Improving the Performance of the Public Mental Health System: Quality Assessment and Improvement

1
Improving the Performance of the Public Mental
Health System Quality Assessment and
Improvement
2
Berwicks Question

• How can one tell whether or not a healthcare
  organization is really serious about improving
  its quality, instead of simply engaging in
  defensive measurement to protect itself against
  the demands of outsiders for information?
• Don Berwick, The Basic Concepts of Quality
  Improvement, unpublished paper, 1987

3
Quality Assessment and Improvement

• Definitions
• The Quality Environment
• The Quality Assessment Component
• DMAIAn Ongoing Process
• Fundamental Principles of Quality Management

4
Definitions

• Performance measurement is the regular
  collection and reporting of data to track work
  produced and results achieved
• Performance measure is the specific quantitative
  representation of capacity, process, or outcome
  deemed relevant to the assessment of performance
• Performance standard is a generally accepted,
  objective standard of measurement such as a rule
  or guideline against which an organizations
  level of performance can be compared

5
Definitions

• Performance management is the use of performance
  measurement information to help set agreed-upon
  performance goals, allocate and prioritize
  resources, inform managers to either confirm or
  change current policy or program directions to
  meet those goals, and report on the success in
  meeting those goals
• Performance measurement is NOT punishment
• Guidebook for Performance Measurement ,
• Turning Point Project

6
Definitions

• Two of the primary uses for the results of
  performance measurement are for
• Making comparisons of performance levels
• Improving the quality of the processes and
  outcomes of the organization
• The American College of Mental Health
  Administration (ACMHA) applied these distinctions
  between comparison and quality improvement in the
  proposed Consensus Set of Indicators for
  Behavioral Health

7
Definitions

• In the ACMHA project, five national accreditation
  entities reached consensus on a set of
  performance measures (CARF The Rehabilitation
  Commission, the Council on Accreditation, the
  Council on Quality and Leadership in Support of
  Persons with Disabilities, JCAHO and NCQA) but
  not on the specifications for measurement
• They concluded that it was important to
  recognize that selecting appropriate measures
  depends on the purpose of assessing performance
• They designated measures as either a comparison
  measure or a quality improvement measure to
  clarify the intended use of each measure and its
  data set

8
Definitions

• For comparison purposes, the standards and
  measures should provide sufficiently valid and
  reliable quantification such that comparison
  across the systems programs and departments can
  be made. By identifying the highest level of
  performance or outcome (the benchmark), an
  organization can duplicate those work processes
  to achieve higher performance overall.
• For improving quality, some standards and
  measures lend themselves more to internal
  monitoring of performance and local
  accountability and are most suitable for
  supporting the improvement of the organization
  rather than for comparability among organizations.

9
Definitions

• Balanced Budget Act (BBA) of 1997 was a
  substantial rewrite of the Medicaid and Medicare
  program rules. Final rules were passed on
  6/14/02 protocols and checklists then rolled
  out. Details in the protocols and checklists are
  critical for an understanding of BBA impact.
• External Quality Review Organization (EQRO) is an
  independent entity that meets competence criteria
  for conducting Medicaid EQR activities EQROs are
  being selected through state procurement
  processes to review the operations of risk
  bearing organizations contracting with state
  Medicaid agencies.

10
Definitions

• Managed Care Organization (MCO) is Medicaids
  term for a health plan that provides health care
  services to Medicaid enrollees
• Examples include Group Health Cooperative,
  Community Health Plan of Washington and Molina
• Prepaid Inpatient Health Plan (PIHP) is
  Medicaids term for a health plan that provides a
  more limited range of services than an MCO, for
  specialty services such as mental health
• Examples include the Washington State Regional
  Support Networks (RSNs) and Oregons Mental
  Health Organizations (MHOs)

11
The Quality Environment

• Crossing the Quality Chasm a New Health System
  for the 21st Century, Institute of Medicine (IOM)
  2001
• Redesign of the health care system based on 10
  new rules
• Build organizational supports for change,
  including the incorporation of care process and
  outcome measures into daily work and revising
  financial methods to support quality work
• Priority Areas for National Action Transforming
  Health Care Quality, IOM 200320 priority areas
  selected including major depression (screening
  and treatment) and severe and persistent mental
  illness (focus on treatment in the public sector)

12
The Quality Environment

• Currently an IOM Committee is studying how to
  adapt the Quality Chasm recommendations to Mental
  Health and Addictive Disorders
• In December 2004, a meeting was co-hosted by the
  National Council for Community Behavioral
  Healthcare (NCCBH), RWJ Center for Health Care
  Strategies and SAMHSA to frame a National
  Initiative for Behavioral Health Care Quality
  Improvement (CMS participated in this effort, an
  opportunity to generate their support as well as
  to foster relationships)

13
The Quality Environment

• The Institute for Healthcare Improvement (IHI)
  and Don Berwick, MD, have led the healthcare
  dialogue from its early beginnings
• IHI is a healthcare industry focal point through
  National Forums, trainings, and Breakthrough
  Series that target reducing adverse drug events,
  medical errors or reducing delays and waiting
  times throughout the system
• IHI partnered with Health Resources and Services
  Administration (HRSA) in developing and staffing
  the Health Disparities Collaboratives for
  Federally Qualified Health Centers (asthma,
  diabetes, depression)

14
The Quality Environment

• National Committee for Quality Assurance (NCQA),
  created jointly by healthcare purchasers and HMOs
  to assess, measure and report on the quality of
  care provided by managed care organizations
• Measures performance through HEDIS, a
  standardized measurement system for MCOs
• Accredits MCOs using standards grounded in
  QIaccreditation is based on a combination of
  accreditation survey scores and scores on HEDIS
  measures (33 of 100 points)
• MAA incorporates most of the NCQA quality
  standards in MCO contracts, collects selected
  HEDIS performance measures, and uses NCQA
  accreditation for a major part of the EQRO review

15
The Quality Environment

• Joint Commission on Accreditation of Healthcare
  Organizations (JCAHO) accreditation process has
  shifted from survey preparation and scores to
  continuous operational improvement in support of
  safe, high-quality care
• ORYX core measure data are used to continually
  assess key performance areas, and eventually will
  be incorporated into the organizations
  performance report as core measures are adopted
  for programs
• JCAHO prepared the CMS protocols for BBA EQROs to
  use in review of Medicaid MCOs and PIHPs

16
The Quality Environment

• BBA requires EQROs to operate within specific
  protocols
• Determine Compliance with Federal Medicaid
  Managed Care Regulations
• Validate performance measures and methods of
  calculating measures of performance
• Validate Performance Improvement Projects (PIPs)
  and methods of conducting a PIP
• Conduct an Information Systems Capabilities
  Assessment (ISCA)

17
The Quality Environment

• BBA rules require that the MHD and PIHPs have a
  Quality Assessment and Performance Improvement
  Program (QAPI) that includes mechanisms to detect
  both under-utilization and over-utilization
• MHD is expected to continuously and consistently
  monitor the appropriateness and quality of the
  consumer care delivery system in PIHPs
• MHD infrastructure is charged with reviewing
  statewide mental health data, recommending system
  improvements, and designing and implementing
  quality improvement projects and processes

18
The Quality Environment

• Sample questions from the EQRO protocol include
• Have any recent QAPI activities been implemented
  to monitor compliance with established standards
  for timeliness of access to care and member
  services?
• What types of information does the program
  provide to support recredentialing of providers?
• How does your PIHP detect over- and
  under-utilization? Provide examples.
• How are enrollee and provider data from all
  components of your network used in your QAPI?

19
The Quality Environment

• So we have a QAPI and PIPsits all about the
  bureaucracy, right?wrong, the work must have
  relevance to the organizational vision, mission
  and goalsits about achieving your purpose and
  serving your consumers
• Requires leadership commitment and a deep
  understanding of the vision and mission of the
  system and/or organization
• If you cannot tell 1) how a project specifically
  relates to your agencys vision and mission, or
  2) (worse) if you cannot tell how your agencys
  mission and vision relate to quality, the project
  should be sidelined until you can
• Hayes and Nelson, A Handbook Of Quality Change
  Implementation For Behavioral Health

20
The Quality Assessment Component

• Are your system decisions made in a data-free
  environment?
• How do you know if your agency is achieving its
  goals?
• How do you know when you should initiate a PIP?
• How will you decide on implementing practice
  guidelines?
• How will you know if (and why or why not) the PIP
  or practice guideline is successful?

21
The Quality Assessment Component

• JCAHO ORYX Core Measurement Sets
• Data Quality Principles (Handout 1)
• May vary by setting or by key issue
• Relate to the basic principles of care, process
  oriented
• Performance measure categories considered useful
  in the accreditation process
• Clinical
• Health status
• Perception of care/service
• Categories not considered useful include
• Financial measures
• Utilization measures, unless related to a
  standard of quality

22
The Quality Assessment Component

• For JACHO Behavioral Health, requirements differ
  by type of organization
• Organizations providing 24 hour care
• Select a minimum of six clinical, health status
  or perception of care measures from the set of
  JCAHO approved measures
• Measures must focus on the clients that receive
  24 hour services
• Organizations providing non-24 hour care and/or
  24 hour care for an ADC of less than 10
• Select at least six measures from any relevant
  source,
• Share data, analytic conclusions and actions
  taken with surveyor
• In future will be expected to select and enroll
  in a listed performance measurement system when
  core measures relevant to their services are
  identified

23
The Quality Assessment Component

• NCQA HEDIS BH Measurement Sets (Handout 2)
• Follow up after hospitalization for mental
  illness
• Antidepressant medication management
• Mental health utilizationinpatient discharges
  and average length of stay
• Mental health utilizationpercentage of members
  receiving services
• Chemical dependency utilizationinpatient
  discharges and average length of stay
• Initiation and engagement of AOD dependence
  treatment
• Identification of AOD services

24
The Quality Assessment Component

• PIHP Measurement Sets
• Verity examples (Handout 3)
• ACMHA Indicators (Handout 4)
• PIHP Master Calendar (Handout 5)
• Pilot Measurement of Initiation and Engagement
  (Handout 6)

25
The Quality Assessment Component
26
(No Transcript)
27
The Quality Assessment Component

• The quality assessment component requires that a
  group of selected indicators are regularly
  tracked and reported
• The data should be regularly analyzed through the
  use of control charts and comparison charts (Stay
  tuned for details!)
• These indicators should tell you if you are
  achieving your agency goals and objectives
• These indicators can provide the basis for
  deciding when a PIP might be indicated and the
  baseline information for measuring the future
  impact of PIPs

28
The Quality Assessment Component

• The quality organization does not wait to be told
  (via regulations or requirements) what processes,
  procedures, or programs to implement. Instead the
  quality organization proactively implements a
  program that it recognizes it may have to alter
  as standards or regulations are developed
• Consider the ACMHA list of indicators as a
  starting placeit includes measures of what
  quality service means to consumers

From Hayes and Nelson, A Handbook Of Quality
Change Implementation For Behavioral Health
29
The Quality Assessment Component

• Quality assessment is an absolutely necessary,
  but not sufficient, step to change from a
  data-free environment to a culture of
  measurement

30
DMAIAn Ongoing Process

• Living in a the plan-do, plan-do world?
• Too busy fighting fires to close the loop?
• Quality assessment indicators must have relevance
  to the organizational vision, mission, goals and
  objectives
• PIPs based on goals and objectives must use a
  Design/Measure/Analyze/
• Improve cycle

31
DMAIAn Ongoing Process

• MHD Implementation and Design Group has
  designated a standard methodology for the
  infrastructure to use in assessing, choosing,
  developing, monitoring and evaluating QI
  opportunities and outcomes
• The Design, Measure, Analyze and Improve (DMAI)
  model is
• Congruent with and supported by JCAHO
• Data drivenintegrates trending, tracking,
  analysis and action into day-to-day processes
• Already used by numerous providers and at least
  one RSN

32
DMAIAn Ongoing Process
From Hayes and Nelson, A Handbook Of Quality
Change Implementation For Behavioral Health
33
DMAIan Ongoing Process

• Designtwo steps at the beginning
• Establish objectives of the project
• Establish the processes used to meet the
  objective
• Measure
• Establish the specific outcome and process
  measures the project will use for baseline and
  post implementation measurement

34
DMAI-An Ongoing Process

• Analyzetwo types of analysis
• Use statistical and numerical methods
• Use comparative methods
• Improve
• Implement revised processes until analysis of
  measures indicates that the objectives have been
  met

35
DMAIAn Ongoing Process

• DMAI adds clarity about objectives into the
  cycle this is implied, but not specified in the
  PDCA or PDSA, and two types of analyses are
  specified

36
Fundamental Principles of Quality Management

• Know your customers and what they need
• Focus on processes
• Use data for making decisions
• Understand variation in processes
• Use teamwork to improve work
• Make quality improvement continuous
• Demonstrate leadership commitment

37
Know Your Customers
Design
Improve

• Identify customers and their needsin
  healthcare there are usually two sets of
  customers
• The people who use your services are the primary
  customers
• The purchasers of your services also have
  requirements
• Set goals based on their needs and DMAI
  objectives based on the goals
• Monitor performance and satisfaction to target
  performance improvementopportunities
• Improve or redesign how work is done

38
Focus on Processes
Design
Improve

• 85 of poor quality is a result of poor work
  processes, not of staff doing a bad job
• When things go wrong, it is often at the point of
  the handoff in the process
• Attend to improving the overall design, not just
  one partsome of the most complex and poor
  quality processes are the result of improving
  and creating work arounds at some steps instead
  of redesigning the entire process

39
Focus on Processes
Design
Improve

• Advice from NCQA, JCAHO and othersmeasure
  processes that are
• High-risk
• High-volume
• Problem proneAnd
• Can be tracked and reported as summary or
  aggregate statistics
• Are being selected by other organizations to
  allow statistically valid comparisons to be made
  (for purposes of benchmarking)

40
Use Data to Make Decisions
DMAI

• Use performance assessment data to target
  improvement
• Use data analysis tools to develop information
• Analyze data to identify root cause
• Use data to monitor performance outcomes

41
Use Data to Make Decisions
Measure
Analyze

• Collection of data on clinical outcome alone does
  not provide useful information about what led to
  the outcome, or how it can be replicated or
  improved
• Pairing collection of outcome data with data on
  key process performance measures associated with
  the outcome will provide information on the
  consistency of the process of care
• Statistical analysis of these sets of data tells
  an organization whether it is improving
  performance on outcomes while improving
  consistency

42
Use Data to Make Decisions
Measure
Analyze
43
Use Data to Make Decisions
DMAI

• Symptom is the indication of a problem, but not
  a statement of cause
• Theory is the preliminary diagnosis about the
  cause
• Analysis includes data that confirms or rules
  out theories
• Solutionis the change that will best address
  the cause
• Information is data that confirms whether the
  solution is having the expected impact

44
Use Data to Make Decisions
DMAI
From Hayes and Nelson, A Handbook Of Quality
Change Implementation For Behavioral Health
45
Use Data to Make Decisions
DMAI
Numerical Tools
Conceptual Tools

• Check Sheet
• Bar Chart
• Histogram
• Pareto Chart
• Control Chart
• Run Chart

• Affinity Diagram
• Brainstorming
• Process Flow Chart
• Interrelational Diagraph
• Matrix Diagram
• Tree Diagram
• Cause and Effect Diagram

46
Use Data to Make Decisions
Design

• Conceptual tools support theory generation
  regarding root causes, a key step in the PIP
  process
• Root causes
• In the logical chain of causes
• Directly and economically controllable
• Can be considered a constant part of (or
  deficiency in) the process under study
• If eliminated, the problem disappears or is
  drastically reduced (the Pareto Principle or
  80-20 rule)
• Initiating a PIP that defines a desired solution,
  rather than a process to be studied can be
  hazardous to your QAPI s health!

47
Use Data to Make Decisions
Design

• Brainstorming for root causestheory generation
  thrives on divergent thinking, so no idea is a
  bad one
• What can go wrong in the process we are studying?
• Problems in hand-offs between steps
• Problems in execution within steps
• Look at machines, materials, methods,
  measurements, and people
• Cause-effect or Fishbone diagram (Handout 7)
• Organizes and displays theories
• Encourages divergent thinking
• Demonstrates the complexity of the problem
• Encourages scientific analysis (rule-out)
• Failure to use a Cause-effect diagram or use of
  an incomplete one, can be hazardous to a PIPs
  health!

48
Use Data to Make Decisions
Measure
Analyze

• Numerical tools support analysis of PIP theories,
  measurement of PIP implementation and ongoing
  assessment
• Specific theories are needed to drive data
  collection and analysis
• Data collection and analysis leads to the vital
  few root causes by narrowing the competing
  theories of cause
• Look for clusters of causes that can be tested
  together
• Use stratifying variables to localize the problem
  and identify likely causes
• Do Pareto analysis of symptoms and theories

49
Use Data to Make Decisions
DMAI

• From Methods and Tools of Quality Improvement
• Institute for Healthcare Improvement

50
Use Data to Make Decisions
Measure
Analyze

• The Four Dimensions of Variability

Shape
Center average, median or mode
Spread range or standard deviation
Sequence trend
From Methods and Tools of Quality Improvement
Institute for Healthcare Improvement
51
Use Data to Make Decisions
Measure
Analyze

• The average by itself is not a good summary of
  data, use a variety of numerical summaries
  (Handout 8)
• Measures of center include
• Average/Mean the total data values divided by
  the total number of observations
• Median the middle value in the data set, half of
  the data value lie above, half lie below the
  median
• Mode the most frequently occurring values in the
  set of data
• Use histograms to look at overall variation
  patterns
• Use line graphs to look at patterns over time

52
Use Data to Make Decisions
Measure
Analyze

• Pareto Principle
• In any group of things that contribute to a
  common effect, a relative few contributors will
  account for the majority of the effect
• These few contributors are call the vital few
  while the many other contributors are called the
  useful many
• The vital few hold the greatest potential gain
  from quality improvement efforts
• Pareto DiagramA fact based tool for priority
  setting in quality improvement efforts (Handout 9)

53
Use Data to Make Decisions
Measure
Analyze

• Control charts (Handout 9)
• Variation in performance data is the result of a
  complex system of causes
• Variation in this system of causes has
  characteristics of random variation
• Used for ongoing quality assessment, control
  charts can help decide when to take action on the
  process based on the data
• Statistics provide standard distributions and
  mathematical methods for testing common and
  special cause variation

54
Understand Variation
Measure
Analyze

• Sources of variation include machines,
  materials, methods, measurements, people,
  environment
• Control charts are pictures of trend data with an
  extra featurethe range of variation built into
  the system
• Common cause variation occurs if the process is
  stable variation in data points will be random
  and obey a mathematical lawit is said to be in
  statistical control, with a large number of small
  sources of variation
• If an organization reacts to random variation in
  a process that is stable/in statistical control,
  it is called tampering and leads to further
  complexity, increasing variation and mistakes

55
Understand Variation
Measure
Analyze

• Special cause variation arises because of
  specific circumstances which are not part of the
  process all the time and may or may not ever
  recurif the recurrence is periodic, clues to the
  root cause may emerge
• Not in statistical control is
• One data point above or below the upper/lower
  control limits (three standard deviations)
• Two out of three consecutive data points beyond
  two standard deviations
• Of five consecutive data point, four are on the
  same side of the mean and beyond one standard
  deviation
• Eight consecutive data points are on the same
  side of mean
• Need to investigate special cause variation
  before making any conclusions about performance
  level

56
Understand Variation
Measure
Analyze

• Dont redesign an entire process (a PIP) when
  there is special cause variation, because there
  is not a consistent process to improve or stable
  baseline data to measure the impact of PIP
  implementation
• A sentinel event is a special cause variation
  requiring root cause analysis
• Examine specific incident(s) of special cause
  variation and make changes to a single element
  only after very careful analysis
• Failure to distinguish between common and special
  cause variation can be hazardous to
  organizational performance!

57
Understand Variation
Measure
Analyze

• Control chart analysis is done before comparison
  analysis to ensure a given process is stable
  before evaluation of relative performance level
• Comparison charts are based on multiple
  organizations performance data (or on
  standards/benchmarks that have been adopted) and
  are used to evaluate relative performance level
• If the process is stable, the only way to make
  improvements is to fundamentally change some
  aspect of the processthrough a redesign of the
  process or PIP
• Use benchmark data to create a new control chart
  that raises the bar on consistency of expected
  average performance and control limits

58
Use Teamwork
Design
Improve

• PIPs need buy-in from all stakeholders
• Process being studied is stable, but complex
• Creative ideas are needed
• Division of labor is needed
• Process often crosses functions
• Solution generally affects many

59
Use Teamwork
Design
Improve

• Those who do the work
• Have theories about the cause
• Have the detailed knowledge needed for conceptual
  analysis tools such as fishbone diagrams
• Have the clinical and intuitive judgment needed
  for design work
• Have ideas about improving the processes
• Improving processes means changeinvolvement in
  planning change and having staff that are seen as
  leaders for the change will be critical to
  successful implementation
• Open, safe communication is critical for
  improving processes (Handout 10)

60
Use Teamwork
Design
Improve

• Provide every team with a clear charge and
  support resources
• Teams should adopt working agreements (everything
  from cell phone rules to decision procedures)
• Teams should have assigned roles of facilitators
  and recorders
• The team process has some predictable stages that
  it is useful to keep in mind
• Forming
• Storming
• Norming
• Performing

61
Make QI Continuous
DMAI
From Hayes and Nelson, A Handbook Of Quality
Change Implementation For Behavioral Health
62
Make QI Continuous
DMAI

• QI is a system-wide approach to assessing and
  continuously improving quality of the processes
  and services over time
• See inter-relationships, not parts
• Understand the flow of work, not the one-time
  snapshot
• Detail the work processes
• Determine cause and effect relationships
• Identify points of highest leverage
• Improve and innovate, not just change for
  changes sake
• A way of doing business, not the exclusive
  responsibility of one individual or a committee

63
Make QI Continuous
DMAI

• Use quality assessment to identify areas for
  improvement
• Charge PIP team and provide support
• Provide DMAI training
• Use tools to understand root causes
• Use data for baseline and analysis
• Design process improvement to address root causes
• Train ... train ... train staff on the newly
  designed process improvement
• Evaluate the impact of process improvements
• If you dont get the results you expected.use
  assessment to understand why, revise accordingly
  and try again

64
Make QI Continuous
DMAI

• Measure improvement over time and compared to
  benchmarks
• Health of the people served
• Customer satisfaction
• Cycle time
• Accuracy/consistent features
• Financial performance
• Quality assessment is critical to measure the
  impact of PIPs and practice guidelines
• Measurement of baseline rates
• Initial and second remeasurement after
  implementation
• Plan to ensure demonstrated improvement can be
  maintained over time
• EQROs are looking for performance measurement
  over time

65
Demonstrate Leadership Commitment
DMAI

• Build a QAPI culture
• Connect the organizations strategic plan to
  performance improvement
• Know and use quality principles
• Encourage all staff to use quality improvement in
  daily work
• Reward improvements
• Assure adequate QAPI infrastructure for quality
  assessment and improvement activities

66
Demonstrate Leadership Commitment QAPI Culture
DMAI

• Clearly stated and enacted constancy of purposea
  deep understanding of the vision and mission
• Regular review of key indicator data
• Decisions made on data rather than hunches or
  opinions
• Long range view supports search for root causes
  and permanent solutions rather than quick fixes
• Focus on systems rather than individuals
• Continued identification of improvement
  opportunities
• Publicize successes (Handout 11)
• Clear communication agency-wide regarding the
  commitment to quality and the change processes
  necessary to implement improvement

67
Demonstrate Leadership Commitment QAPI
Infrastructure
DMAI

• Governance
• Oversight and accountability
• Program structure
• Who will do what when, with what processes for
  recommending or deciding
• Staff
• Support for ongoing monitoring and analysis, for
  training and facilitating improvement activities
• Data system
• Collect data and report in a user friendly way

68
Demonstrate Leadership Commitment QAPI
Description
DMAI

• Goals
• Organizational structures, responsibilities, and
  flow of information
• Quality council/committee
• Method for selecting PIP projects, charging PIP
  work teams
• Scope
• Programs/services/staff included
• Processes included
• Process for using quality assessment results to
  plan changes

69
Demonstrate Leadership Commitment QAPI Work Plan
DMAI

• Goals
• Important aspects of care/services
• Activities that involve a high volumes, high
  degree of risk and/or tend to produce problems
  for patients or staff
• Monitoring activities associated with important
  aspects of care/services
• Methods of measurement, frequency, timelines for
  reporting
• Consumer satisfaction monitoring

70
Demonstrate Leadership Commitment QAPI Work Plan
DMAI

• Planned PIPs (in process, new) and timelines
• Evaluation of PIPs now implemented and timelines
• Annual evaluation of QAPI workplan and program
  description, with proposed revisions

71
Berwicks Answer

• Quality improvement is grounded in values,
• It begins with a commitment at the top of the
  organization,
• It takes money,
• It has mechanisms for horizontal integration of
  quality measurement and control up and down the
  line of management and is relevant to front line
  staff,
• It requires statistical sophistication,
• The focus is on design, not simply on
  performance,
• Management is responsive and looks for ways to
  remove obstacles to improvement, and
• There is a strategy to drive out fear