Transcript Best Practices for Clinical Transformation: Partners HealthCare

Knowledge Management Challenges
in the Healthcare Delivery Market
Tonya Hongsermeier, MD, MBA
Corporate Manager, Clinical Decision Support and Knowledge Management,
Clinical Informatics Research & Development
Partners HeatlhCare System, Inc.
Agenda


About Partners Healthcare
Knowledge Management and Informatics
 Knowledge
 Knowledge
 Knowledge

Application
Discovery
Asset Management
Challenges in Healthcare Delivery
 Weak
Organizational Alignment
 Weak Investment in Asset Management
 Implications for Clinical R&D
 Implications for Personalized Medicine
Partners HealthCare







Massachusetts General Hospital, Brigham and
Women’s Hospital and several other hospitals in
the network
Licensed Beds
3196
Births
18,478
Admissions
134,991
Patient Days
871,321
Average LOS
5.31
Total Outpatient Visits
2,324,073
Partners Information Systems







Much published on innovative use of informatics
in healthcare (Bates, Teich, Glaser, Kuperman,
Barnett, Chueh, and many others)
800 applications
520 active projects
680 employees based in 19 locations
FY02 operating budget of $92.3M
FY02 capital budget of $47M
These are relatively generous numbers as a
percentage of operating expenses
Some Current Clinical Knowledge
Assets Developed at Partners








Medication Data Dictionary and DDIs
Inpatient alerts and interactive order rules
Gerios and Nephros for proactive filtering of
drug doses for elderly and/or renal insufficient
Radiology Ordering decision support
Preventive health reminders
Outpatient lab result decision support
Outpatient documentation templates
Piloting outpatient drug-lab, drug-disease
interactive reminders
Current State Challenges Typical of Many
Academic Healthcare Delivery Organizations






7 homegrown and 2 commercial CPOE systems,
plan to evolve to “next generation CPOE” in next
5 years
Limited implementation of structured (encoded)
clinical documentation
Proprietary approaches to knowledge encoding
Not re-usable or sharable
Much updating/maintenance is bottlenecked by
resource constraints
Research datawarehouse in place, but struggle
to expand in face of fragmented clinical systems
environment
Typical Committee and Project Structures
Related to Medication Safety Illustrate
Organizational Alignment Problem
Information Technology Projects





Physician Order Entry
Team
Clinical Data
Repository Team
Pharmacy System
Team
Clinical Documentation
Team
Electronic Medication
Adminstration Team
Committees and Departments






Pharmacy and
Therapeutics
Patient Safety
Quality or Performance
Improvement
Policies and
Procedures
Formulary
Infection Control
Medication Use Process Organization
Medication Safety Steering Committee
Chief Medical Officer, Chief Nursing Officer, Chief Information Officer, Chief Quality Officer
Interdisciplinary Medication Use Process Advisory Team
Physicians, Nurses, Pharmacists, Clinical Systems Architects
Information Technology Projects





Physician Order Entry
Team
Clinical Data Repository
Team
Pharmacy System Team
Clinical Documentation
Team
Electronic Medication
Administration Team
Committees and Departments






Pharmacy and
Therapeutics
Patient Safety
Quality or Performance
Improvement
Policies and Procedures
Formulary
Infection Control
Knowledge Management: The Core Processes
Knowledge
Application
Knowledge
Asset Management
Knowledge
Discovery
A Continuum of Clinical Decision Support
and Knowledge Discovery*
 Surveillance  Interactive
Reference
Knowledge
Linking
Event
Monitori
ng
Safety
Net
 Proactive
Anticipati
on
 Learning
Understan
ding and
Predicting
Performan
ce
 Monitoring
patient
data with
passive
decision
support
 Intercepting  Making the
right
incorrect
decisions
clinical
the easiest
decisions
decisions
*modified from the First Consulting Group Model of Clinical Decision Support
 Predictive
Modeling
 Casebased
Reasoning
 Learning
Knowledge
Repository
Medication Decision Support Categories at Partners
REFERENCE INFORMATION

Drug-information knowledge linking via
info button adjacent to drug name

Partners handbook provides access to
numerous drug information databases

Planned drug-information knowledge
linking via info button in electronic
medication administration record in FY
04
SURVEILLANCE AND MONITORING

Drug-induced abnormal lab result
notification of physician

Drug-induced abnormal lab result
notification of pharmacist

Renal function decline in patient on
renally excreted drug notification of
physician and pharmacist
INTERACTIVE DECISION SUPPORT FOR
PHYSICIAN AND PHARMACIST:

Drug-allergy checking

Drug-drug interaction checking

Drug-food interaction checking

Drug-herb interaction checking

Drug-disease interaction checking
INTERACTIVE DECISION SUPPORT FOR
PHYSICIAN ORDER ENTRY ONLY:

Drug-lab interaction checking

Consequent order recommendations

Relevant lab display

Indication-required orders

Height, weight, allergy update required
notification

Dose calculation tools

Intravenous to oral conversion
recommendation on renewal of intravenous
order when patient receiving other oral
medications

Formulary substitution alerts

Antibiotic restriction alerts
PROACTIVE DECISION SUPPORT

Gerios for elderly patient medication dosing

Nephros for dosing in renal insufficiency

Preventive health reminders

Problem-linked order sets
Laboratory Notification with consequent
order recommendations
Alternate Procedures, Redirects, DrugAllergy, Drug-Drug, Drug-Lab etc.
Gerios: Dose-filters for age
Nephros: Dose-filters for renal function
Inappropriately sedated elderly inpatients on average incur $5600 excess
costs over expected for severity of illness
Preventive Reminders
Problem-level anticipatory
decision support



Today, order sets and documentation templates
are static which means that clinician must
change them to personalize them to patient
We plan to use inferencing to dynamically
generate problem-driven order sets and
documentation templates that account for
multiple co-morbidities
Must be able to incorporate future onslaught of
gene diagnostic and prognostic data
Knowledge Application must anticipate
these dimensions of the clinical encounter
Clinical Standardization
Standards of Practice,
Role/Venue Requirements
Billing/Regulatory Requirements
Improvisation
User Personalization
Patient Preferences
End-user workflow preferences
Learning and User-defined
Poly-hierarchical inferencing with
actionable advice – surveillance,
This is an example from clinical decision support company called Theradoc
A Continuum of Clinical Decision Support
and Knowledge Discovery
 Surveillance  Interactive
Reference
Knowledge
Linking
Event
Monitori
ng
Safety
Net
 Proactive
Anticipati
on
 Learning
Understan
ding and
Predicting
Performan
ce
 Monitoring
patient
data with
passive
decision
support
 Intercepting  Making the
right
incorrect
decisions
clinical
the easiest
decisions
decisions
*modified from the First Consulting Group Model of Clinical Decision Support
 Predictive
Modeling
 Casebased
Reasoning
 Learning
Knowledge
Repository
Current Initiatives




Quality data warehouse with Clinician
Dashboards
Early identification of patients at risk for case
management
Longer term knowledge discovery goals to use
performance data to enhance knowledge
repository
Need to evolve towards non-human dependent
modes of knowledge acquisition
Knowledge Asset Management Infrastructure:







Analysis of clinical performance data to understand
where knowledge deficits are to support performance
goals
Authoring and support of virtual, asynchronous
collaborative authoring by knowledge editors and
leaders of research, safety and quality improvement
initiatives (reference knowledge specs for
encoding
Knowledge acquisition from commercial/etc
knowledge bases
Validation and audit trail maintenance (metaknowledge)
Inventory (knowledge librarian)
Publishing and Sharing
Reference information and knowledge model
What are the challenges today:




Healthcare delivery organizations purchase
systems but don’t invest in knowledge asset
management, they install plumbing
Vendors sell knowledge editors, not knowledge
management support infrastructure
There is no repository of “best clinical IT
practices” at a national level, few among the
vendors
No knowledge encoding and representation
standards to facilitate knowledge sharing
Partners-Wide Knowledge Management Model
KNOWLEDGE ASSET MANAGEMENT
Signature Initiatives and
Sub-Committees set
Enterprise-wide Strategy,
Clinical Standards and
Performance Measures
Decision Support Design Teams
direct the design of crossfunctional knowledge to be
encoded
Performance Feedback to Leaders,
SMEs, Committees, and End-users
Subject Matter Expert (SME) Panels
Advise on Entity, Venue, Role, Specialty,
Primary Care, Disease Management, and
Safety related requirements for
application function and knowledge bases
Partners Genetics
Computing Platform
Applications for Virtual Collaborative
Knowledge Authoring and Maintenance
Knowledge
Repository
Knowledge
Building Blocks
Common Services
Knowledge Editors
Information
Model
Clinical Workflow
Applications and
Services
DECISION SUPPORT
(APPLIED KNOWLEDGE)
Data Warehouse
PERFORMANCE
and OUTCOMES
(KNOWLEDGE
DISCOVERY)
Knowledge Asset Management:
Translating Goals into a Knowledge Repository Taxonomy
Goal Framework: Safety, Quality, Efficiency, Research
Data/Knowledge
Seeking
Assessment
Dx/Rx
Decision Making
Order Fulfillment,
Communication and
Applications
Coordination
Billing
Reporting
Transfer/
Handoff
Care
(Results, Observations, Orders, Tasks/Proc/Mar,Messaging, CDS,
CORE CARE PROCESS
AUTOMATION TAXONOMY
Measurement)
and Knowledge
Bases
Medical Management,
Research,
and Reporting
Care Applications
(Results, Observations, Orders, Tasks/Proc/Mar,Messaging, CDS,
Clinical Knowledge forMeasurement)
Personalized Medicine Taxonomy
and Knowledge Bases
Role and Venue Domain Taxonomy
Requirements
Care Applications
and Knowledge Bases
MEDICATION USE PROCESS: Acetaminophen in a 2.5 Kg Premature Infant
Patient Venue
Clinical Domain
Role-based
workflow
Informatics
Support
Neonatal Intensive Care Unit
Fever Management
Physician Orders Pharmacist
Dispenses
Physician Order Pharmacy
Entry with
Verification and
Pediatric dosing Dispensing
calculation engine System
Nurse Administers Nurse Assesses
Electronic
Clinical
Medication
Documentation
Administration
Record with
Calculation Engine
Clinical Data Repository with longitudinal patient information such as
gestational age, height, weight, allergies, active and historical orders,
assessment data, laboratory data, and other diagnostic data
Medication
Knowledge
Acetaminophen, Infant
Acetaminophen,
10 mg/kg PO Q 6 Acetaminophen 0.25 ml PO Q 6
hrs
100mg/ml solution hours
Desired
Outcome is
fever abated
Sample High-level Example Taxonomy for Knowledge Assets
CORE CLINICAL PROCESSES
Admission
Patient Scheduling and Tracking
Clinical Data Review
Clinical Knowledge Retrieval
Patient Assessment
Diagnostic/Therapeutic Ordering
Order Fulfillment
Results Reporting
Billing
Patient Transfer/Hand-off
Care Oversight
Medical Management
Population Management
VENUES
Medical-Surgical
Coronary Care Unit
Surgical Intensive Care Unit
Medical Intensive Care Unit
Post-operative Care Unit
Thoracic Care Unit
Emergency Department
Operating Room
Cardiac Catheterization Lab
Rehabillitation Ward
Subacute Care Unit
Psychiatric Care Unit
Ambulatory Care
Home Care
Self Care
MODE OF INFORMATICS
APPLICATION
Results Review
Notifications
Interactive Decision Support
Reminders
Clinical Messaging
Order Sets
Clinical Documentation
Vital Signs, Intake and Output
Reports
Dashboards
WORKFLOWS
Physicians
Nurses
Pharmacists
Medical Technicians
Case Managers
Social Workers
Respiratory Therapists
Occupational Therapists
Physical Therapists
Speech Therapists
Chart Abstracter
Risk Manager
Administrative Support
Quality Improvement
Patient/Consumer
CLINICAL DISCIPLINES
Surgical Service Lines
Cardiac (surgical)
Cardiac (interventional)
General Surgery
Gynecology
Orthopedic
Transplant
Trauma
Urology/ Renal (Genitourinary)
Vascular Surgery
Pediatric Surgery
Colorectal Surgery
Neurosurgery
Non-Surgical Service Lines
Cardiac (medical)
Endocrinology
Gastroenterology
General Medicine
Hematology
Oncology
Newborn/ Neonatology
Neurology
Infectitious Disease
Respiratory/ Pulmonary
Rheumatology/Immunology
Behavioral Health
Patient Safety
Medication Safety
anticoagulation management
sedation management
renal/electrolyte management
elderly medication management
pediatric medication management
Disease or Risk Management
Diabetes
Congestive Heart Failure
Hypertension
HIV
Immune Compromised
Coronary Artery Disease
Hyperlipidemia
Elderly/Geriatric Status
Dementia or Alzheimer's Disease
Transition Management
Multiple Sclerosis
Rheumatoid Arthritis
Nosocomial Infection Prevention
Antibiotic Management
Wound and Catheter Management
Ventilator Management
Other Bedside Safety
Falls Prevention
Decubitus Ulcer Prevention
Restraints Management
Center for Clinical Knowledge Engineering
Welcome to the National Knowledge Engineering Repository
Go
Content search HEDIS
Advanced Search Filters (press Ctrl to select more than one):
Clinical Discipline: Surgical
Cardiothoracic Surgery
Interventional Cardiology
Orthopedics
Etc.
Clinical Discipline: Non-Surgical
Cardiovascular
Endocrinology
Gastroenterology
Etc.
Clinical Discipline: Safety
Nosocomial Infection Control
Medication Safety
Decubitus Ulcer Prevention
Etc.
Clinical Discipline: Disease Management
Diabetes Mellitus
Congestive Heart Failure
Multiple Sclerosis
Etc.
Informatics Mode
Interactive Rules
Surveillance and Notifications
Documentation Templates
Etc.
Age
Adult
Pediatrics
Neonate
Etc.
Role
Nurse
Physician
Case Manager
Etc.
Venue
CCU
Ambulatory Care
Emergency Department
Etc.
Search File Hierarchy
Knowledge Asset Management Toolkit
Link to references, survey instruments, diagrams,
descriptions, process flow diagrams, etc on
Partners and VA approaches to asset management
Submit Content to Editor
About Us
ABRIDGED SAMPLE KNOWLEDGE SPECIFICATION
INVENTORY INFORMATION:
Knowledge Specifications
For Encoded Knowledge
Vs
Meta-knowledge about
The knowledge
TITLE:
INFORMATICS MODE:
FILE NAME:
DATE LAST UPDATED:
DATE FIRST ACTIVATED:
AUTHORS:
CONTACT:
VERSION ID:
SITE:
KNOWLEDGE EDITOR:
STATUS:
Digoxin Order Abnormal Lab Result
INTERACTIVE CPOE
BICS_DIGORDER_LAB
12/03/2003
03/02/2001
John Smith, Jane Doe
[email protected]
001.005.003
Safe Hospital
Joe Brown
IN PRODUCTION
PURPOSE:
Warn of potential digoxin toxicity when ordering digoxin in a patient w/
hypokalemia which predisposes patient to an adverse drug reaction
EXPLANATION:
Alert Evokes when digoxin is ordered. Recent chemistry values are
checked for hypokalemia, Recent orders are also checked for current
potassium supplements which would indicate the electrolyte problem is
already being addressed.
KEY WORDS:
Digoxin, Potassium, K, Dig
REFERENCES: Links to Articles
DATA ON IMPACT: Links to Reports on Usability and Clinical Impact
SPECIFICATION INFORMATION
EVOKE: on DIGOXIN ORDER
LOGIC: DIGOXIN ORDER and MOST RECENT POTASSIUM RESULT
<3.4 WITHIN 30 DAYS and ACTIVE ORDERS DO NOT INCLUDE
POTASSIUM CHLORIDE; POTASSIUM PHOSPHATE etc…
ACTION: SEND NOTIFICATION
SUPPLEMENTATION etc….
SCREEN SHOTS: Link
PROCESS FLOW DIAGRAM: Link
RECOMMENDING
POTASSIUM
Future State KM Model
Collaborative
Knowledge
Authoring
Tools
Portal
Meta-Knowledge
Repository
Workflow Applications
Knowledge-based Services
Knowledge Repositories
Information Model
Barriers to Success at the Intersection
of Clinical Informatics and KM





Leadership inadequately committed
Products inadequate to support
processes
Business case intangible
Fear of exposure (technology increases
transparency)
Few roadmaps to success are proven in
the healthcare arena
Market Drivers will Propel Progress




Aging population: computer literate and population
growth will outstrip service capacity, informatics must
support self-management
Business community will aid transition from commodity
to value based purchasing by employers and
consumers, they know that the current inflation rate of
the commodity is untenable
Leapfrog and Government are beginning to purchase
quality
Genomics: personalized medicine will require
technologies for personalization, these same
technologies will enable more user-friendly safety
solutions
Where are we?