Transcript Document

Managing Evidence at the Speed of Change
Tonya Hongsermeier, MD, MBA
Corporate Manager,
Clinical Knowledge Management for Decision Support
Clinical Informatics Research and Development,
Partners Healthcare System
Outline
•
New Market Demands for Evidence-based Practice
•
Challenges and Opportunities of Integrating
Evidence-based Decision Support with the
Electronic Health Record
•
The Partners Healthcare Model for Managing
Evidence at the Speed of Change
Current State of Translating
Evidence into Practice
•
•
•
17 year innovation adoption curve from discovery
into accepted standards of practice
Even if a standard is accepted, patients have a
50:50 chance of receiving appropriate care, a 5-10%
probability of incurring a preventable, anticipatable
adverse event
The market is balking at healthcare inflation, past
utilization management measures have not
succeeded
Carolyn Clancy, MD, Dir. AHRQ
Gandhi et al, NEJM 2003;348(16):1556-1564
Gurwitz et al, JAMA 2003;289:1107-16
McGlynn et al, NEJM 2003;348(26):2635-45.
•Employers and consumers are
now paying for performance
•They will not wait for the data to be
right or fair because they believe if
they wait, it never will be
•Instead, they are using the
process of rewarding performance
to force the healthcare providers to
“make the data right”
•Defined Contribution lays even
greater purchasing responsibility at
the door of the consumer
Computer-Based Clinical
Decision Support
•
•
•
•
Not a panacea, doesn’t substitute for organizational
alignment required for bringing evidence to practice,
but when done well….
55-83% decrease in hospital non-intercepted serious
ADEs using CPOE
22-78% increased adherence to preventive health
reminders
At Partners, proprietary Drug-Drug Interaction
checking intercepts 5% of physician orders,
physicians change their decision about 33% of time
Kaushal R, et al. Arch Intern Med. 2003
Bates, JAMA 1998
And Unpublished
The Volume and Velocity
of Knowledge Processing Required
for Care Delivery Grows
•
Medical literature doubling every 19 years
•
2 Million facts needed to practice
•
A typical drug order today, decision support
accounts for, at best, Age, Weight, Height, Labs,
Other Active Meds, Allergies, Diagnoses
•
Already, there are 3000+ molecular diagnostic
tests on the market, genomics and personalized
medicine will increase the speed of change of
evidence exponentially
Covell DG, Uman GC, Manning PR.
Ann Intern Med. 1985 Oct;103(4):596-9
When do you order this test?
How do you use the test result?
Leading the News: Roche Test Promises to Tailor Drugs
to Patients --- Precise Genetic Approach Could Mean
Major Changes In Development, Treatment
June 25, 2003
Roche Holding AG is launching the first gene test able to predict how a person
will react to a large range of commonly prescribed medicines, one of the biggest
forays yet into tailoring drugs to a patient's genetic makeup.
The test is part of an emerging approach to treatment that health experts
expect could lead to big changes in the way drugs are developed, marketed and
prescribed. For all of the advances in medicine, doctors today determine the
best medicine and dose for an ailing patient largely by trial and error. The
fast-growing field of "personalized" medicine hopes to remove such risks and
alter the pharmaceutical industry's more one-size-fits-all approach in making
and selling drugs.
When Knowledge Changes…
How quickly can
you change the
content of your
rules, order sets,
templates, and
reports?
Clinical Decision Support Challenges
•
How do we supply meaningful decision support that
both improves quality of care for patients and
quality of life for clinicians (and self-managing
patients)?
•
How do we affordably develop, acquire and
maintain the knowledge bases required to deliver
meaningful decision support?
Three Facets of
Clinical Decision Support
Quality Performance
Evidence, Safety
Process Requirements
Regulatory Requirements
Improvisation
User Quality of Life
Patient Preferences
End-user role, workflow preferences
Typical Vendor System Challenges
•
Task-interfering approach to decision support
- Siskel and Ebert Model
•
Editors don’t support “content vetting” process or the
development of disease management approaches
•
Labor of developing and maintaining decision
support knowledge is vastly underestimated
•
Consequently, clinical systems implementations are
under-resourced with adequate knowledge to meet
workflow and quality needs
New Opportunities
•
Content suppliers are emerging for decision
support content:
•
•
Primarily order sets and some rules
Zynx, Wolters-Kluwer, Micromedex
•
New systems are available to support the vetting
process without required attendance at committee
meetings
•
New systems are available that make it possible to
manage knowledge at the “goal level”
•
Manage Diabetes or Coronary Artery Disease related
knowledge rather than managing just rules or order sets or
documentation templates
Where are we?
Partners has a long track-record in
applied clinical decision support
•
•
•
•
•
•
•
•
•
Physician-oriented Drug-Drug Interaction Checking
Inpatient interactive order rules and notification alerts
Inpatient templated orders (hundreds)
Proactive Dosing support for Geriatric, Pediatric, Neonatal,
Renally-Impaired, and Heme-Onc populations
Radiology Ordering decision support
Preventive health reminders
Documentation templates
Disease management reports/dashboards (Diabetes, others to
follow)
Outpatient drug-lab, drug-disease interactive reminders
Heterogeneous Infrastructure
•
•
•
5 internally developed POE applications (inpatient
and outpatient)
3 versions of Meditech and Siemens POE
Despite this heterogeneity:
•
•
•
•
Many sites share common expert dosing services
Many sites share a common allergy repository
Moving toward common problem list and medication list
over the next 2 years
Partners is very committed to maintaining highquality common logic to support Evidence-based
Practices
Knowledge Management
Strategic Goals
•
Reduce the cost and increase the speed of translation of
clinical innovation and evidence into clinical practice
•
Proactive, anticipatory decision support -- avoid
“interruptive” decision support
•
Improve Partners’ organizational effectiveness as a
learning organization through data-driven performance
improvement
•
Align knowledge assets with business, regulatory, safety
and quality requirements, Only build what we cannot
buy
•
Partners has created some of the best decision support
in production in the world, the goal here is to keep the
knowledge up to date
Knowledge Life-Cycle Challenges:
Committee, Department,
Researcher, or Other
Proposes to Implement Content
•Governance and Stewardship
Guideline is Defined and Validated
Functional Knowledge Specification
For Encoding is
Designed and Validated
Specification is
Engineered into Production Generating
a Technical Specification
Ongoing Revisions or
Eventual Sunset
Of Encoded Guideline
•Inadequate tools and personnel to support vetting and update
of knowledge
•Lack of transparency of knowledge already in production
•Bottlenecks the time it takes to agree on content
•Project and resource competition with other engineering
projects, prioritization processes unclear
•Editors inadequate
•Lack of access to analytic data available on decision support
content or
impact on clinical outcomes impact to direct updating
•No content management tools to support process and ensure
timeliness
Tactics we have deployed:
Transparency and Governance (2004):
•
Build and deploy a document library to provide enterprise wide
•
access to information about of decision support knowledge in
production
Worked with existing committees to evolve more effective
governance models for content
Collaboration and Content Life-cycle Tools (2005):
•
•
Collaboration Portals aligned with clinical goals of Partners
Content Management infrastructure to support content
management processes using lifecycles and workflows
(knowledge maintenance)
Content Editing Re-architecture (2006-7):
•
Reduce the editing bottlenecks
•
Once decision is made for knowledge to change, change will be
implemented rapidly
Clinical Content Committee:
Prioritizes and Sponsors Operational Stewardship of Content
Safety
SME
Groups
Quality
Disease Management
Primary Care
Disease Areas
Adult, Geriatrics,
Pediatrics,
Women’s Health
CAD/CHF, Diabetes,
Heme-Onc, Asthma,
ID/HIV, Nephrology,
Psych
Trend Management
Pharmacotherapy
Medication
Knowledge
Committee
Knowledge Analysts facilitate
Knowledge Editors update
Production Knowledge Repositories
PCHI
P&T
Imaging Studies
MGH
ROE
BWH
Precipio
The Knowledge Management Portal:
Can aggregate all content for Coronary
Artery Disease Management
Compare Content Across Organizations
Geriatric Dosing:
Case Example of Collaborative Content
Development using Documentum E-Room
•
•
•
•
•
•
New pay-for-performance incentive for Geriatric Prescribing
at Partners
Gerios in production in one inpatient order entry system for
several years, must deploy across the enterprise
Content review showed gaps (missing drugs) and need for
update to accommodate all care settings
160+ Row Decision Table not easily reviewed with Outlook,
geriatricians don’t have time for meetings
Even if this is commercial content, tools like this are still
needed for committee review
Outcome: successful review, updated geriatric dosing is
now integrated with more order entry applications
Example:
Setting aside the challenge of “who decides on content”,
This screenshot shows a common approach to decision support design:
MS office folders, documents related by title and common location only,
Difficult to know what changed from one version to next or why, people move
on to new jobs and folders get lost…..
Must maintain the “Metaknowledge”:
the who, what, when, where, and why about decision support content,
Partners maintains a geriatric dosing database that supports either default dosing
more appropriate for geriatric population or substitution recommendations
Votes
Other poll approaches
E-room report illustrating aggregation of expert input,
dramatically improves efficiency for subject matter experts
and medication services design teams
Vioxx alert shared, removed from
database
After updating the content in the medication knowledge base, updates are
published quarterly to the portal to share across sites, and we are still working
with vendors to achieve integration of Gerios dosing with Meditech and Siemens
Diabetes:
Case example of the
knowledge editing challenge
• Epidemic, associated with obesity
• Estimated avg $21,000/year per diabetic employee
in absenteeism, disability and medical costs (study
of 6 employers with 375,000 employees)
• HEDIS measures drive reimbursement
• Maintain HbA1c <7 (diet, oral agents and/or insulin)
• If Renal Disease and no contraindication, should be on
ACE inhibitor or ARB
• If lipid disorder and no contraindication, should be on a
statin like lipitor
Methodology
•
•
•
•
•
•
•
Identify relevant guidelines to achieve a given
measurement goal
Decompose guidelines into the draft components
for rules, documentation templates, order sets,
and reporting
Evaluate feasibility of current EHR system to
deploy
Determine what content can be purchased
Establish baseline performance
Implement content
Re-measure and refine
Editing Evidence into
Clinical Decision Support -•
Imagine a HEDIS measure:
•
Patients with Diabetes and Renal Disease or
Hypertension should be on an ACE inhibitor or ARB
unless there is a “contraindication”
•
Must define who has Diabetes, Renal Disease
and Hypertension (problems, documentation,
medications, test results)
•
Define “Contraindication to ACEi or ARB”
•
•
•
•
•
•
Allergy,
Cough symptom on adverse reaction list
Hyperkalemia on problem list or high K test result
Pregnancy (Many components)
Patient refuses
Must be the same in rules, documentation templates,
and reporting tools
Composite Decision Support Application: Diabetes Management
Guided Data Interpretation
Guided Observation Capture
Guided Ordering
Managing change across
content types, across silos
•
The rate of change for contraindication definition today is
very slow, it’s a challenge for most EHRs to provide
decision support for this
•
With the advent of molecular medicine, this rate of
change could become daily
•
Our approach to build a knowledge staging infrastructure
that enables our editors to manage propagation and
inheritance across related content areas
•
Change in one content area can trigger notifications to
edit related content or to validate and automate
propagation to relevant content
Conceptual Architecture for Knowledge Management
Knowledge Management Staging Area
eRoom Content Vetting and Content Editing Portals
Knowledge Search, Retrieval, and Reporting Services
Editors
Rules, Templates, Reports
Validation & Integration
3rd Party
Content
NDDF
Zynx
Micromedex
Clineguide
Etc.
Collaboration
Services
MetaKnowledge
Complex Definitions
Health Language Inc. Terminology
Documentum Content Management Server
Documentum
Deployment Agent
Quality Data Management
Knowledge Repositories
Transaction
Knowledge Repositories
Web
Service
Lookup
Production Knowledge Bases
Partners Services
And Applications
Managing Evidence at the Speed of Change
•
Requires strong leadership commitment to invest in
knowledge, people and systems to manage knowledge
•
Health systems vendors will need to vastly improve their
capabilities to support knowledge acquisition and usable
decision support
•
Standards for content representation and interoperability
remain a challenge, the current state benefits vendors and
consultants
•
Market forces (aging, genomics, pay-for-performance) create
decision support imperatives that will drive technology
innovation
Questions?
Never, ever think outside of the box!!!