Using the Electronic Health Record to Encourage EvidenceBased Practice

1
Using the Electronic Health Recordto Encourage
Evidence-Based Practice

• Jonathan S. Einbinder, MD, MPH
• Partners HealthCare
• (jseinbinder_at_partners.org)
• August 22, 2006

2
Outline

• Meeting information needs the challenge
• Potential of electronic health record (and
  cautions)
• Partners HealthCare Signature Initiative
• Smart Forms
• Registries

3
Patient care information needs in a teaching
hospital

• 24 doctors and students observed
• 5 clinical questions per patient
• 74 concerned patient care

--Osheroff 1991
4
Usefulness of medical information
relevance x validity work to access
Usefulness
--Shaughnessy, Slawson, Bennett 1994
5
The role of technology (the 64 question)

• How can informatics applications
• identify information needs?
• provide useful information?
• change clinician behavior?
• change patient behavior?

6
A good place to start is basic management of
information to help us do things that we know we
should do.

• Make sure that patients reliably receive
  interventions known to be beneficial, leading to
  improved outcomes.

7
Getting past 55 performance
-- EA McGlynn et al. NEJM 20033482635-2645
8
Established patient care strategies are not used

• Random sample of adults in 12 metro areas
• Survey and medical records review
• 439 indicators of quality of care for 30 acute
  and chronic conditions, as well as preventive
  care
• Key finding Participants received 54.9 of
  recommended care

-- EA McGlynn et al. NEJM 20033482635-2645
9
Impediments to using current best evidence in
clinical practice

• When to look for new evidence
• Where to locate relevant information
• How to synthesize it to determine optimal approach

Technology (electronic health record) can
facilitate effective use of information
--Haynes et al., JAMIA 1995
10
Diabetes Population Management
But, consider the following examples

• Intervention
• Novel population management software (registry)
• NP used weekly and emailed suggestions to PCPs.
• Results More testing (A1C and LDL) at
  intervention clinic. No significant effect for
  med prescription, LDL and A1C levels.

Grant et al. Diabetes Care 200427(10)2299-2305
11
OnCall Population Management for Diabetes
12
OnCall Population Management for Diabetes
13
Computerized reminders for CHF/CAD

• Intervention Evidence-based cardiac care
  suggestions, approved by a panel of local
  cardiologists and general internists, were
  displayed to physicians and pharmacists as they
  cared for enrolled patients.
• Results The intervention had no effect on
  physicians' adherence to the care suggestions
  (23 for intervention patients versus 22 for
  controls).

14
Computerized reminders for diabetes and CAD

• Intervention Computerized reminders.
• Results Diabetes reminders resulted in the
  recommended action in 19 of patients in the
  intervention group versus 14 of patients in the
  control group. Annual cholesterol testing HR
  1.41, 95 CI 1.151.72
• ACE inhibitors HR 1.42, 95 CI 0.942.14
• Statin use NS
• Annual eye exam - NS

Sequist et al. J Am Med Inform Assoc.
200512431437.
15
LMR Summary with diabetes reminders
16
Partners Signature Initiative III Goals and Scope

• Problem Proven effective medical interventions
  are not uniformly applied across health care
  settings.
• Goal Set standards, improve clinical
  performance, and achieve national leadership
  levels of performance reliably across the system.
• Acute Myocardial Infarction (AMI)
• Congestive Heart Failure (CHF)
• Diabetes
• Pneumonia
• Smoking cessation

17
EMR Framework
Adoption Get an EMR and use it
Effective Use Use key EMRfeatures fully
Smart Use Leverage EMRdecision support
18
Diabetes Decision Support Team

• Goal Increase the number of providers using and
  patients with diabetes being managed with
  standard evidence-based decision support.
• How
• Develop and deploy an integrated suite of
  information system tools to assist providers in
  caring for their diabetic patients. Specifically,
  this suite will include the Diabetes Smart Form,
  the Diabetes Registry, and the Diabetes
  Population Manager.

19
What are Smart Forms?

• Composite application for clinicians, combining
  documentation, orders, decision support
• Use by clinician during an office visit
• Support efficient workflow with decision support
• Built upon a foundation of clinical applications
  and services
• Data review
• Documentation
• Actionable decision support and orders

20
(No Transcript)
21
Decision Support Highlights
22
Actionable decision support and orders
23
Smart Form origins

• AHRQ Smart Forms and Quality Dashboards grant
• 3 year, 1.2M project to evaluate Smart Forms for
  acute respiratory infection (ARI) and coronary
  artery disease (CD) in a randomized trial
• Signature Initiative 3
• Potential of Smart Forms to promote adherence to
  evidence-based guidelines for diabetes. Smart
  Form for diabetes will be evaluated in randomized
  trial, along with ARI and CAD.

24
Smart Forms Clinical and Quality Opportunity

• Address a critical obstacle to effective use of
  the EMR integration of documentation and
  actionable decision-support in a streamlined
  workflow
• Positively impact clinical processes and outcomes
• Positively impact attainment of performance
  targets in reimbursement contracts
• Provide mechanism to help achieve goals and
  objectives for patient safety, disease
  management, and trend management (in ambulatory
  care settings)

25
Knowledge Management Current
Codes Rules Templates Logic
Clinical Content and Guidelines
Functional Knowledge Specification
Technical Knowledge Specification
For ARI, CAD, DM Iterative, painstaking, manual
process involving clinical sponsors, developers,
analysts shepherded by KM.

• Able to build first Smart Forms
• Needs, issues became apparent
• Knowledge Specification approach defined.

26
Knowledge Management Future
Codes Rules Templates Logic
Clinical Content and Guidelines
Functional Knowledge Specification
Technical Knowledge Specification
Over next year, infrastructure and processes will
evolve

• Processes
• Guideline approval
• Functional knowledge spec.
• Technical knowledge spec.
• Maintenance and updating

• Tools
• Concept Dictionary
• Order Catalogue
• Template Editors
• Rule Editor

27
(No Transcript)
28
(No Transcript)