaxis
Patient
Selection
Detection
Prophylactic
Antibiotics
Antibiotic
Selection
Surgery
- Duration
- Sterile Technique
- Operative Findings
Treatment
Delivery
- Timing
For iatrogenic infections, any
given type of infection can be
dissected into the hierarchy of
contributing layers.
Post-Op
Wound Care
Calling out the contributing layers
helps the team think through the steps
ripest for change.
Tools for Engineering Change:
Run Charts
• Our brains understand graphics better than tables
• Tabular information doesn’t convey trends over time very
well
• Keep it simple
• In center of horizontal axis place: baseline mean
performance
• In center of vertical axis place: implementation point
• Can add upper and lower control limits, but usually not
needed
Tools for Engineering Change:
Run Charts
Percent Sliding Scale Insulin Only
80
70
50
10/20/03
New Order Set
40
30
01/20/04
CPOE - TH
20
10
Au
g04
Ju
n04
04
Ap
r-
Fe
b04
-0
3
D
ec
ct
-0
3
O
Au
g03
Ju
n03
03
Ap
r-
Fe
b03
-0
2
D
ec
ct
-0
2
0
O
Percent
60
Tools for Engineering Change:
Run Charts
Percent with Frank Hypoglycemic Events
16
14
10
8
10/20/03
New Order Set
6
March 2003
Team Forms
4
2
CPOE
TH - 1/04
HC - 8/04
Ju
n04
Au
g04
O
ct
-0
4
D
ec
-0
4
Fe
b05
Ju
n03
Au
g03
O
ct
-0
3
D
ec
-0
3
Fe
b04
Ap
r04
Fe
b03
Ap
r03
-0
2
D
ec
ct
-0
2
0
O
Percent
12
Tools for Engineering Change:
Run Charts
Percent with Optimal/Acceptable Glucose Readings
100
90
80
60
10/20/03
New Order Set
50
March 2003
Team Forms
40
CPOE
TH - 1/04
HC - 8/04
30
20
10
Ju
n04
Au
g04
O
ct
-0
4
D
ec
-0
4
Fe
b05
Ju
n03
Au
g03
O
ct
-0
3
D
ec
-0
3
Fe
b04
Ap
r04
Fe
b03
Ap
r03
-0
2
D
ec
ct
-0
2
0
O
Percent
70
Engineering Change:
The Multidisicplinary Team Asks “How?”
• How?
 can you make it easy to do the right thing?
• You cannot destroy productivity
– Changes must maintain, or enhance, workplace efficiency or balance
• You must devote as much attention to fitting changes into clinical
work flow as you do to the evidence-based guideline
– Changes must be blended into the flow of clinical care
– Important variables to consider: staffing, training, supplies, physical
layout, information flow, and educational materials
Engineering Change
Improve incrementally. Learn through action.
Plan Do Study Act
PDSA PDSA  PDSA  PDSA PDSA  PDSA
Test your changes. Assess their effect.
Then re-work the changes and do it again…and again…
Engineering Change:
PDSA
(the Benefits of Repeated Cycles)
• Increases belief that change will result in
improvement
• Allows opportunities for “failures” without
impacting performance
• Provides documentation of improvement
• Adapts to meet changing environment
• Evaluates costs and side-effects of the change
• Minimizes resistance upon implementation
Engineering Change:
PDSA
• Overview:
 scientific method for action-oriented learning:
shorthand for testing a change in the real world setting
 test a change by: planning it, trying it, measuring its
results… and then trying to do it better the next time
 multiple rounds of changes – some failures and some
successes - should lead to improved aggregate
outcome
Engineering Change:
PDSA
• Principles for Success:
 start new changes on the smallest possible scale, e.g.
one patient, one nurse, one doctor
 run just as many PDSA cycles as necessary to gain
confidence in your change – then expand
 expand incrementally to more patients
 expand to involve more nurses, more doctors, more
departments
 balance changes within system to ensure other
processes not adversely stressed
What do we want to achieve?
What changes will drive our progress?
How will we measure our progress?
How should we modify our latest changes?
modified from: The Foundation of Improvement by Thomas W. Nolan et. al
Engineering Change
What do we want to achieve?
Set an outcome aim.
(It should be ambitious, must be measurable and must
specify a time-period and a definite population in your
hospital.)
List the outcome aim again, then:
–
–
–
–
ask “why” three times,
ask “how” three times,
look at the new aim statements, and
pick the best one
“Function
Expansion”
modified from: The Foundation of Improvement by Thomas W. Nolan et. al
Engineering Change
What changes will drive our progress ?
Select change(s) to your system, the one(s) most likely to
improve outcomes.
(Recognize that not all changes improve outcomes or offer
balance.)
modified from: The Foundation of Improvement by Thomas W. Nolan et. al
Engineering Change
Principles of Measurement:
Seek usefulness, not perfection.
Integrate measurement into the daily routine.
Use qualitative and quantitative data.
Use sampling.
Plot data over time.
How will we measure our progress?
Define what you will measure quantitatively.
(Collect data, chart measures regularly over specified
time-period, and chart against benchmarks & goal lines.)
Three Types of Measures:
1) Outcomes
2) Process
3) Balancing measures
(Use a balanced set of measures for all
improvement efforts.)
modified from: The Foundation of Improvement by Thomas W. Nolan et. al
Engineering Change
How should we modify our latest changes?
Test your changes.
(Run PDSA cycles to learn from the work setting.)
modified from: The Foundation of Improvement by Thomas W. Nolan et. al
Engineering Change:
Hints for Success
•
•
•
•
•
•
•
•
•
•
•
Empower nursing
Expedite order set and protocol passage through appropriate medical staff committees
Better to implement an imperfect, compromise change than no change at all
Pilot newest changes on smallest scale
Provide hot line or support for difficult implementation situations
Use your new system as a shared baseline, with clinicians free to vary based on individual patient
needs
Follow metrics continuously as you implement
Feed metrics back into subsequent PDSA cycles
Measure, learn, and over time eliminate variation arising from professionals; retain variation arising
from patients
Keep big picture in mind
Negotiate ‘speed bumps’





Time delays in getting data
Incomplete buy-in
Go around obstacles instead of through them (can always go back to them later)
Some who disagree with you may be correct
Make changes painless as possible: make it easy to do the right thing
QI Theory:
Quality Improvement in the Hospital
• Suggested next steps:
1) Share this primer in QI Theory with other hospitalists in
your group
2) Identify an important QI project at your hospital
3) Lead the QI project using all available resources
4) Learn from your experience and be among the first to
mentor other hospitalists
Use SHM’s topic-specific resource rooms to ask questions,
share experiences & tools, review the literature, and to
download presentations to help you educate others.
Acknowledgments
• Brent James, MD, MStat (Intermountain Health Care's Institute for
Health Care Delivery Research): concepts, content, figures
• Thomas Nolan, PhD (Institute for Healthcare Improvement): concepts,
content, figures
• Greg Maynard, MD, MSc (University of California, San Diego):
editorial composition and review
• Jason Stein, MD (Emory University School of Medicine): editorial
composition