Presented to:

1
Organizing for High Reliability in Health Care
Systems
Presented to Rhode Island ICU Collaborative Novem
ber 6, 2008 by Charles M. Watts, MD Chief
Medical Officer Senior Vice President, Medical
Affairs
2
NMH Is Mission Driven with a Strategic Plan that
Guides the Organization

• Mission
• NMH Is an Academic Medical Center Where the
  Patient Comes First
• Strategic Plan
• To Provide the Best Patient Experience from the
  Patients Perspective
• To Recruit, Develop and Retain the Best People
  who Share the Organizations Values and Achieve
  Results
• To Develop the Resources to Achieve Our Mission
  and Vision through Exceptional Financial
  Performance

3
Focus on Quality and Patient Safety

• This is our mission and core to our physicians,
  nurses and clinicians
• Medical Staff Bylaws outline physicians
  responsibility for quality care and management
• Remains a critical area of focus
• Our patients continue to have preventable serious
  adverse events and do not consistently receive
  reliable evidence based care
• Publicly reported and available data is used
  increasingly to drive payment as well as
  reputation
• Professional liability expense is a significant
  component of the operating budget
• Focus on this agenda is expected by our Board of
  Directors
• Key measures reported to both PSC and NMH boards
• Only area of focus for the past 4 years of annual
  board leadership retreats

4
Donabedian Model for Measuring Quality (or Safety)

• Structure (How care is organized)
• Availability of appropriate resources
• Credentialing and privileging
• Presence of EHR, CPOE, protocols
• Process (What we do)
• How often do we use evidence-based protocols,
  care
• Outcome (The results achieved)
• How often do we harm patients?
• Culture (The context in which care is delivered)
• Ability of staff to raise concerns
• Ability of senior leaders to listen to concerns
  and act on them
• How conducive a system is to change

5
Achieving the Cultureof aHigh Reliability
Organization
6
Reliability

• Reliability Number of defects per opportunity
  for that defect
• Defects must be able to be validly measured as
  rates
• Rates need a clearly defined numerator (defect)
  and denominator (population at risk) and must be
  devoid of reporting biases. Can be process or
  outcome measures.
• Examples
• of CRBSI / 1000 catheter days
• Shoulder dystocia cases associated with injury /
  Shoulder dystocia cases (or number of deliveries)
• of patients receiving appropriate VTE
  prophylaxis / of patients eligible
• Problems
• Availability/measurements of baseline data
• Quality of data (measure fewer things better)
• Failure to couple measures with high performance
  grades

7
Characteristics of High Reliability Organizations

• Anticipating and becoming aware of the unexpected
• Preoccupation with failure
• Reluctance to simplify interpretations
• Sensitivity to operations
• Containing the unexpected when it occurs
• Commitment to resilience
• Deference to expertise

8
Anticipating and becoming aware of the unexpected
Preoccupation with Failure

• Chronic wariness is the tone in a safe
  systemhubris is the enemy
• Treat any lapse that occurs as a symptom that
  something may be wrong with the system, something
  that could have very severe consequences if
  several separate small errors happen to coincide
• Encourage reporting of errors
• Elaborate the experiences of a near miss for
  what can be learned. Near misses are not viewed
  as proof that a system has enough checks in it to
  prevent errors but rather as opportunities to
  better understand how the system could fail and
  fix it (an invitation to improve)
• Continuous effort to define/articulate mistakes
  you dont want to make, and what can be done to
  prevent them

9
Anticipating and becoming aware of the unexpected
Reluctance to Simplify Interpretations

• HROs refuse to simplify they accept that the
  nature of the work is complex, that their systems
  can fail in ways that have never happened before
  and that they cannot possibly identify all of the
  ways that their systems could fail in the future.
• Do not assume that failure or near misses are the
  result of a single, simple cause.
• Position yourself to see as much as possible.
  Less simplification allows you to see more. Look
  for differences from what you expect or have seen
  before.
• Example Use of RCA, FMEA

10
Anticipating and becoming aware of the
unexpected Sensitivity to Operations

• Must have deep knowledge of the technology, the
  system, ones co-workers, and most of all oneself
• Pay close attention to operations, to the front
  line where the real work is done (get house
  officers, nurses, pharmacists involved)
• Maintain situational awareness this is the only
  way that anomalies, potential errors, or actual
  errors can be quickly identified and addressed
• HROs are aware of close ties between sensitivity
  to operations and sensitivity to relationships
• Recognition that systems are not necessarily
  orderly, stable, and routinized, but rather
  dynamic with unfolding events, variation in the
  strengths and quality of hand-offs, complicated
  sequences of activities, variation in the way
  that intelligence is woven into work or stripped
  away from it, and the possibility that problems
  are sometimes the outcomes of attempted
  solutions.

11
Containing the unexpected when it
occursCommitment to Resilience

• Capability to detect, contain, and bounce back
  from the inevitable errors that are part of an
  indeterminate world
• An HRO is not error free, but errors dont
  disable it
• Combination of keeping errors small and
  improvising
• Imagine worst case conditions and practice their
  own equivalent of fire drills
• Managers in HROs take pride in the fact that they
  spend time putting out firesregard successful
  firefighting as evidence that they are resilient
  and able to contain the unexpected (not as a
  distraction that makes it difficult to do their
  real work)
• Assume you will be surprised, and concentrate on
  developing resources to react and record
• Fast real time learning (both content and
  process)
• Fast and accurate communication, feedback
• Modeling. Imagine worst case conditions and
  practice them
• Improvisation, maintain flexibility

12
Containing the unexpected when it
occursDeference to Expertise

• Distributed expertise
• Team members and organizational leaders defer to
  the person with the most knowledge/expertise
  relevant to the issue they are confronting
• Usually someone at the sharp end
• Not necessarily the most experienced person or
  the person highest in the organizational
  hierarchy
• Staff at every level need to be comfortable with
  this, either escalating concerns or asking the
  advice of someone lower in the pecking order
• De-emphasis on hierarchy is critical for
  organizations to prevent and respond to problems
  more effectively

13
A Mindful Infrastructure for High Reliability
Fundamental Characteristics General
Orientation Impact on Processes
Ultimate Outcome
Preoccupation with Failure
Reluctance to Simplify
Exceptionally Safe, Consistently High Quality
Care
Capability to Discover and Manage Unexpected
Events (High-Reliability)
State of Mindfulness
Sensitivity to Operations
Commitment To Resilience
Deference to Expertise
Weick, Karl, Managing the Unexpected
14
Sensemaking Making Sense of Events

• When an unexpected event occurs we first have to
  notice it, then we have to make sense of it, and
  finally we have to do something about it
• Sensemaking require a forum devoted to analysis
  and medical inquiry to make sense and create
  safe patient care
• Conversation among staff about a particular event
• Each person brings in their own unique knowledge
• Conversation is the mechanism that combines that
  knowledge
• Examples of sensemaking conversations
• Root Cause Analysis
• Failure Mode and Effect Analysis (FMEA)
• Data Mining
• Probabilistic Risk Assessment (PRA)

15
Creating HRO Culture

• James Reason A safety culture is an informed
  culture, one in which those who manage and
  operate the system have current knowledge about
  the human, technical, organizational, and
  environmental factors that determine the safety
  if the system as a whole.
• Intelligent wariness Collect and disseminate
  information about incidents, near misses, and the
  state of the systems vital signs
• Need consistent
• Reporting culture what gets reported
• Just culture who is blamed
• Flexible culture adaptation to sudden changes,
  events
• Learning culture Use lessons learned to
  reconfigure assumptions and actions

16
Structure and Process of Quality Management at
NMH
17
Quality Committee Structure
Provides Operational Framework To Improve
Quality And Patient Safety
18
Key Elements of National Healthcare Quality
AgendaAligned with NMH Strategic Plan

• Six IOM Aims
• Safe
• Effective
• Patient Centered
• Timely
• Efficient
• Equitable

• Best Patient Experience
• Deliver Safe Care
• Deliver Effective Care
• Coordinate Care
• Deliver Timely Convenient Care
• Provide the Best Physical Environment for Care
• Offer Advanced Expertise Through RE
• Be the Trusted Source of Information
• Improve the Health of Our Community

Crossing the Quality Chasm A New Health System
for the 21st Century
19
How Do We Select Measures for Best Patient
Experience?

• Domains are driven from strategy
• System level measures
• Patient satisfaction
• Effective, safe, convenient, timely, personalized
  care
• Key internal drivers
• Strategic new/expanded programs
• Clinical leaders recommendations
• Institutional experience (claims and pipeline
  review, CCEC cases)
• Control metrics for DMAIC projects
• Selected high-credibility external drivers
• JCAHO, CMS
• IHI, NQF, AHRQ, Leapfrog
• Magnet
• Every measure has a goal and a timeline

20
Comprehensive Claims Data ReviewPrioritization
of Drivers
21
Clinical Care Evaluation Committee
Adverse Event Data Review FY06 Overview
228 Cases were Reviewed, Leading to 396 Actions
and Referrals
Follow-up Categories
Follow-up Actions
8 follow up items led to a Root Cause Analysis
and associated improvements
Root Cause Analysis
228 Cases Reviewed
331 follow up items were referred to departments
and leaders for analysis and action or resulted
in PI projects
Improvement Actions
14 Severe Events (F-I)
86 Non-Severe Events (A-E)
57 items were referred to Quality Management
Committees, usually for review of physician
practice
Referred to QM Committee
22
Plan InputsData Sources for Quality Planning
Process
2007 Quality Planning Approach reviewed over 640
performance measures from both internal and
external sources

• Internal Sources
• BPE Dashboard
• Internal Committee Dashboards
• Surgical Services
• Department of Medicine
• Department of Obstetrics and Gynecology
• Nursing
• Emergency Department
• Critical Care
• Pathology
• Radiology
• CCEC/CAC Cases
• PurpleSurg (surgical resident risk database)
• Claims Study including PCEs

• Hospital-wide External Sources
• CMS/JCAHO
• NQF
• IHI
• AHRQ
• UHC
• Leapfrog
• HealthGrades

• Department Specific External Sources
• Department Regulatory Agencies/Professional Org.
• Surgical Care Improvement Project (SCIP)
• NSQIP (National Surgical Quality Improvement
  Program)
• National Perinatal Information Center (NPIC)
• Council of Womens Infants Specialty Hospitals
• Chi Solutions, Inc.
• College of American Pathologists
• National Nosocomial Infection Surveillance System
  (NNIS)
• National Database of Nursing Quality Indicators
• Consumer Assessment of Healthcare Providers and
  Systems

23
Internal Data Analysis Identifies...
Key Drivers of Patient Harm
Obstetrics
Surgical Services

• Shoulder Dystocia
• Uterine Rupture
• Hyper-stimulation
• Fetal Monitoring
• HIV
• Group Beta Strep
• 3rd 4th Degree Lacerations
• Bleeding or Hemorrhage

• Foreign Body Retained
• Procedural Complications
• Unintended Injury
• Surgical Site Infection
• Diagnosis Missed or Delayed
• Positioning

24
Summary of Current Key Risk Drivers for
Obstetrics Gynecology

• Key Risk Drivers identified from current open
  claims and recent 12 months of incidents reviewed
  by CCEC and CAC (Claims Advisory Committee) with
  severity index of E-I
• Current Key Risk Drivers are consistent with
  those identified in 2004 Claims Lookback Study
• Incident Key Risk Drivers are based on recent 12
  months of data, but lack historical trending data
  to determine improvement Modifying process and
  systems to support trending data for the future
• How do we know if interventions are working?

25
Process Improvement Delivers Measurable Results

• Adopted Industry Process Improvement Tools
  DMAIC (2002), Lean (2006)
• Since launching the program in 2002
• More than 400,000 patient encounters impacted
• More than 130 DMAIC projects completed
• 7 Million in annualized financial benefit, over
  40M in benefit to date
• 53 staff trained as Improvement Leaders through
  an advanced 6-month course
• More than 400 staff (including 98 of management
  team) trained in one-day DMAIC Foundation course

Define
Measure
Analyze
Improve
Control
26
OB Shoulder Dystocia What is it?
Intrauterine pressure is caused by maternal
contractions
Anterior shoulder impacted on symphysis pubis
Brachial plexus stretching
DANGER

• Difficult to predict
• Severe brain damage or death due to
  hypoxia/acidosis if delay in delivery
• Brachial plexus damage

A
D
M
C
I
27
OB Shoulder Dystocia Improvements

• Pre-Planned Emergency Response Protocol
• Training Sessions / Drills
• Birthing Simulator

28
OB Shoulder Dystocia Results
Compliance with protocol has averaged over 90
and the brachial plexus injury rate has decreased
by more than 60 since implementing pre-planned
response to shoulder dystocia
Compliance with Protocol
Brachial Plexus Injuries as a of Shoulder
Dystocias (Lower is Better)
Protocol Implemented
Protocol Implemented
Began Training
Began Training
Definition Documented compliance with
recommended shoulder dystocia practices for all
coded occurrences of shoulder dystocia.
Numerator all newborns whose medical record
indicates injury to brachial plexus (ICD9 767.6)
and whose mothers record indicates a shoulder
dystocia (ICD9 660.41). Denominator All cases of
shoulder dystocia.
M
A
D
I
C
29
NMH/NMFF Compliance with Pitocin Protocol
Significant improvement in protocol compliance
has been sustained. Expect dose advancement
compliance to improve from current level of 90
in order to meet goal. Strip management
compliance is 97.

• Dose and Advancement Correct dosage (the right
  concentration (15 units in 250 cc 0.9 normal
  saline) and initial dose (2 mu/min)
• Proper dose advancement (advancement at
  increments 2 mu/min gt 15 min intervals)
• Strip management interventions tachysystole
  (more than 7 contractions in 15 minutes) reduce
  the pitocin to the previous setting and
  reevaluate the tracing in 15 min if the
  tachysystole persists, continue to reduce the
  dosage until it is no longer present
• Prolonged deceleration gt 2 minutes discontinue
  Pitocin
• Included Population Obstetrical patients who had
  pitocin administered for labor induction/augmentat
  ion

30
Preventing Birth Trauma at Prentice (NMH)
NMH Birth Trauma Rate per 1,000 Births (AHRQ
Data)
Infants with 5 min Apgar lt6 and gt750 gmsl
FY2006 Average 3.76
13
11
7
7
6
FY2007 Average 2.51
4
Birth Trauma ICD-9 codes
767.0 Subdural and cerebral hemorrhage (due to
trauma or to intrapartum anoxia or
hypoxia) 767.11 Epicranial subaponeurotic
hemorrhage (massive) 767.3 Injuries to skeleton
(excludes clavicle)
767.4 Injury to spine and spinal cord 767.7 Other
cranial and peripheral nerve injuries 767.8 Other
specified birth trauma 767.9 Birth Trauma,
unspecified
31
Preventable Codes Outside ICU Per 1000 Patient
Days
Since the implementation of the Rapid Response
team (RRT) in January 2006, there has been an
overall reduction of 47 in preventable codes
outside the ICU

• Clinical judgment Staff nurses are trained to
  recognize the signs of decomposition and to
  activate the RRT
• Proactive rounding RRT nurses round on all ICU
  transfers within 24 hours of transfer
• Electronic surveillance - Electronic vital sign
  data is used to supplement clinician judgment,
  reduce the subjectivity of activating the RRT
  team and to earlier identify patients at risk

32
Eliminate Avoidable Severe Adverse Events
Increased Incident Reporting with Downward Trend
for Severe Avoidable Adverse Events from FY04 Q1
through FY08 Q3. FY08 Q1-Q3 achieved an 87
reduction in avoidable severe adverse events
compared to FY04 Q1-Q3.
Total Incidents Reported
of Severe Harm Events
of Incidents Reported
Severe Harm
Severe Events
G - Event may have contributed to or resulted in
permanent patient harm H - Event occurred that
required intervention necessary to sustain life I
- Event occurred that may have contributed to or
resulted in the patients death
33
Are continued reductions in lawsuit filings
expected?
FY08Q1-Q3 achieved an 87 reduction in avoidable
severe adverse events compared to FY04Q1-Q3.
Based on typical two year filing lag, FY06 and
FY07 events are considered predictors of FY08 and
FY09 lawsuits respectively, thus continued
reductions in lawsuit filings were expected but
have not materialized.
NETSImplementation
Total Incidents Reported
Predictor for FY08 Lawsuits
Predictor for FY09 Lawsuits
G-IIncidents
In FY05Q4 count methodology changed for G,H,I
CCEC cases from Date of Event to Date
Presented to CCEC Baseline period Severe
Events G - Event may have contributed to or
resulted in permanent patient harm H - Event
occurred that required intervention necessary to
sustain life I - Event occurred that may have
contributed to or resulted in the patients death
34
NMH Has Been Recognized for Quality and Excellence

• UHC Top 10 for Quality and Accountability
  2006-2007
• Achieved Magnet Status, 2006
• 10 Specialties in 2008 U.S. News World Report
  of Best Hospitals
• 2005 National Quality Health Care Award
• Most Preferred Hospital for 13 Years (NRC)
• Leapfrog Groups Top Hospitals List, 2006
• Named to 100 Best Companies for Working Women
  for 8 Years
• Most Wired for 8 Years

35
NMH Is a Top Performer Among Academic Medical
Centers
Data Source 2007 UHC Quality and
Accountability Effectiveness Score (core measures
and readmission rate)
36
Imagine a world
37
What About the Other 13?
38
Measures
Stretch Outcome Measures (Zero-defects)
Process Measures
Outcome Measures
15
85

• Culture
• Not Measurable
• Not Predictable

• Standard Process
• Defined problem
• Metrics
• Goal improvement project
• Implementation

39
Central Line AssociatedBlood Stream Infections
40
Overview of Catheter-Related Infections in ICUs
N
Engl J Med 20063552781-2783
Annual Patient Stays in the 6000 Acute Care
Hospitals and Associated ICUs in the United
States. About half the days patients spend in
ICUs (ICU days) are associated with the use of a
central venous catheter and therefore with a risk
of subsequent bloodstream infection (five
infections per 1000 catheter-days).
41
Using Bundles to Improve Reliability

• A Bundle is a grouping of evidence based
  processes with proximate time and space
  characteristics that when performed together can
  improve outcomes
• Small in number
• Bundles demand all or none thinking and
  measurement
• Bundles facilitate identification of failures,
  and redesign of process
• Bundles are associated with improved teamwork and
  communication

42
Process MeasuresCentral line insertion
bundle(CLABSI Central Line Associated Blood
Stream Infection)

• Hand hygiene (insertion and maintenance of lines)
• Maximal barrier precautions upon insertion
• Chlorhexidine skin antisepsis
• Optimal catheter insertion site selection (avoid
  femoral line insertion)
• Remove unnecessary lines promptly (convert to
  non-central lines)

43
Central Venous Catheter Aggregate Infection Rate
A number of interventions over time have
contributed to a significant decrease in
infection rates within the intensive care units.
Audit volume remains variable due to manual data
collection. The number of patients can range from
3-20 per month and reflects central venous
catheters (CVC) placed initially in the ICU,
excluding those placed in other procedure areas.
Beginning in March 2008, the CVC dressing data is
no longer part of the NMH CVC bundle.
Line Carts
PICC Team
Phlebotomy
ICU-based Committees

• CVC Guidelines
• Practice hand hygiene
• Maximal barrier precautions upon insertion
• Use of chlorhexidine skin antisepsis

• Central line associated bloodstream infection in
  a patient with microbiologic and/or clinical
  infection manifestations and no apparent source
  for the infection at another body site
• Includes MICU, SICU, NSICU, CTICU and CCU

44
(No Transcript)
45
What does it take to get to zero?

• Effective improvement is achieved through
• Methodical application of successive iterations
  of systems based solutions
• Additional process measures, bundles
• Checklists
• Daily goals
• Provision of necessary infrastructure at the
  appropriate time and place
• Equipment
• Technology
• Training (team, individual, simulation)
• Feedback information
• Embracement of necessary cultural and behavioral
  change

46
Michigan Keystone CR-BSI Rate Project
For each line insertion
CVC Policy
Line Cart
Daily Goals
Checklist
47
What does it take to get to zero?

• Characteristics of High Reliability Organizations
• Anticipating and becoming aware of the unexpected
• Preoccupation with failure
• Reluctance to simplify interpretations
• Sensitivity to operations
• Containing the unexpected when it occurs
• Commitment to resilience
• Deference to expertise

48
Michigan Keystone CR-BSI Rate Project
Empower Nursing
49
Creating HRO Culture

• Arm yourself for guerilla warfare.
• Create an error-friendly learning culture.
• Reward reporting. Encourage clinicians, patients
  and others to be vigilant in identifying
  potential or actual errors, and take appropriate
  steps to prevent or mitigate harm.
• Provide feedback, share widely.
• Engage providers. Make adverse events real, tell
  stories, involve the patient, say you are sorry.
• Cognitive reframing of the possible.

50
Imagine a world
51
Without Avoidable Harm