Quality Improvement and Patient
Safety:
Part 1: What is a quality chasm and
why do we care???
Part II: Making Mistakes; the why’s
and what if’s
Part III-IV:You and I Can Do QI
Danielle Scheurer, MD, MSCR
Chief Quality Officer
Learning objectives:
Understand what quality and QI are
 Understand why and how QI evolved
 Understand why QI is important
 MUSC and DOM QI structure

What is quality?
Meeting the needs and exceeding the
expectations of those we serve
 Delivering all and only the care that the
patient and family needs

What is improvement?
Reducing variation in care processes
 Shifting and tightening of the bell curve
 A change in the design of care
 Ultimate goal is outcome improvement

Quality Improvement
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:
Patient Population
worse
Quality
Tail
better
worse
worse
Quality
What put quality on the “front
burner”
Libby Zion: 1984
 Lewis Blackman: 2000
 Josie King: 2001
 Young
 Healthy
 Smart parents
 97,999 low profile cases

What put quality on the “front
burner”

To Err Is Human
◦ 1999 IOM Report
◦ 98,000 deaths in US hospitals a year due to
preventable medical errors
◦ Due to “faulty systems, processes, and
conditions that lead people to make mistakes”

Crossing the quality chasm
◦ 2001 IOM report
◦ Defined 6 dimensions of healthcare quality
◦ Suggestions for improvement
Institute of Medicine

Annual deaths in the US
from:
Medical errors: 44,000 –
98,000
 Motor vehicle accidents:
43,458
 Breast cancer:
42,297
 AIDS:
16,526

Institute of Medicine
The American health system needs fundamental change
“The nation’s health care system lacks… the
capabilities to ensure that services are safe,
effective, patient-centered, timely, efficient and
equitable…Between the health care we have and
the care we could have lies not just a gap but a
chasm.”
Dimensions of Healthcare Quality
Safe: Avoid patient injuries from the care
that is intended to help them
 Effective: Match care to science; avoid
overuse of ineffective care and underuse of
effective care
 Patient-centered: Honor the individual
and respect choice
 Timely: Reduce waiting for both patients
and providers
 Efficient: Reduce waste
 Equitable: Close racial and ethnic gaps

Quality Improvement:
Bridges the Implementation Gap
Scientific
understanding
Progress
Implementation
Gap
Patient care
Time
Why is it important?
Making mistakes: It happens





Transposition error rate (mixing
drug doses)
Commission error rate
(misreading a label)
Calculation error rate
Omission error rate (failure to
turn off a switch)
Error rates under stress or
danger
Most serious medical errors are committed by
competent caring people doing what other competent
caring people would do…..
Don Berwick

.0006

.003
.03
 .01


.25
Why is QI important?
•
•
Patients: Better care, fewer errors
Regulation and accreditation
• Reputation
• Financial
•
•
•
Life long learning (ACGME and MOC)
Academic activity
Front-line clinicians
– Empowerment to change
– Jobs more rewarding
– Less daily frustration
Quality and the ACGME

System based practice: Residents must
demonstrate and awareness of and
responsiveness to the larger context and
system of health care, as well as the ability
to call effectively on other resources in
the system to provide optimal health care
System based practice: Expectations






Work effectively in various health care delivery
settings and systems relevant to their clinical specialty
Coordinate patient care within the health care
system relevant to their clinical specialty
Incorporate considerations of cost awareness and
risk-benefit analysis in patient care and/or populationbased care as appropriate
Advocate for quality patient care and optimal patient
care systems
Work in interprofessional teams to enhance patient
safety and improve patient care quality
Participate in identifying system errors and in
implementing potential systems solutions
Quality and the ACGME

Practice based learning and improvement:
Residents must demonstrate the ability to
investigate and evaluate their care of
patients, to appraise and assimilate
scientific evidence, and to continuously
improve patient care based on constant
self-evaluation and life-long learning.
Residents are expected to develop skills
and habits to be able to meet the
following goals:
Practice based learning and
improvement








Identify strengths, deficiencies, and limits in one’s knowledge
and expertise
Let learning and improvement goals
Identify and perform appropriate learning activities
Systematically analyze practice, using quality improvement
methods, and implement changes within the goal of practice
improvement
Incorporate formative evaluation feedback into daily practice
Locate, appraise, and assimilate evidence from scientific
studies related to their patients’ health problems
Use information technology to optimize learning
Participate in the education of patients, families, students,
residents, and other health professionals
Quality in Life Long Learning
American Board of Internal Medicine
 “Maintenance of Certification” requires
evidence of participation in quality /
process improvement
 Subspecialty certification requires QI

Career
No longer “soft science”
 Peer reviewed publications
 Considered “translational research”
 Subject to IRB for dissemination

QI and ACGME
Medicine
Peds, Optho, Neuro, Derm, Radiology, Anesthesia
……
Resources
Curriculum
 Early project development
 Data resources
 Statistical help

PI Project at MUSC
Quality Improvement at MUSC
IMPROVE structure for QI
 Lean and Six Sigma management
philosophies, along with strong leadership
and willingness to innovate, to improve
quality

Quality Philosophy: Lean
Management philosophy derived from
Toyota Production System which focused
on the elimination of waste
 Waste is anything that adds to a process
without adding value “non-value added”

What is “Waste”: 8 types








Waiting: Patients, doctors, staff
Overproduction: 3 admits in 20 minutes to
the same person
Rework: Medication reconciliation
Motion: Looking for charts
Processing: Several people doing the same
steps to discharge a patient
Intellect: Doing things “below your pay
grade”
Inventory: Wasted “stuff” laying around
Transportation: Patients, materials
Quality Philosophy: Six Sigma

Management philosophy that seeks to
reduce variation and defects by applying a
highly structured, systematic approach to
operational success
◦ Asthmatics should all get beta-agonists
◦ Sickle cells admits should all get IVF

Initially developed by Motorola
IMPROVE
•
•
•
•
•
•
•
Identify the problem
Measure the impact
Problem analysis
Remedy the critical issues
Operationalize the solutions
Validate the improvement
Evaluate over time
Identify the problem
What is the problem?
 Who identified it?
 When was it identified?
 When and where is it occurring?
 Pick something that matters to you, and
state WHY it matters (how does it fit into
the domains of quality)
 Who else cares about the problem (who
are the stakeholders?)

Measure the impact
How is the data collected (survey,
administrative data, chart review)?
 Is it valid/accurate?
 Is it a manual process or automated?
 Is there a clear definition of the outcome
(or can it be interpreted different ways)?
 Who is going to measure?
 Can you sample?

Problem Analysis: What is causing
the problem?
•
•
•
•
•
•
•
•
Time of day, day of week
Department specific / system wide
Inefficient staffing (numbers or skill set)
Poor communication
Inadequate process or policy
Lack of controls to keep the problem from
occurring
Poor individual performance (usually not the
only issue)
Pick an appropriate process analysis tool to
further analyze the problem/process
Remedy the critical issues






Pick a remedy based on the process analysis
What will the group try to improve?
What are the barriers?
What evidence is there that it will have an
impact (has someone tried and succeeded or
failed)?
How “reliable” is the intervention?
Do you need more than one intervention to
make it nearly impossible to recur?
Remedies (in order of reliability)
Education
 Reminders
 Checklists
 Order sets
 Protocols
 Pathways
 Templates
 “Hard stop” order entry

Operationalize
How are you going to make it work?
 How will the barriers be removed?
 What assistance is required from senior
leadership?
 What is the plan to roll out and
implement solutions?

Validate
How will we know we made a difference?
 What are you measuring?
 How often are you measuring it?
 Is the measure meaningful?
 Are you measuring “unintended
consequences”?

Evaluate
How to sustain the improvement?
 Who is responsible for monitoring and
measuring over time?
 What is the plan to react if the measures
slip?
 How will future staff be made aware of
the new process?

What is a Safety Culture?
Not focused individual blame
 Safe system to disclose (potential) issues
without repercussion
 System approach to proactive error
reduction
 Workers are involved in the work
process and process re-design
 Workers are empowered to make change

How to create a safety culture
No blame problem analyses
 Anonymous safety reporting (eg PSN)
 Providing feedback to front line staff of
events and reporting
 Frontline provider engagement in
changing the system

Making mistakes
The why’s and what if’s
Danielle Scheurer, MD, MSCR
Chief Quality Officer
Learning objectives:
Understand basic definitions in patient
safety (errors, adverse events, negligence)
 Understand fundamentals of “safety
culture” and “system approach” to safety
 Understand most common types of
healthcare errors and how to mitigate
 Understand the role of disclosure,
apology, and renumeration in safety
events

What is an Error?
Failure of a planned
action to be
completed as
intended
 OR the use of a
wrong plan to
achieve an aim

- To Err is Human, Institute of Medicine 2000
Types of errors: Latent
System defects that make active errors
more likely to happen
 Ex: Poor design of IT data display making
it easy for clinicians to misunderstand

◦ Nursing understanding of heparin protocols
Inadequate staffing making people “rush”
or routinely “multi-task”
 Difficult to measure; may exist for years
before effect is seen

Types of errors: Active
Occur at the frontline provider level
 Usually easier to measure; negative
outcome almost immediately apparent

◦ Nurse hangs the wrong medication
◦ Doctor orders medication intended for
another patient
◦ PICC line inserted into the wrong patient
◦ Doctor amputates the wrong leg
Active versus latent errors

A “safety culture” system:
Proactively looks for latent errors all the
time
 Retroactively looks for latent errors
when an active error has occurred

What is an Adverse Event?
An injury that is caused by medical
management rather than the patient’s
underlying condition
 Preventable adverse event = adverse
event due to error

Negligence
Errors
Near
Misses
Preventable
Adverse Events
Adverse
Events
The points of intersection
Not all errors result in an adverse event
(“near miss”) Ex: double dose lasix with
no effect to the patient
 Not all adverse events are the result of an
error Ex: hyperglycemia from steroids
 Of errors/adverse events that are
preventable, only a small portion are due
to “negligent” active error

Patient Safety
 Freedom
from
accidental injury
 Involves the
establishment of
systems to
decrease all
errors, and
intercepting them
when they occur
What is a Safety Culture?
Not focused individual blame
 Safe system to disclose (potential) errors
without repercussion
 System approach to proactive error
reduction
 Workers are involved in the work
process

How to create a safety culture
No blame M&M conferences or RCAs
 Simulating possible adverse events (eg
FMEA)
 Anonymous safety reporting system (eg
PSN)
 Providing feedback to front line staff of
events and reporting

System Approach to Error:
Fundamentals
Analyze the system for contributing
factors that allow errors to occur
(majority of errors due to faulty systems)
 Individual blaming will not prevent future
errors from occurring
 Individuals for the most part are trying to
do their best; they do not come to work
with the intent to harm

Making mistakes: It happens
Transposition error rate (mixing drug
doses)
 Commission error rate (misreading a
label)
 Calculation error rate
 Omission error rate (forget)


Error rates under stress or danger

.0006

.003
.03
 .01

10X
A system approach

“We cannot
change the
human
condition, but
we can change
the conditions
under which
humans work”
Reason; BMJ 2000;320:768-770
Human Factors Engineering

“The study of the
interrelationships
between humans,
the tools they use
and the
environment in
which they live
and work”
Factors that contribute to error

Environment: Distractions, interruptions,
noise, ergonomics, clutter, lighting
◦ Forgot to enter order due to multiple
pages/competing priorities
◦ Lost handwritten sign out sheet and forgot to
check a lab
◦ Discharge paperwork incomplete due to
inability to find all needed papers
Factors that contribute to error

Human: Stress, fatigue, emotions
(boredom, frustration) communications
◦ Feel asleep and did not check a lab
◦ Failure to clarify sign out questions due to
discomfort with confrontation
◦ Transcription error in filling out discharge
med list due to stress or fatigue
Factors that contribute to error

Patient: language, compliance, fear, disease
acuity, knowledge deficit
◦ Patient neglects to mention a pertinent
review of symptoms due to fear or
embarrassment of what it would mean
◦ Patient neglects to take meds correctly due to
finances, lack of understanding, low literacy
◦ Patient does not understand instructions due
to language barrier
Factors that contribute to error

Equipment: design, variability, quality,
labeling, training, availability
◦ Patient gets wrong heparin dose due to poor
nurse training on how to program smart
pump, multiple heparin doses in drug library,
or no availability of a pump
◦ Patient gets wrong chemotherapy dose due to
confusing labeling during packaging
◦ Hand hygiene does not occur due to
unavailability of purell/soap
Factors that contribute to error

Systems: staff level, staff experience,
supervision, handoffs, workflow
◦ CLA-BSI occurs due to low staffing and single
provider putting in line without checklist or
help with sterile procedure
◦ Med rec not completed because it is out of
workflow (does not connect to CPOE or
inpatient documentation)
◦ Patient not seen upon transfer out of ICU
due to inadequate communication from ICU
to floor team
Layers of defense to reduce errors
Most common types of errors in
healthcare
Medication errors
 Procedural errors
 Communication errors
 Diagnostic errors

Medication Errors

ADE:
◦ Increases LOS by 8-12 days
◦ Increases cost $16,000 to $24,000
◦ Most commonly due to wrong dose (not time,
route, drug, or allergy errors)

pADE:
◦ Most commonly occurs at physician ordering
stage (not administration or dispensing)

Near miss: Caught before given
Mitigating Medication Errors

Order stage:
◦ CPOE with decision aids for
dose, route, interval,
indication

Prep stage:
◦ Double check with pharmacy
for drug/dose
appropriateness
◦ Remove open floor stocks of
high risk medications
(heparin, insulin)
Mitigating Medication Errors
Administration stage:
 Bar coding to match
patient with medication
dose, time, and route
 Teach back technique
with the patient/family
during medication
administration
Procedural errors
Wrong patient
 Wrong site
 Wrong/Bad
equipment

Mitigating procedural errors
Procedural “timeout” with checklists
 Team training
 Culture of safety
with frontline
empowerment
 Teach back technique
with the patient for
procedural marking
(prior to sedation)

Mitigating procedural errors
Procedural simulation training
 Consistent equipment stocking

Communication errors
Illegible handwriting
 Inadequate handoffs
 Poor or no
communication
among or between
teams
 Poor, no, or
inappropriate
communication to
the patients

Mitigating communication errors
Template EMR
documentation
 Multi-disciplinary
team training and
rounding
 Standardize template,
time, and place for
handoffs (without
noise, distraction)

Mitigating communication errors
Use read back
techniques for
critical information
 Use SBAR technique
(or other) to
communicate new
patient findings

Diagnostic errors
Delayed or missed diagnosis which affects
patient outcome or leads to additional
complications / procedures
 Account for about 1/3 of errors
 Associated with proportionally higher
morbidity than other errors
 Can be due to heuristics, most commonly

◦ Pattern recognition
◦ Algorithmic thinking
Pattern recognition:
What do you see?
Pattern Recognition:
A heuristic (shortcut)
You see it, you know it
 Relies on experience
 Requires a well defined prototype
 Is quick and “easy” and used commonly
 Prone to representative error and
premature closure

Algorithmic Thinking
Hyponatremia
Hypovolemic
Normovolemic
Hypervolemic
SIADH
Solute loss (eg vomiting)
Cirrhosis, CHF, nephrotic syndrome
Avoiding diagnostic error

Slow down
◦ Take a diagnostic time
out
◦ Always make a
differential diagnosis
◦ Ask what does not fit,
and could it be more
than one diagnosis
Avoiding diagnostic error
Always consider the
worst it could be
 Focus on the history
and physical exam

◦ Listen to the patient
◦ Reduce reliance on
testing
Avoiding diagnostic errors

Understand Bayesian
theory
◦ Use a mathematical
approach to decision
making
◦ Use pre and post-test
probability to know
the value of diagnostic
testing
Avoiding diagnostic errors

Acknowledge how
the patient makes
you feel
◦ Emotions (good and
bad) can affect how
you test and treat the
patient
Medical negligence


Failure to meet the
standard of practice of
an average qualified
physician practicing the
specialty in question
Physician practicing in a
system not grossly
“latent error prone”
that breaches standard
practice
Disclosure: part of patient safety

Errors and near
misses have to be
transparent
◦ To providers
◦ To patients
◦ To the system

Disclosure with
compensation
reduces claims,
lawsuits, and legal
costs
Kachalia A. Ann Intern Med 2010;153:213.
What physician think
What Patients Want
Jargon-free statement
that error occurred
 Description of the error
 Why it happened
 Implications of the error
for their health and how
to deal with the
consequences

Gallagher TH, et al. JAMA. 2003:289:1001-7.
83
What Patients Want
Outline of steps that
will be taken to
prevent future errors
 An apology from the
health care worker

Gallagher TH, et al. JAMA. 2003:289:1001-7.
84
Why an apology?
Barriers to Disclosure for Physicians

Fear of litigation
◦ Unlikely to apologize due to concern about
consequences of admitting fault

Discomfort with discussing such issues
◦ Physician may choose words carefully to avoid
explicitly stating that an error occurred

Concern that information may harm
patient
◦ Belief that disclosure may impact patient’s
trust in the physician
Gallagher TH, et al. JAMA. 2003:289:1001-7; Robinson AR, et al. Arch Intern
Med. 2002:162:2186-90; Wu AW, et al. JAMA. 1991;265:2089-94.
86
Errors happen: 3 part plan
RECOGNIZE the mistake
 MITIGATE harm associated with error
 DISCLOSE the error

MUSC approach to apology: DO
Take care of the patient
 Call Risk management
 Call your attending
 Be open and honest

Disclosure process:
If there is a change in plan of care:
Attending must disclose
 If there is not a change in plan of
care: Attending must be informed,
but resident can disclose
 Always call risk management first

How to disclose
DO






Apologize
Explain outcome,
current condition,
treatments
Provide support and
empathy
Offer family meeting
Be accessible
Give contact info
DO NOT
Use terms negligence
or accident
 Blame or discuss
fault
 Admit or speculate
about liability
 Discuss peer review
or root cause
analysis

How to document
DO
Objective clinical
information
 Medical response
 Follow up plans
 If late entry, mark
“late entry” with
date of time of event
and entry

DO NOT
Alter prior
documentation
 Mention risk
management,
occurrence
reporting, sentinel
event

Steps for Risk
Patient follow up plan after analysis of
what happened
 Financial plan
 Attending follow up

Quality Improvement: Part 3-4
You and I Can do QI
Project Design and Publication
Danielle Scheurer, MD, MSCR, SFHM
Medical Director of Quality
Medical University of South Carolina
Learning objectives:
Understand the fundamentals of forming a
team and team meetings
 Give an example of a goal and an aim
 Give examples of types of measurement
 Identify some ways to evaluate a process
 Understand benefits of PDSA

Team Members
Champion: Respected, able to surmount
barriers, external voice of the project
(mentor, attending)
Team Members
Team Facilitator: QI expert, helps with
design and operations (optional but nice)
 Process owner: Schedules meetings, sets
the agenda, records activity, reports
findings. “Owns” the success of the
project.

Team Meetings
All members and opinions are equal
• Team members will speak freely and in turn,
no side bar conversations, one at a time
• Cells phones and pagers on silent alert
• Discuss/analyze/attack problems, not people
• Everyone gets their homework done
• Silence (or absence) equals agreement
• Meetings start / end on time
• Get creative on meeting times, use
conference calling
•
IMPROVE
•
•
•
•
•
•
•
Identify the problem
Measure the problem
Problem analysis
Remedy the critical issues
Operationalize the solutions
Validate the improvement
Evaluate over time
Identify the problem
What is the problem?
 Who identified it?
 When was it identified?
 When and where is it occurring?
 Pick something that matters to you, and
state WHY it matters (how does it fit into
the domains of quality)
 Who else cares about the problem (who
are the stakeholders?)

Measure

How to get data
◦ (survey, administrative data, chart review)?
Is it valid/accurate?
 Is it a manual process or automated?
 Is there a clear definition of the outcome
(or can it be interpreted different ways)?
 Who is going to measure?
 Can you sample?

Problem Analysis: What is causing
the problem?
•
•
•
•
•
•
•
•
Time of day, day of week
Department specific / system wide
Inefficient staffing (numbers or skill set)
Poor communication
Inadequate process or policy
Lack of controls to keep the problem from
occurring
Poor individual performance (usually not the
only issue)
Pick an appropriate process analysis tool to
further analyze the problem/process
Set Goal
•
Goals are general statements of what you
are trying to achieve
– Reduce CA-UTI
Set Aims
•
Aims are SMART
– Specific: Reduce CA-UTI on 8E
– Measurable: Current 8/1000 catheter days
– Achievable: Reduce to 5/1000 catheter days
– Realistic: Depends on HOW to achieve it
– Time delimited: In 6 months
Examples of Aims

> 75% of CHF patients on 8E will be
discharged with instructions within 6 mo

Decrease all cause mortality rates in
COPD from 10% to 5% by Jan 2012

Reduce unplanned extubations by 50%
within 3 months of intervention
106
Types of Measurement

Process measures: The process of delivering
care
◦ HbA1c for diabetics
◦ Completion of an order set for asthmatics

Outcome measures: The actual end point
◦ Death
◦ Readmission rates

Balancing measures: What are the other
(sometimes adverse or unintended)
consequences of our intervention
◦ Time to complete work
◦ Longer length of stay
Measurement: How will we know
that a change is an improvement?
Pick measures that are related to aims
 Develop operational definitions
 Determine patient eligibility, denominator,
numerator
 Keep measures simple and practical

108
Examples of Measures

Mean time from initiation of discharge process
to patient walking into elevator (in minutes)
◦ [Process measure]

% PNA patients readmitted with 30 days
◦ [Outcome measure]

PNA length of stay
◦ [balancing measure]
109
Principles of measurement
Seek usefulness, not perfection (your goal is
improvement, not measurement)
 Integrate measurement into the daily routine (instead of
waiting 2 months for administrative data)
 Use qualitative and quantitative data (ask nurses how it
is going?)
 Use sampling (make your measurement easy, not hard)
 Plot data over time (to observe trends and patterns)
 Use balancing measures (measure anticipated
unintended consequences, such as time or cost)
 Measure both process and outcome measures (% use of
a discharge order set, and % discharged on intended
medication)

Measuring processes:
◦
◦
◦
◦
◦
◦

Cause and effect diagrams
Process maps
Swim Lane Maps
Decision flow diagrams
Spaghetti diagrams
Cycle time charts
Basic 2 types of data analysis tools
◦ Pareto chart
◦ Run chart
Process evaluation:
Processes are complicated, involving multiple
steps and decision branch points
 If a process is flawed, inefficient, or resulting
in an undesirable outcome, the process
should be systematically evaluated
 How to evaluate a process:

◦
◦
◦
◦
◦
◦
Cause and effect diagrams
Process map
Swim Lane map
Decision Flow Diagram
Spaghetti diagrams
Cycle time charts
Cause and Effect Diagram
Fishbone “Ishikawa” diagram
Graphically displays possible factors
leading to an event or process
 Used to organize ideas during
brainstorming sessions
 Helps to logically and systematically
evaluate a process, and decide which
steps are flawed, inefficient, or need repair

Cause-and-Effect Diagram
Example: Adverse Drug Events (ADE)
Drug
Administration
Errors
Ordering
Errors
Nurse
Physician
Pharmacist
Transcribing
Rate
Dilution
Spelling
Route
Nurse
Dose
Route
Scheduling
Psychiatric
Electrolyte
Hepatic
Renal
Past Allergic
Reaction
Absorption
Physiologic
Factors
Patient
ADE
Unforeseen
Weight
Gender
Race
Patient
Errors
Order Missed
Wrong
Drug
Age
Compliance
Place outcome here
Dosage
Time
Cognitive
Physician
Pharmacy
Nurse/Clerk
Expected
Drug/Drug
Drug/Food
Drug/Lab
Pharmacokinetics
Pharmacodyamics
Pharmocologic
Factors
Pharmacist
Patient
Physician
Dietician
“Macro” Process Map
Displays the number of steps needed to
successfully implement a process
 Graphically displays the “current state” at
a “high level”
 Each step is displayed, which can later be
broken down into smaller steps “micro
level”
 The process, and each point, can be
further investigated based on where the
failures are occurring

Macro Process Map
Patient needs
a procedure
INR checked
and elevated
FFP ordered
and given
Swim Lane Map
Shows simultaneous workflows of
different disciplines
 Can see the interaction of disciplines
 Helps to facilitate better communication
among disciplines in complicated
processes

Decision Flow Diagram
Displays the steps of a process with
decision branch points
 Creates transparency of a process
regardless of complexity
 Can identify areas of improvement

◦ How the process is supposed to work
◦ How the process is actually working
◦ Variability in the process

Circle (input/output), diamond (decision
branch), square (task to do)
Decision Flow Diagram

Lamp not
working
Stop
Y
N
Plugged
in?
Plug it in
Y
Wor
king?
N
Bulb
burned
out?
N
Buy new lamp
Y
Put in new bulb
Wor
king?
Y
Stop
Spaghetti diagram
Graphically displays motion (and usually
waste) of a process (or set of processes)
 Can be effective for using before-after an
intervention, although less “rigorous” an
outcome measure
 Can be used in conjunction with time
motion studies, pedometer readings, etc

Outpatient clinic
Cycle times
Graphic depiction of what processes are
occurring, in what order, and how long
each “batch” is taking
 Anytime the “batch” times are uneven,
there is waste built in the process
 Efficient processes have “load balancing”
 Can be used to identify waste within a
process, then used to design a solution,
which can be re-measured

ED process: cycle times
Time (minutes)
30
25
20
15
Time (minutes)
10
5
0
Registration
Triage
Nurse
Doctor
Get more granular; see if you can
load balance work
30
25
20
Non-doctor work
Time (minutes)
15
10
5
0
Registration
Triage
Nurse
Doctor
Once you have collected data:
Data Analysis

Data can be displayed (and analyzed) in a
number of ways, the most common (and
simple) of which are:
◦ Pareto Chart (Bar graph)
◦ Run Chart
Pareto Chart (Bar graph)
Graphically displays relative frequency of
events
 Used to focus choices, by designating the
weight of each category
 Can help teams prioritize which
processes to focus on first
 Based on this graphic depiction of the
causes of an adverse event, might focus
on the top 3 contributing causes first

Pareto Chart
100
Causes Contributing to Adverse Drug Events
90
80
% Contributing
70
60
50
40
30
20
10
0
Causes
Causes
Run Charts
Graphically displays information over time
 The horizontal axis usually has a mean
performance
 The vertical axis has interventions placed
to determine if the interventions are
affecting the outcome

Run Chart
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
Ap
r-0
4
Fe
b04
-0
3
De
c
ct
-0
3
O
Au
g03
Ju
n03
Ap
r-0
3
Fe
b03
-0
2
De
c
ct
-0
2
0
O
Percent
60
Now it is time to implement the
change
Start on a small scale (a clinic, a unit, a
doctor)
 Improve incrementally
 Learn through action
 Test your changes
 Assess their effect
 Re-work the changes and do it again and
again and again…
 PDSA cycles: PLAN, DO, STUDY, ACT

Value of PDSA cycles
•
•
•
•
•
Allows “failures” without impacting large
scale performance (“fail fast”)
Adapts to changing environments
Evaluates costs and unintended
consequences of the change
Empowers front line users to experience
the change on a small scale
Minimizes resistance for larger
implementation
Bottom line with PDSA
• Common sense approach to change and
improvement

It is not complicated

Manageable
◦ big ideas tested on a small scale

Cycles happen quickly
◦ allows for successive cycles and sustainable
change

Guides the test of a change to determine if the
change is an improvement
Changes that
result in
Improvement
A P
S D
A P
S D
A P
S D
A P
S D
Big Idea
Time
Source: D. Berwick
IMPROVE: Next week
•
•
•
•
•
•
•
Identify the problem
Measure the impact
Problem analysis
Remedy the critical issues
Operationalize the solutions
Validate the improvement
Evaluate over time
Remedy the critical issues






Pick a remedy based on the process analysis
What will the group try to improve?
What are the barriers?
What evidence is there that it will have an
impact (has someone tried and succeeded or
failed)?
How “reliable” is the intervention?
Do you need more than one intervention to
make it nearly impossible to recur?
Operationalize
How are you going to make it work?
 How will the barriers be removed?
 What assistance is required from senior
leadership?
 What is the plan to roll out and
implement solutions?

Validate
How will we know we made a difference?
 What are you measuring?
 How often are you measuring it?
 Is the measure meaningful?
 Are you measuring “unintended
consequences”?

Evaluate
How to sustain the improvement?
 Who will monitor / measure over time?
 What is the plan if the measures slip?
 How will future staff be made aware of
the new process?

Remedies
At least one needs to be “highly reliable”
 Usually need several to be effective and
sustainable

◦
◦
◦
◦
◦
◦
Admission / Discharge orders
Order entry “hard stops”
Decision aids / Protocols
Educational material
Pre-populated documentation in EMR
Standing orders / Default pathways
Remedies (in order of reliability)
Education
 Reminders
 Checklists
 Order sets
 Protocols
 Pathways
 Templates
 “Hard stop” order entry

What is “Reliability”
Failure free operation over time
 You get the results you intended to get
 Inverse of the system’s failure rate

Making Systems Safer:
Way to Redesign Systems
 Elimination
 Confirmation
 Standardization
 Simplification
 Constraints
 Reminders
and
forcing functions
 Redundancy
 Differentiation
 Detection
 Mitigation
 User-centered
design
144
Organizational Reliability
Aviation and nuclear power: Failure rate is
1 in 1,000,000 (“Six Sigma”)
 Medical processes: we routinely “fail”

◦ Hand hygiene: 8 in 10
◦ Patient understanding of medication: 1 in 3
◦ VTE prophylaxis in inpatients: 9 in 10
What is “High Reliability”
•
High reliability system
• Known to be complex and dangerous, but safe
with infrequent error rates
• Has a preoccupation with failure
• Defers decisions to expertise (not seniority)
• Routinely (and proactively) evaluates what could
go wrong, instead of what did go wrong
• Seeks to standardize processes; doing the “right”
thing is easier than doing the “wrong” thing
• DOES NOT rely on vigilance or memory alone
Not Highly Reliable
Highly Reliable
Why do we care?

Ways to think about designing
interventions that will be
◦ EFFECTIVE
◦ SUSTAINABLE
Elimination
Eliminate some choices on purpose
 Reduce # procedural trays available

◦ Combine with simulation to ensure
understanding and competency

Reduce # medications available
◦ Combine with CPOE and formulary aids
150
Elimination
Remove high risk
medications from floor
stock (concentrated KCl,
insulin vials)
 Do not allow verbal orders
 Remove razors from all
procedural areas

151
Standardization

Prescription writing
◦ 0.1 mg but not 1.0 mg
◦ Approved abbreviations only (no “QD”)
◦ Spell out units: “micrograms”, not “µg”

Metric units only (kg, cm, Celsius)
◦ Standardize all thermometers, bed weight,
order sets, patient instructions
152
Standardization

Standardize equipment
◦ Pumps
◦ Ventilators
◦ Keyboard

Standardize documentation
◦ Flow sheets
◦ Progress notes
◦ Discharge summaries
153
5S Standardization

Workplace organization
methodology
Sort: Keep essential items,
discard others
Straighten: Everything has it’s
place
Sweep/Shine:Visual cue for
mess
Standardize: All units identical
Sustain: Avoid entropy

Creates safer environments





Constraints / Forcing Functions
CPOE
 Orders with pre-printed boxes checked
 Metric system for orders
 Scan armband to allow medication release
 Not allowing blood release from blood
bank until 2 blood types verified

156
Redundancy
2 nurse verification
for chemotherapy
 Pharmacist verifies
admission
medications from
physician medication
history
 Pharmacist “signs
off” on all new
medication orders

157
Redundancy
2 patient identifiers
before blood drawn
 Critical lab read back
technique
 Patient weight on
every order page
 Team contacted
twice for ICU
transfers

158
Differentiation
Confirmation
Allowing for question time during sign out
 Read-back of verbal orders
 “I’m clear, you’re clear, we are all clear”
during codes
 Procedural time out

160
Simplification
1-2 medications from
each category
available for ordering
(ACE, BB)
 Reduce number of
places information is
stored (vitals signs,
allergies)
 Universal influenza
vaccination for all
inpatients

Simplification
Make things
predictable and easy
to find (oxygen
tubing, a central line
kit, LP tray)
 Have central line kits
with only necessary
equipment included
 Reduce med rec
from a 2 to a 1 step
process

Reminders:
Avoid relying on memory

Checklists: first developed in mid 1930’s
by the Army Air Corps in response to
increasing complexity of flying an airplane.
◦ Resulted in the adoption of the B-17, which
helped to turn the course of WWII

Adopted as a way to reduce central line
associated infections,VAP, wrong site
surgery, etc, etc, etc
163
Detection
Create a system so
that errors are
detectable in real
time
 Bar coding / scanning
meds, makes it easy
to recognize you are
about to give the
wrong med to the
wrong patient

Detection
Hard stop in CPOE
for allergy and drug
interactions
 Sponge count at the
end of surgeries
 Xrays at the end of
all trauma cases

Mitigation
When something goes wrong, make the
impact as innocuous as possible
 Standard reporting for events
 Immediate RCA to reduce harm
associated with error and recurrence
 Give medication antidote if available

User-centered designs

Creating a work space that enables us
◦ To make safe and sound decisions
◦ To implement those decisions into a seamless
care plan
Not user centered
QI/PS Resources
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
IHI (http://www.ihi.org/IHI/Topics/PatientSafety)
AHRQ (http://www.ahrq.gov/qual)
AHRQ 2009 quality report (http://www.ahrq.gov/qual/nhqr09/nhqr09.pdf)
JC (http://www.jointcommission.org/Library)
NPSF (http://www.npsf.org/rc/mp)
CDC (http://www.cdc.gov/ncidod/dhqp/index.html)
IOM (http://www.iom.edu)
SHM (http://www.hospitalmedicine.org)
HRSA (http://www.hrsa.gov/quality)
SQUIRE guidelines/resources (http://squire-statement.org/)
AHA HRET (http://www.hret.org/hret/publications/qualitypubs.html)
American Society of Quality (http://www.asq.org/learn-about-quality/cause-analysistools/overview/overview.html)
QI/PS Journals
International Journal for Quality in Health Care
International Journal of health care quality assurance
Joint Commission journal on quality and patient safety
Journal of health care quality
Enhanced Reading
Davidoff F, Batalden P, Stevens D, Ogrinc
G, Mooney S for the SQUIRE
development group. Publication guidelines
for quality improvement studies in health
care: evolution of the SQUIRE project.
JGIM 2008;23:2125.
 Nolan and Berwick. All-or-none
measurement raises the bar on
performance. JAMA 2006;295:1168.

Enhanced Reading
www.ihi.org/IHI/Topics/Improvement/Impr
ovementMethods/HowToImprove
 http://iom.edu/~/media/Files/Report%20Fil
es/1999/To-Err-isHuman/To%20Err%20is%20Human%2019
99%20%20report%20brief.pdf
 IHI open school

◦ QI 101: Fundamentals of improvement
◦ PS 101: Introduction to patient safety
Writing an abstract for a
scientific meeting:
Learning objectives
Recognize the differences between
abstracts for manuscripts and for
meetings
 Compare Traditional vs QI abstracts
 Understand key steps in writing a good
abstract
 Identify specific strengths & common
problems in abstracts submitted for
presentation at scientific meetings.

Outline
Why write a great abstract?
 Review the anatomy of an abstract
 Review common errors in abstract
writing
 Small groups to critique previously
presented abstracts
 Large groups to discuss findings and
address specific comments

Why Write “Abstracts for Scientific
Meetings”
Chaotic ideas find form and substance
through drafting
 Promotes original thought
 Sets goals for project completion

The Project Life Cycle
Project
Idea
Obtain Funding
and Career
Advancement
Preliminary
Data
Advance Science
Abstract for
Scientific
Meeting
PeerReviewed
Publication
Purpose of an Abstract for a
Scientific Meeting
Flash compared to detail
 Limit data presented but also present
most important data

◦ What data supports your conclusions
directly?
Guidelines for Abstract Writing
Title
Concise
 Direct
 No abbreviations

Introduction
Why this study?
 Concisely state the goals and rationale
(i.e., describe why the work was done in 1
or 2 sentences)
 10% to 15% of total length

Methods

What was done?
◦ Describe the population studied, techniques
used, and how the data were analyzed
◦ Definitions

30% to 35% of total length
Results

What was found?
◦ Summarize only the important findings

Tables & graphs can be used in abstracts
written for meetings as a means to
emphasize the results
◦ A table or graph is not a substitute for results
but rather a means by which to efficiency
display the results

35% to 40% of total length
Discussion

The Answer!!
◦ Most important things learned
◦ OK to state implications of findings but don’t
extrapolate too much
◦ Keep conclusions within scope of data
investigated

10% to 20% of total length
QI vs. Traditional Manuscripts
Traditional Manuscript
Completely defined process
 Generalizability
Methods section
 Describe design
 Details about all of process
Results section
 Present summary measures
 Adjust for variability
 Tables and Figures
Discuss: Interpret data and
describe impact

QI Manuscript
Evolving process
 Local context
Methods section
 Describe context/personnel
 Detail about initial process
Results section
 Describe process changes
 Show all the data/variability
 “Run” and SPC charts
Discuss: implications of outcome
data and changes in process

Davidoff F, Batalden P, Qual Saf Health Care 2005
The Writing Process: Drafts

Create an outline or free-writing under
headings

A first draft will likely be disappointing
◦ You can’t revise until you’ve written
The Writing Process: Drafts

Obtain feedback from peers
◦ Encourage them to provide an honest and
constructive critique
◦ “I want this to be a good abstract so please don’t
worry about me being offended”

More is better
◦ Solicit feedback from multiple people from diverse
disciplines
◦ Different reviewers focus on different areas & content
(e.g., research design, statistical analysis, or writing
style)
The Writing Process: Drafts
Provide feedback to peers when asked
 Participate in group feedback sessions- you get
to hear a variety of viewpoints
 The more abstracts you review, the easier it is
to write your own

Writing Styles & Hints
No jargon
 Only use standard, well-accepted
abbreviations and symbols
 Shorter is better, always

◦ I have made this letter longer than usual because
I lack the time to make it shorter (Blaise Pascal,
Provincial Letters, XVI. 1657)
Common Problems

Lack of sentence clarity (i.e., the message
is not clear)
◦ Question omitted
◦ Question vaguely stated
◦ Answer is not stated (only implications
stated)
Common Problems

Poor organization
◦ No clear ordering of sections
◦ Mixing content (e.g., giving results in the
methods section)
Common Problems

Lack of coherence
◦ Too many abbreviations
◦ Use of “respectively”
 Cure was achieved in patients with aa, bb, cc, dd,
and ee in 0.5%, 0.5%, 0.6%, 0.7%, and 0.8%,
respectively
 Better to use a table
Common Problems

Excessive detail
◦ Reporting data for minor results
◦ Providing details for well-known techniques
◦ Providing exact data rather than odds ratio or
percent change
◦ Duplicating data
◦ Including references
Common Problems

Failure to consider non-specialist readers
◦ Underdeveloped rationale
◦ Overly technical language
Common Problems

Make sure the entire abstract develops
and concludes the main point YOU want
to express
◦ Abstracts evolve as you write them
 Critically review the introduction statements after
finishing conclusions
The Project Life Cycle
Project
Idea
Obtain Funding
and Career
Advancement
Preliminary
Data
Advance Science
Abstract for
Scientific
Meeting
PeerReviewed
Publication
Publication Rate
Carroll AE, et al. Pediatrics 2003;112:1238-1241
It Takes Time, Be persistent
Carroll AE, et al. Pediatrics 2003;112:1238-1241