Transcript Introduction to Database Management

HA 608 - Lecture 6
Decision Support Systems
Knowledge Management
Janet Guptill – HA 608 – October 1, 2007
1
MAJOR TYPES OF SYSTEMS






EXECUTIVE SUPPORT SYSTEMS (ESS)
[Strategic Level]
MANAGEMENT INFORMATION SYSTEMS (MIS)
DECISION SUPPORT SYSTEMS (DSS)
[Management Level]
KNOWLEDGE WORK SYSTEMS (KWS)
OFFICE AUTOMATION SYSTEMS (OAS)
[Knowledge Level]
TRANSACTION PROCESSING SYSTEMS (TPS)
[Operational Level]
2
Key Terms
 Decision Support – analyzing data, often from
different sources, to make better decisions
 Decision Support Systems (DSS) automate
decision support
 Expert Systems automate decision-making
 Executive Information Systems (EIS) provide
“dashboards” to assess operational performance
 Clinical Decision Support Systems (CDSS)
enhance patient care decision-making
 Knowledge Management (KM) incorporates
evidence- and experience-based information
3
Decision Support - A Possible Definition:


Decision Support - an organization’s
use of data in order to improve its
managerial and clinical decisionmaking effectiveness.
NOTE: Above definition makes NO
mention of computers!
4
Steps in Using Data to Make Decisions:







Formulate the Decision Problem
Obtain Appropriate Data
Summarize Data
Create a Model
Use Model to Evaluate Alternatives
Choose an Alternative
Implement the Alternative
5
Approaches to Incorporating
Data in Decision Making




Manual – collect data and logically
organize it to support decision process
Spreadsheets, Statistical Software, etc.
Request “IT Department” to generate a
report
Use Decision-Support System (DSS) that
integrates needed data and provides
analysis framework
6
A Sample Decision Making Problem



You are the Executive Director of a 21physician multi-specialty clinic.
You currently purchase MRI services on
a discounted fee-for-service basis from a
local hospital.
You have begun to think about building
the capability of providing these services
in-house.
7
A Sample Decision Making Problem
(Cont’d)




You are trying to decide when to begin
installing the necessary MRI equipment
and when to start a search for a
radiologist with expertise in this area.
You realize that this decision should be
based on data.
How do you make this decision?
What data do you need?
8
Test the feasibility of a freestanding
center
Map forecasted procedures by zipcode,
overlay existing imaging provider sites
Estimate demand based on 3-5 mile
radius and realistic market share targets
“Reality test” the demand numbers to be
certain sufficient volumes are attainable
Create the financials, staffing plans, and
marketing strategies
9
St Louis 2001 Major Imaging
Procedures by site of care
180,000
160,000
140,000
120,000
100,000
Hospital OP
Freestanding
80,000
60,000
40,000
20,000
0
MRI
CT
PET
SPECT
MISC
10
Major Imaging Procedures expected to grow
over next 5 years in St Louis due to greater
adoption of imaging technology in the market
400,000
350,000
300,000
250,000
2002
2007
200,000
150,000
100,000
50,000
0
MRI
CT
PET
SPECT
MISC
11
St Louis area map—Focus on St Charles
County
12
St Charles County - Existing Imaging
Centers - 2001
13
What volumes are needed to be worthwhile?
What are realistic start-up volumes?
(assumes 260 days/year)
Breakeven
Volume/Day
Year 1
Volume/Day
Year 2
Year 3
Volume/Day Volume/Day
MRI
7
10
15
20
CT
5
10
15
20
PET
4
4
6
8
SPECT
5
5
7
10
14
What decisions would you make with
these data?
Mkt 1:
63301
63303
63304
63376
Market 1
Mkt 2:
63385
63366
63367
Market 2
2007 Proc
Forecast
Breakeven
Mkt Share
Capacity
Mkt Share
CT
25,319
5%
21%
MRI
14,277
13%
36%
PET
529
196%
392%
5,022
26%
52%
11,225
12%
46%
MRI
6,329
29%
82%
PET
232
448%
896%
2,191
59%
119%
SPECT
CT
SPECT
15
Desirable Attributes of a Decision
Support System





Easy Interaction With the System
Executives Can Retrieve Data
Themselves
Data are Displayed in a Meaningful
Format
System has Modeling Capability
System Generates Clear Reports
16
First, a look at Database Management
Systems (DBMS)
Describe How Database Management
Systems Organize Data
Identify 3 Database Models,
Principles Of Database Design
Discuss Database Trends
17
Two Definitions:
Database - collection of data carefully
organized to be of value to a user
Database Management System (DBMS) software used to manipulate the database
18
Example Database
Field
Employee
Name
Date of
Hire
Social
Security #
Language
Fluency
Ken L. Watt
03/03/86
111-23-3223
None
Jane Sargent
11/10/90
356-29-0588
German
Record
File
Mary Smith
05/05/97
334-44-9876
Spanish
.
.
.
.
.
.
.
.
.
.
.
.
Robert Cardin
09/12/92
056-88-4848
French
19
An Overview of Database Models
Hierarchical Data Model - stores data as
nodes in a tree structure
Department
ROOT
FIRST
CHILD
2nd CHILD
Employees
Equipment
Technician
Maintenance
Records
20
Use of “Pointers” to connect records
Field In One Record Is Address Of Next
Record In Sequence
RECORD 1
POINTER
RECORD 2
RECORD 3
POINTER
POINTER
21
Types Of Relationships
Found in a Hierarchical Data Model
ONE-TO-ONE:
Department
ONE-TO-MANY:
Technician
Employees
Equipment
Maintenance
Records
Supervisor
22
An Overview of Database Models
Network Data Model - stores data as
nodes in a network
Department
Employees
Technician
Equipment
Maintenance
Records
23
NETWORK DATA MODEL
Variation Of Hierarchical Model
Useful For Many-to-Many Relationships
Example: Student Class Schedules 
Many Students in many classes
MANY-TO-MANY:
CLASS
1
STUDENT
A
CLASS
2
STUDENT
B
STUDENT
C
24
An Overview of Database Models
Relational Data Model - stores data in
individual files, or tables
 Data In Table Format – each record is an
event with a standard set of event
characteristics
 Relation: Table
 Tuple: Row (Record) In Table
 Field: Column (Attribute) In Table
25
Relational Data Model
26
Comparison of the Models
Network Models - seem to have little
application in health care; some research
applications reported
Hierarchical Models - appropriate where
data form a natural hierarchy; radiology
reporting system; data on individual
patients
Relational Models - emerging as the most
popular and widely used; supports ad hoc
queries
27
Healthcare example of relational
databases
Hospital
Patient ID
Patient Age
DRG
Attending
Physician
ID
Admit Date
Discharge
Date
Discharge Disposition
1234509
42
465
1389
6-05-06
6-10-06
Home
3445676
75
110
3409
5-16-06
6-10-06
SNF
5678932
22
322
6704
6-07-06
6-08-06
Home
7890111
6
201
3422
6-10-06
6-11-06
Transfer
Physician
ID
Physician
Name
Physician
Specialty
Physician
Practice
Name
Physician
Office
Zipcode
Medical Staff
Activation
Date
Physician
Date of Birth
1389
Jones
OB/GYN
Jones PC
63144
11-1996
5-12-1959
3409
Richards
Internal
Medicine
Medical
Specialists
63106
5-1987
2-22-1947
6704
Jackson
Orthopedics
Sports
Medicine
63118
6-2000
4-25-1977
3422
Craig
Pediatrics
Family Health
63105
6-2004
8-16-1980
28
Comparison of the Models
Network
Hierarchical
Relational
Processing Medium
Efficiency
Flexibility Medium
High
Medium
Low
High
User
Low
Friendly
Program
High
Complexity
Low
High
High
Low
29
Database Management System (DBMS)
Software To Create & Maintain Data
Enables Business Applications To
Extract Data
Independent Of Specific Computer
Programs
30
Database Management Systems
Data Definition Language (DDL) - used
to define and describe the data
elements in the database
Data Manipulation Language (DML) used to access, edit, and extract
information from the data contained in
the database
Data Dictionary - used to store a
detailed description of the data
31
Two Views Of Data
 Physical View: Where are Data Physically?
 Software Application or other data source
Drive, Disk, Surface, Track, Sector, Record
 Tape, Block, Record Number
 Logical View: What Data are Needed By
Application?
 Information needed to make the decision(s)
 Name, Type, Length Of Field
 Ability to link or relate to other needed data
elements

32
Data Definition Language
Creates a link between the “User View” of
the data and the Physical View of the data
User defines a schema, or view of the
database
Schema includes file description, record
description, and information about fields
33
Data Manipulation Language
Provides ease of interaction between the
user and the database
Allows user to add new data; sort, delete,
edit, or display data; generate reports
Two methods for interacting:


“Embedded Statements”
Direct Interaction (Query Languages)
34
Query Languages
 Natural Language Queries - English-like
statements
 “Please give me the number, name and
department name of all pieces of
equipment that are associated with the
department having the number 19.”
35
Query Languages

Query-by-Example (QBE) – Microsoft Access
36
Query Languages
Structured Query Language (SQL) developed in the 1970s and adopted as a
standard relational language in 1986
 SELECT EQUIP_NO, EQUIP_NAME,
DEPT_NAME FROM EQTABLE, DEPTABLE
WHERE
DEPTABLE.DEPT_NO=EQTABLE.DEPT_NO
AND DEPT_NO=19
37
Data Dictionary
 File that stores detailed information about the
data elements used in a database -name
type (numeric, alphanumeric, logical ...)
storage allocation
person authorized to change
date of last change
38
Creating A Database
Conceptual Design
Physical Design
39
CREATING A DATABASE
CONCEPTUAL DESIGN:
Abstract Model, Business Perspective
How Will Data Be Grouped?
Relationships Among Elements
Establish End-User Needs
40
Creating A Database
Physical Design:
 Detailed Model By Database Specialists
 Entity-Relationship Diagram (ERD)


Documents the conceptual data model
Normalization - Process of making the
database structure efficient
Physician
Fills Out
Medication
Order
41
Elements Of Database Environment
DATA
ADMINISTRATION
DATABASE
TECHNOLOGY &
MANAGEMENT
DATABASE
MANAGEMENT
SYSTEM
DATA PLANNING
& MODELING
METHODOLOGY
USERS
42
DATABASE TRENDS
Object- Oriented: Data and Procedures
Stored Together; Can be Retrieved,
Shared
Hypermedia: Nodes Contain Text,
Graphics, Sound, Video, Programs;
Organizes Data as Nodes
Linking Databases
Via The Web
43
DATABASE TRENDS
Distributed Databases - a set of “smaller”
databases into which an organization
might choose to store its data.


Benefits include: data are closer to user;
multiple copies exist; data access is more
efficient; applications are more
balanced.
Disadvantages: more complex;
potential for loss of synchronization.
44
3 Final Concepts
 Data Warehouse - enables the collection and
organization of disparate data sources, both
internal and external, to an enterprise
 Clinical Data Repository
 Master Patient Index (MPI)
 Standardization of Terminology & Data Format
 Data Mart
 Meets the demands of specific department(s) or
decision types
45
Standardization Issues
 Health Level-7 (HL7) – a set of standards
designed to develop a cost-effective approach
to system connectivity
 SNOMED (Systematized Nomenclature of
Medical Reference Terminology) – a standard
vocabulary of clinical terms
 EMPI (Enterprise-wide master person index) – a
relational database containing IDs of all
patients seen anywhere in the system
46
DATABASE ADMINISTRATION
 Defines & Organizes Database
Structure And Content
 Develops Security Procedures
 Develops Database
Documentation
 Maintains DBMS
47
Conceptual Model of a Decision-Support System
External
Databases
Financial
Databases
Clinical
Systems
Other
Data Sources
DSS
Database(s
)
DBMS
Model
Library
Model
Manager
User Interface
User
Report
Writer
General Uses of a DSS
and the System to Support Each Use
General Use:






Retrieve Data Item
Perform Ad Hoc
Analysis
Present Aggregated
Data
Determine Impact of a
Proposed Decision
Propose Decisions to
Management
Make Decisions
According to a Rule
Type of System:






Simple DBMS
Generic Statistical
Package
Executive Information
System
DSS with “What-If”
Modeling Capability
DSS with Optimization
Modeling Capability
Expert System; DSS with
Artificial Intelligence 49
Sources of Information for
Decision Support



Internal Transaction Processing Systems
(e.g., ADT systems)
Specially Constructed Databases (e.g.,
medical staff roster)
External Data Sources (e.g., market
demographics data)
50
Storing the Information for
Decision Support

Databases
Example: Lab Data in a Single Hospital
Many Databases throughout the System
Often Represent Disparate Systems
Critical to have a key field to link
disparate databases together, e.g.,
patient ID, procedure code, etc.
51
Storing the Information for
Decision Support

Data Repositories
Example: Clinical Data Drawn from Multiple
Hospitals Related to Day-to-Day Practice
Virtual vs. Physical Repository
BJC Clinical Desk-Top (ClinDesk) System is
a Physical Clinical Data Repository
EMR would be a real-time data repository
52
Storing the Information for
Decision Support

Data Warehouses
More Information than Repository
Used for Retrospective Research
Avoids “Bogging Down” Operating
Information Systems
Often “Interrogated” With Data Mining
Classic example is risk-adjusted patient
discharge data for outcomes analysis
53
Information Needed for
Decision Support






Information to Support Strategic Planning
Information to Support Marketing
Information to Assist in Resource Allocation
Information to Support Enhancement of
Productivity and Operating Efficiency
Information to Support Outcomes
Assessment
Information to Support Contract Negotiation
54
Approaches to Developing a DSS




Write the necessary programs from
scratch in a suitable language
Use tools such as spreadsheets,
database tools, data “cubes,” etc.
Customize a package
Purchase a turnkey system

What are examples of healthcare turnkey
systems?
56
Expert Systems automate the
decision-making process

Expert System: “A system capable of
reproducing “the reasoning process a
human decision maker would go through
in reaching a decision, diagnosing a
problem, or suggesting a course of
action.”
Mallach, E.G. 1994. Understanding Decision Support
Systems and Expert Systems. Burr Ridge, IL.: Irwin
57
Components of an Expert System





Knowledge Base (Rule Base) - contains
expertise of the system
Database - Information which knowledge
base is matched against
Inference Engine - generates conclusions
User Interface - facilitates interaction between
system and user
Workspace - where system stores facts about
a situation
58
Conceptual Model of an Expert System
User
User
Interface
Inference
Engine
Workspace
Knowledge
Base
Database
59
Typical Applications of Expert Systems




Program to find “fraud and abuse” in
insurance claims
Program to support hospital bed
assignment
Program to handle scheduling of
outpatient procedures
Program to check drug interactions or
inappropriate dosages
60
Executive Information Systems (EIS)

Definition - “An information system
which draws from multiple applications
and multiple data sources, internal and
external, to provide executives and other
decision makers with the necessary
information to monitor and analyze the
performance of the organization.”
Hoven, J. van den. 1996. “Executive Support Systems and Decision
Making.” Journal of Systems Management 47(2):48-55.
61
Typical Areas of Interest to the Executive






Financial Performance
Clinical Outcomes
Human Resource Utilization
Access and Continuity
Customer Satisfaction
Market Share
62
Getting the Executive to Use the EIS




Executive Participates in the Design
EIS Must Provide Relevant and
Desired Information
Output of EIS Must be Pleasing with
High-Quality Graphics
System Should be Relatively Easy
to Use
63
Comparison of DSS & EIS




Both Integrate Clinical and Financial
Information from a Variety of Sources
Both Enhance Management Decision
Making
DSS Supports Greater Depth of
Analytical Probing & Modeling
EIS provides performance indicators
at a glance, similar to a scorecard
64
Clinical Decision Support Systems
(CDSS)


Broad Definition: Any automated tool that
helps clinicians improve the delivery or
management of patient care
Ideal Definition: A set of knowledge-based
tools fully integrated with both the
“physician component” of the computerized
patient record and a repository of complete
and accurate clinical data and test results
65
The Two Erroneous Extremes
in DSS Design


Overly Simplistic - System is able to
merely collect and aggregate data rather
than to serve as a DSS
Overly Sophisticated Computing
Technology - Clever “bells and whistles,”
but limited ability to interface with
operational systems and to provide the
executive with needed information
66
The Proper Balance
Executives must look upon the
installation of a DSS in their
organization as one of THEIR
strategic projects, rather than a
necessary
activity
to
be
delegated to the IT staff!
67
Knowledge Management and DSS
DSS contain information to make
managerial decisions.
Data become Information; but to create
Knowledge one needs Experience.
Knowledge Management links user
experience to data and information to
enhance executive decision-making.
68
What is Knowledge
Management?
John Seely Brown:
“Because information is not knowledge, data is not wisdom,
bits are not experience. The difference is us: we make
knowledge out of information together, in our communities
of practice.”
The Social Life of Information
69
Knowledge Management entails the use
of “Online Communities”
What is a
“community”?
Why are they moving
“online”?
Why should this
matter to you?
What is hard about it?
What makes it
easier?
70
Knowledge Management: people connecting
through shared needs
Who else faces similar
challenges to mine?
Who has
expertise in
this area?
Is anyone else working
on this same problem?
What ideas have been
tried and tested?
How can I share what I
have learned?
Is there a recommended
way to do this?
71
Knowledge Management
Lessons Learned from Leading Business Managers
Define the desired performance outcomes – link knowledge
transfer activities contributing to these results – APQC.org
Saved “tens of millions of dollars” by creating a
worldwide repository of “best practices”
$1.5 million in savings from 2 of its communities of practice
More than $1 billion in documented bottom-line savings since 1995
Gained $1.5B in annual wafer manufacturing capacity by sharing “best practices”
$50 million a year in travel cost avoidance and $6 million annually by
finding information more quickly through its KM initiative
Saved over $150M in the first year of an initiative to identify and share marketing best practices
72
What are communities of practice?
 Communities of practice are social networks
 Online communities are supported by web
technologies
 They exist to solve problems
73
What are online communities in
healthcare?
Consumer communities –
Disease support groups, weight loss/stop
smoking, connect patients and families
Professional communities –
Professional societies, physician networking,
hospital business alliances, software users
Employee communities –
Best practice adoption, process improvement
teams, peer networking
74
Online communities are it!
MySpace
Generation
Collaboration
Expectations
Customer
Interactions
Engagement
75
Virtual Communities are used across
hospitals to:
 Create relationships across time and space for
peer learning and experience sharing
 Identify successful practices, lessons
learned, and critical success factors for
achieving better results
 Encourage and reward adoption of innovative
practices and data-driven business processes
76
Virtual Communities are used within hospitals
to:
 Collaborate across roles, departments, or
functions to solve operational problems
 Simplify access to experts and expertise,
encourage new ideas
 Save time and money by re-using work done
in another department or area
77
What are key building blocks for
effective communities?
 Value Add – both the individual and the
organization have to see it as useful
 Culture – a new way of working – inherently
more open and collaborative
 Infrastructure – making it look simple is hard
work to begin with
 Communicating Impact – stories drive change
IHI Profiles in Improvement
Who's improving health care? People are.
Listen to the story of Jennifer Dunscomb of
Columbus Regional Hospital.
Source: www.IHI.org
78
Examples of Knowledge Management
Systems in Healthcare
CHCA
3 pronged approach to knowledge management:
Peer Networking Forums, Performance
Improvement Collaboratives, Race for Results
Awards Program
CHI
Embedded knowledge transfer and learning:
Knowledge Communities, Practice in Action,
Calls to Learn, Relay Reports, LEARN
79
CHCA Case Study
Background of CHCA
Overview of Forums, Collaboratives,
Race for Results
Strategic Impact to date
Lessons Learned
80
Improving the Performance of Children’s Hospitals
Knowledge Transfer to Improve Performance:
A Case Study
 42 non-competing hospitals US, Canada
 $14 billion combined revenue (1)
 Average per member revenue of $330 million
 If Fortune 500 would be ranked 142
 IDN influence:
 500,000 inpatients; 10 million outpatients (2)
 102,000 employees (2)
 >20,000 pediatric physicians (5,162 medical specialists;1,985 surgical
specialists(2)
 Top 5 among U.S. health systems/IDNs
Sources: (1) Estimated from Goldman Sachs report to CHCA, July 2004; (2) Estimated from personnel
report in AHA Guide 2003/ 2004
81
CHCA’s 3-pronged strategy
Peer Networking
Performance
Improvement
Spread
 Online communities
 Peer group meetings
 Collaboratives
PEOPLE &
PROCESS
 Teleconferences
 List serves
 Forum directors
 Special reports
 Benchmarking
 PDSA approach
 Results reported to
peers and executives
 Dedicated PI staff
 Awards process with
external judges
 Peer reviewed
publication
 Ambassador program
 External published
results
 Real time tools and
resources
STRATEGIC
IMPACT
 Individual employee
improvement in
productivity
 Satisfaction + individual
hospital improvement in
results
 Organization-wide
improvement, e.g., cost
reduction, error
reduction, safety
improvement
 Accelerate improvement
 Safe, efficient and
effective
 Focus on spread
 Knowledge available
when you need it
 Best practices
 Peer assistance
TECHNOLOGY
C,
C, c
 RACE for Results
 Juried annual award
82
1 - Peer Networking Forums
 Internet site for Forum members only
 Exclusivity, confidentiality, knowledge of colleagues
 Dedicated staff facilitator – Supports 3-5 Forums
depending on content knowledge and required expertise
 Share documents, post weblinks, initiate discussions,
find resources
 Technology combined with meetings keeps the
group connected
 Teleconferences, webcasts, bi-annual meetings
 Ad hoc conversations, focused research, group
problem-solving
 Rapid response to posted questions
 Benchmarking and identifying variation
83
Peer Networking Forums are Highly Active
2006 Hospital Participation in Forums
Ambulatory
22
Materials Management
33
Cardiac
28
OR Directors
31
CFO
40
PACT
34
CHAPs
17
Patient Financial
Services
21
CIO
36
Payor Contracting
33
CNO
40
Pediatric Practice
Exec.
22
COO
40
Pharmacy Buyers
40
Corporate Compliance
28
Pharmacy Directors
39
Customer Service
20
PHIS
37
Dietary
33
Physician Relations
22
Executive Dialogue
40
Quality and Safety
Leaders
42
Facilities Management
33
Radiology Directors
33
Health Information
Mgmt
33
Respiratory Directors
32
Home Care
17
Risk Managers
25
Human Resources
32
SMAC
30
JCAHO
35
Social Work
Community
15
Lab Directors
32
SPBD
28
Overall
2006
satisfaction
5.24 of 6.0
(87%)
84
Peer Networking Forums webpage example
85
2 - Performance Improvement Collaboratives
 Dedicated Performance Improvement staff and
resources
 Trained in IHI improvement methodology
 Hospitals agree to share results, post data and publish results
 Use industry and hospital expert panels to validate clinical
direction
 Combine research and rapid cycle - essential for academic
engagement
 Technology tools and partners integral to success
 Knowledge repository available real time
 improvements, tool kits, lessons learned, comparative data, audios of
webcasts and lessons learned
 Strategic partners essential to spreading results and gaining
credibility
 AHRQ Partnership for Quality Grant helped fund participation and
training for all 42 hospitals
 Data-sharing agreements developed to expand comparative data sets
86
(Vermont Oxford Neonatal Network and others)
Performance Improvement Collaboratives
Example: Reduced Adverse Drug Events
 16 teams (89%) had a reduction in ADE rate
 Average among teams with a reduction:
64% reduction
 Average for all teams: 49% reduction
 11 teams (61%) had at least a 50% reduction in
ADE rate
0%
-40%
Goal
-60%
Avg.
-80%
CHCA Hospitals with reduction in ADE rate
o
m
N
ew bus
Yo
rk
Pi
tts B
bu
rg
h
Pa
lo
Al
to
Ft
W
C
or
or
pu
th
s
C
h
N
ew rist
i
O
rle
an
s
At
l
a
Ka
nt
a
ns
as
Ci
ty
N
as
hv
ille
C
in
c
in
St
Pe nat
i
te
rs
bu
Bi
rg
rm
in
gh
am
D
ay
to
n
ng
e
ffa
l
ol
u
C
Bu
O
ra
er
ag
e
-100%
87
BETTER
-20%
Av
Atlanta
Birmingham
Buffalo
Cincinnati
Columbus
Corpus Christi
Dayton
Fort Worth
Kansas City
Miami
Nashville
New Orleans
New York/
Morgan-Stanley
New York/
Komansky Center
Orange
Palo Alto
Pittsburgh
St. Petersburg
% change
Hospital Teams:
Collaboratives have own web-sharing spaces
88
Adverse Drug Event Collaborative webpage
89
Blood Stream Infection Collaborative webpage
90
Surgical Infection Prevention Collaborative
91
3 - Awards Program Encourages Spread
 Formal RACE for Results awards program
 Formal application process with strict submission requirements
 External judges panel representing industry experts in quality and
patient safety
 Results announced at award ceremony during annual Quality
& Safety Meeting
 Winners required to serve as Ambassadors during subsequent year
to teach techniques and encourage adoption of proven practices
 Formal marketing campaign to publicize event
 Emails, posters, web notices to promote the competition and
publicize winners
 Email-based Relay Report to report progress as proven practices
are replicated across the alliance
 Resources and contacts posted on the intranet to facilitate
connections and encourage adoption
 Benchmarking reports regularly published to document
improvements
 Improve Today Webcasts connect colleagues
92
RACE for Results
Awards Program
2004
2005
2006
Little Rock:
Reducing
Catheter-Related
Bloodstream
Infections through
Repeated Rapid
Cycle
Improvements
Palo Alto:
Decreasing ADEs
By Implementing
Safety Best
Practices
Omaha: "Asthma
Attack“
Cincinnati:
Reducing Cost
through Improving
Quality
11 Entries
Washington DC:
Using PHIS to
Target Reducing
Infections in VP
Shunt Surgeries
12 Entries
2007
Dayton:
Reducing
CatheterAssociated
Bloodstream
Infections in
Children
17 Entries
30 Entries 93
RACE Results in Performance Improvement
94
Conclusions for CHCA
Strategy Drives Approach
 Informal peer networking builds a culture of sharing and collaboration
 Formal collaboratives are needed to create immediate results
 Systematic rewards and support are needed to spread initial results
Knowledge Transfer involves Technology, People/Process, and Strategy
 Technology enables information sharing and people directories
 People processes ensure productive interaction and knowledge exchange
 Strategy determines impact measures and ensures organizational
momentum
CHCA Case Study Results:
 42 children’s hospitals participate in 30 peer networking forums, regularly
sharing improvement tools and resources, exchanging best practices and
learning from industry experts
 18 children’s hospitals averted 13,478 adverse drug events (ADEs),
representing $2.7 million
in net savings, and reduced PICU blood stream infections (BSIs) by 57%
 More than 60 intensive care units are working to sustain and spread
improvements in ADEs and BSIs based on the initial collaboratives’ work
95
CHI Case Study
Background of CHI
KT&L Strategy and Scope
Relay Report results to date
Lessons Learned
96

CHI Fast Facts
u

u
u
 Multi-institutional System of
Catholic Healthcare Providers
 Dedicated to the healing ministry of the Catholic Church
 National Offices: Denver, Northern KY, and Minneapolis
 Market Based Organizations (MBOs)
 19 states
 68 rural and urban communities
 71 hospitals (63 acute care, 5 behavioral, 2 rehabilitation, 1 long
term acute care)
 43 long-term care, assisted living facilities and residential units
 5 Community Health Services Organizations
 Licensed acute care beds range from 15 to 1,546
 $7.1 Billion in Annual Revenues
 66,000 Employees (and growing)
97
Vision for CHI
Catholic Health Initiatives’ Vision is to live
out its Mission by transforming health care
delivery and by creating new ministries for
the promotion of healthy communities.
98
CHI Strategic Plan: 2007 - 2011
99
Leveraging the Knowledge Within
“Our goal is for CHI to become known as
an innovative organization. That will be
our legacy for the future health care
system – that CHI learns to leverage the
wisdom of the whole, efficiently,
effectively, and humanely.”
- Kevin E. Lofton, FACHE, CEO, Catholic Health
Initiatives
100
Knowledge Leadership
Knowledge Leaders are Leaders who are effective at…
 Embracing and driving change
 Sharing experiences and applying learning
 Modeling the expected behaviors grounded in the culture
of the organization
… in order to tap into the intellectual capital of the
organization and harness it to innovate and grow
101
Knowledge Transfer & Learning at CHI
Knowledge
Communities
Communication
& Collaboration
Strategic
Priority
Consulting
Formal
Education
Practice in
Action
102
CHI’s KT&L Strategy
Communication
& Collaboration
Knowledge
Transfer
Integrated
Learning
TECHNOLOGY
 Knowledge Communities
 Relay Report
 Live Meeting webconferencing
 List serves
 Practice in Action - Proven
Practices database
 Pathfinders collection of
expert resources
 Calls to Learn
 Annual Events &
Conferences
 System-wide LMS LEARN
PEOPLE &
PROCESS




Calls to Learn
National office sponsors
KC Chartering process
KC metrics reports
 Integrated into Annual
Planning Budget Review
process
 Formal process to confirm a
proven practice
 KT&L staff support systemwide initiatives e.g., CHI
Connect, service line
development
 Centralized calendar of
Calls to Learn across the
organization and beyond
 LMS intended to address
CHI-wide practices
 Integration of
organizational
effectiveness research with
delivery of new education
STRATEGIC
IMPACT
 Enable innovation
 Focus on strategic
priorities
 Defined and implemented
new standards of practice
 Organization-wide
improvement, e.g., cost
reduction, error reduction,
safety improvement
 Accelerate improvement
 Safe, efficient and effective
 Strategic Priority Consulting
 Compliance adherence
 Enable delivery of
education in support of
strategic priorities across
the system
103
Knowledge Communities: Collaborate
and Innovate
An environment that enables innovation, supports the
development and spread of new ideas and builds the
organizational social network to save time and reduce costs.
Value of Pharmacist
as part of bedside
patient care team proved
and $53 Million saved
104
Why CHI uses online communities
 Connect peers and experts across CHI
 Common space across distance and time
 Supports the work of the Knowledge
Community:
Enable and leverage knowledge sharing
Learn before doing
Problem solving: find, innovate, and accelerate solutions
Reduce costs, save time, and increase social fabric
A strategic resource
105
Making it Easier to Connect with
Knowledge Communities
106
Relay Report: Communicate, Connect,
Celebrate
Improve connectivity, celebrate successes and increase
awareness and utilization of KT&L resources.
Accelerate the
implementation of
clinical imaging
technology, resulting
in accelerated
NPSR of $1.5 - $3.0 M
107
Practice in Action:
Transfer Critical Knowledge
Increase adoption of reliable, evidence based practices, identify
organizational expertise, and recognize facilities that have
achieved success.
Avoided medication
errors through
improved reconciliation
of home and hospital
medications.
108
Strategic Priority Consulting: Support
Organizational Priorities
Accelerate achievement of strategic priorities by creating a plan
to leverage available knowledge transfer and learning
resources.
Accelerate the
implementation of
ERP technology
(Lawson) and the
realization of
projected savings
109
Formal Education: Sustain Change
Coordinate and share system resources to insure that education
and training help employees learn the skills, behaviors and
competencies they need to move strategic priorities forward.
Avoid dangerous /
deadly events in OB
through a targeted
Advanced Fetal
Monitoring curriculum
110
Knowledge Transfer is both Organic AND
Strategic
Accelerate Learning & Enable Innovation through…
Knowledge
Communities
Communication
& Collaboration
Strategic
Priority
Consulting
…leading to new models of
care delivery and creative
solutions…
…resulting in improved
outcomes:
• Quality & Patient Safety
Formal
Education
Practice
in Action
• Employee Satisfaction
& Engagement
• Increased Operating
Margins
111
Conclusions for CHI
Knowledge Leadership is at the core of CHI’s business strategy
 New Leadership competencies are based on collaboration & change
 Knowledge communities build a culture of sharing and innovation
 KT&L team has become core resource for national strategic initiatives
Knowledge Transfer has become the way CHI works
 Web tools support connectivity and facilitate communicating about key
knowledge resources and success stories
 New roles have evolved as collaboration has become embedded in the
way CHI works
 Strategic initiatives rely on knowledge tools to speed adoption
CHI Case Study Results:
 A total of 48 knowledge communities involve over 1600 associates across
the system
 Knowledge communities have yielded both “hard” and “soft” dollar
savings, impact patient outcomes through improved practices, increase
reuse of proven practices
 After 5+ years, CHI leadership expect KT&L resources to be utilized as
part of strategic priority projects and leaders will be held accountable to
Knowledge Leadership competencies
112
Communities Are Here!
A 2001 best practice study (Using Communities of Practice to
Drive Organizational Performance and Innovation) found:
“…strong evidence that communities are the next step in the
evolution of the modern, knowledge-based organization.
Communities … are a legitimate way to spend time, engage
an amazing percent of employees, are held accountable for
producing and stewarding business-critical knowledge (and
often results), and are assuming a formal voice in the
organization, based on the power of their knowledge, not
their position.”
113
Knowledge Management Building Blocks
in Healthcare
Peer
Networking
Performance
Improvement
Spread
TECHNOLOGY
Create MySpace for
your employees,
physicians, and
customers
Create public
campaigns for targeted
improvement goals
Publish results on the
hospital website –
customize for each
audience
PEOPLE &
PROCESS
Create online people
directories, create peer
group moderator roles,
highlight personal
success stories
Develop a dedicated PI
staff – this may
incorporate Six Sigma,
IHI Collaboratives, etc
– or may be internally
developed
Incorporate proven
practice sharing into
annual awards
ceremonies, dept
budget reviews,
employee performance
reviews
STRATEGIC
IMPACT
Enhanced employee
satisfaction and
productivity, strong
customer satisfaction
scores
Focused improvement
in targeted areas, e.g.,
patient safety, financial
performance, wait
times, turnover, etc.
Faster decisions,
quicker adoption of
proven practices, rapid
innovation absorption
114
Key Topics
 Decision Support – analyzing data, often from
different sources, to make better decisions
 Decision Support Systems (DSS) automate
decision support
 Executive Information Systems (EIS) quickly
assess performance and trends
 Clinical Decision Support Systems (CDSS)
enhance patient care decision-making
 Knowledge Management (KM) incorporates
evidence- and experience-based information
115