Toward The Intelligent Electronic Health Record The VA experience

1
Toward The Intelligent Electronic Health Record
- The VA experience

• Steven (Hank) Rappaport, MD
• Enterprise Systems Manager for Provider Systems
• Office of Information, Veterans Health
  Administration, Department of Veterans Affairs
• May 11, 2007

2
Objectives

• Describe the VHA Electronic Health Record and
  computing environment, and its intelligence
• Compare the capabilities and intelligence of
  other systems
• Describe what is needed for a truly intelligent
  system

3
No Current Relevant Financial Relationships with
Commercial entities(However, starting May 21,
2007 I begin work with Microsofts Azyxxi project)
Steven (Hank) Rappaport, MD
Disclosures
4
Topics

• Background VHA and VistA
• Intelligence
• Overview
• Data Browsing
• Ordering / Order checks / Order suggestion
• Narrative documentation
• Guideline compliance
• Treatment planning and care coordination / panel
  management
• Patient tracking / workflow monitoring
• Data Mining
• Problem oriented charting
• Decision support / links to knowledge resources

5
Background

• VHA and VistA

6
Background - Outline

• VHA environment
• Brief history of the VHA EHR
• EHR Capabilities
• Impact of EHR use
• Enablers, limitations constraints

7
VA Environment

• Patients 4.5M veteran users
• Sites of care 150 medical centers, 800 clinics
• Outpatient visits per year 50M
• Admissions per year 550k
• Staff 200k
• Bills sent to insurance 21M/year
• Budget 1.4B IT

8
VistA Usage

• 1.7B orders, 1M/day
• 640M Images, 910k/day
• 906M Documents, 640k/day
• 885M Medication admin via BCMA, 620k/day
• 1.1B Vitals, 725k/day
• 1M lab results /day
• 200M Outpatient Rxs dispensed/year

9
History of the VHA EHR

• Underground railroad in 1978
• Official deployment of DHCP in 1982
• Lab, pharmacy, scheduling
• VistA Imaging 1992
• CHCS
• Order Entry / Results Reporting 1994
• CPRS GUI 1998
• BCMA 2000

10
History of the VHA EHR

• Opensource
• Global deployment
• Mexico
• State Vet Homes in Oklahoma
• Midland Memorial Hospital, Texas
• American Samoa
• VistA Office EHR
• World VistA (www.worldvista.org)
• Vendor consortium
• www.hardhats.org

11
VistA Packages Clinical

• CPRS - Order Entry / Results Reporting
• Pharmacy
• Laboratory
• Radiology / Imaging
• Surgery
• Nursing
• Nutrition Food Service
• Audiology and Speech Pathology
• Social Work
• Medicine / Clinical Procedures
• Clinical Registries

12
VistA Packages Clinical Admin

• Admission / Discharge / Transfer
• Patient Information Management
• Appointment Management / Scheduling
• Integrated Billing / Fee Basis
• Patient Care Management Module
• Prosthetics

13
VistA Packages Hospital Management

• IFCAP (purchasing)
• Engineering
• Time and attendance

14
VistA current architecture

• M database
• 128 production instances, facility-centric
• M server side programming language
• Windows applications written in Delphi
• National Health Data Repository and other
  services
• Migrating to Java, relational databases, 3-tier
  service oriented architecture

15
EHR Functional Capabilities

• CPOE
• Narrative documentation
• Role-based document management
• Request management
• Data browsing
• Alerts management
• Algorithmic reminders
• Panel View and query

16
(No Transcript)
17
(No Transcript)
18
Enablers

• Single database
• Minimal interface issues
• Captive audience to some degree
• Clinical Indication Data Capture mutiny
• Willingness to iterate
• Designed by end-users, for end-users
• Implementation approach
• Local flexibility for development and
  configuration
• Iterative
• Clinical Application Coordinators

19
Constraints

• Physician time
• Regional data
• Incomplete data
• Standards
• Privacy / security

20
Recognized Limitations of the VHA EHR

• Limited coded clinical data
• Non-standardized clinical data
• Limited knowledge representation language
• Limited ability to exchange knowledge across
  sites
• Becoming increasingly difficult and expensive to
  maintain / enhance the software
• Vendor dependence
• Query tools limited
• Limited coding of patient context
• Limited links between orders and
• Documentation
• Problems
• Follow-up

21
Intelligence
22
Wheres the intelligence?
23
Intelligence Definition

• Any sufficiently advanced technology is
  indistinguishable from magic.
• Arthur C. Clarke, "Profiles of The Future", 1961
  (Clarke's third law)

24
Intelligence Types in EHR

• Links to knowledge resources
• Context sensitive
• Intelligent prompting for documentation
• Presenting likely information of interest
• Assessment of proposed therapy
• Therapy advice
• Therapy monitoring
• Learning

25
Is there any intelligent life out there?
26
Searching for Intelligence

• Data Browsing
• Ordering / Order checks / Order suggestion
• Narrative documentation
• Guideline compliance
• Treatment planning and care coordination / panel
  management
• Patient tracking / workflow monitoring
• Data Mining
• Problem oriented charting
• Decision support / links to knowledge resources

27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
(No Transcript)
38
(No Transcript)
39
(No Transcript)
40
(No Transcript)
41
Data browsing

• Wheres the Intelligence?
• What kind of intelligence would be helpful?
• What more is needed?

42
Intelligent Browsing

• Display related information
• Trend identification
• Links to related information
• Problem focused
• Plan focused

43
(No Transcript)
44
(No Transcript)
45
(No Transcript)
46
(No Transcript)
47
(No Transcript)
48
Ordering / Order checks / advice
49
Ordering - Outline

• Simple
• Quick orders
• Order menus
• Order sets
• Closing the loop BCMA
• Problem oriented ordering

50
(No Transcript)
51
Basic med ordering
52
Basic med ordering
53
Basic med ordering
54
Basic Med Ordering
55
Personal quick orders
56
Personal quick orders
57
Order menus
58
Order menus
59
Orders by category
60
Menus of quick orders
61
Order sets
62
Disease specific order menus
63
Add education
64
Basic decision support
65
Ordering as a byproduct of care
66
(No Transcript)
67
Utilization Diagnosis-SpecificGabapentin
68
Ordering

• Wheres the intelligence?
• What kind of intelligence is helpful?
• What more is needed?

69
Narrative documentation
70
Narrative Documentation outline

• Simple templates
• Linked to notes
• Patient data objects
• Starting templates for notes in progress
• Interactive dialogs
• Controls
• Branching logic
• Saving data as well as text
• Links to reference sources
• Linking notes to consults / procedures
• Actions as a byproduct of documentation
• Images linked to templates
• Templates on the fly

71
(No Transcript)
72
(No Transcript)
73
(No Transcript)
74
(No Transcript)
75
Dialog templates
76
Dialog templates
77
(No Transcript)
78
(No Transcript)
79
(No Transcript)
80
(No Transcript)
81
Links to reference materials
82
(No Transcript)
83
(No Transcript)
84
(No Transcript)
85
(No Transcript)
86
(No Transcript)
87
Narrative documentation

• Wheres the intelligence?
• What kind of intelligence is helpful?
• What more is needed?

88
Narrative documentation issues

• Protected text
• Copy / paste
• Default templates for clinical context
  (appointments, patient condition, etc)
• Editing complex, unsigned notes
• Data standardization

89
Guideline Compliance
90
Guidelines - Outline

• VA - DoD Clinical Practice Guideline Council
• Clinical reminders for executing guidelines
• Due now
• Performance measures
• Limitation completeness of data (single site vs
  VHA)
• Athena
• MUEs

91
VA DoD Clinical Practice Guideline Council
92
Automation-assisted guidelines

• Hypertension
• Depression screening
• Lipid management
• Mamography / Pap
• Hepatitis C screening
• Diabetes management
• Colorectal cancer screening
• Influenza / pnemovax
• Exercise / weight / smoking

93
CPG implementation HTN
94
CPG implementation - HTN
95
CPG implementation Antipsychotics and
Involuntary Movements
96
Antipsychotics
97
Athena
98
Medication Utilization
99
Guidelines summary

• Wheres the intelligence?
• What kind of intelligence is helpful?
• What more is needed?

100
Treatment planning / Care Coordination / Panel
Management
101
Care Coordination outline

• Treatment planning / monitoring
• Care Coordination
• Care Management
• Dashboard
• Query tool

102
Treatment Planning / Monitoring

• Specify goals of care, future data needs, future
  conditional actions
• System monitors for events defined by plan and
  generates alerts
• Long way to go
• Desire to do this in Care Delivery Support System

103
Care Coordination

• Home telehealth
• Integration of multiple vendor systems
• Monthly summary progress notes
• Nurse case managers
• Additional signer capability
• Alerts

104
Panel Management
105
Panel Management
106
Panel Management
107
Treatment Planning / Care Coordination / Panel
Management

• Wheres the intelligence?
• What kind of intelligence is helpful?
• What more is needed?

108
Patient tracking / workflow monitoring
109
Patient Tracking / workflow monitoring

• Build models of care delivery process
• Track patients progress through the model
• Notify when progress not meeting expectations

110
Scenarios

• Emergency department
• Outpatient surgery
• Chemotherapy clinic

111
Approach

• Workflow engine
• Monitor for clinical events
• Order placement
• Result availability
• Contact with staff
• Alert using existing tools
• ED whiteboard

112
Patient tracking
113
Patient Tracking
114
Patient tracking / workflow monitoring

• Wheres the intelligence?
• What kinds of intelligence are helpful?
• What more is needed?

115
Data Mining
116
Data Mining - Topics

• Health data repository
• Seasonal variation in blood pressure
• Problem order associations

117
Health Data Repository

• National repository of VHA clinical data
• Allergies
• Medications
• Laboratory results
• Vital signs
• Demographic information
• More coming
• Official medical record QA/research tool

118
Seasonal and geographic variation in blood
pressure
119
Seasonal and geographic variation in blood
pressure
120
Problem Order associations
121
Data Mining summary

• Wheres the intelligence?
• What kinds of intelligence are helpful?
• What more is needed?

122
Problem oriented charting
123
Problem oriented charting - topics

• Clinical Indications
• Drug database
• Euclid assist

124
Clinical Indications Data Capture

• Desire to have a diagnosis for each order
• Caused near mutiny

125
(No Transcript)
126
(No Transcript)
127
Current approach to Dx for orders
128
Problem-focused display of related data
129
Problem oriented charting summary

• Wheres the intelligence?
• What kinds of intelligence are helpful?
• What more is needed?

130
Decision support / links to knowledge resources
131
Decision support topics

• Errors the VHA EHR does not yet prevent
• Links to knowledge
• Differential diagnosis
• Therapy suggestion

132
Errors that the VHA EHR does not yet prevent

• Incorrect drug
• Incorrect dose
• Incorrect associated orders

133
Links to knowledge

• Context sensitive links
• Calculators

134
Differential diagnosis / therapy suggestion
135
Decision support / links summary

• Wheres the intelligence
• What kinds of intelligence are helpful
• What more is needed

136
Intelligent EHR VHA experience

• Summary
• Browsing
• Ordering
• Documentation
• Planning / plan monitoring
• Problem focused care
• Panel management
• Diagnostic support
• Data mining as a way to derive intelligence
  automatically

137
The Intelligent EHR is an enabler for the
delivery of high quality, cost effective care
that leads to satisfied staff and patients