Transcript Assessment of Data Warehousing in Healthcare - IEEE-USA

Data Warehousing for
Telehealth/Telemedicine
2001 Symposium on Applications & the Internet
(Saint 2001)
San Diego, CA
January 10, 2001
Rose Cintron-Allen, MHP
[email protected]
Warren Sterling, PhD
[email protected]
Agenda
•
•
•
•
•
•
•
•
•
•
Data Warehouse Definition
Building a Warehouse
Warehouse Uses in Telehealth/Telemedicine
National Medical Knowledge Bank (NMKB) Project
NMKB Features
Virtual Healthcare Events
Intelligent Agent-Based Active Learning Framework
Case Finder
Evidence Finder
Summary & Conclusions
What is a Data Warehouse?
Data Warehousing is a process, not a
product
It is a process for properly assembling
and managing data from various
sources for the purpose of answering
business questions and making
decisions that were not previously
possible.
Source: NCR Corporation (and Teradata Corp.) Copyrights 1982-1996. All rights reserved.
Building a Warehouse
 Define Business Opportunity & ROI
 Sourcing the Data
 Leverage
Lessons Learned
&
ROI Evaluation
 Operation &
Maintenance
 Planning and
Design
 Physical
Implementation
Telehealth/Telemedicine Uses
Distance Healthcare Education
Teleradiology
Telepathology
Healthcare Informatics
Patient Medical Record
National Medical Knowledge Bank
An Advanced Technology Program Joint Venture
Allegheny-Singer Research Institute, Pittsburgh PA
Millennium Healthcare Solutions, Edison NJ
NCR Corporation, Dayton OH
MCP Hahnemann University, Philadelphia PA
AT&T Government Markets, McLeansville NC
Sponsored by:
The United States Department of Commerce
National Institute of Standards and Technology
What is It?
The National Medical Knowledge Bank is an advanced
repository of specialized medical experience and
knowledge, featuring easy capture of information, with
massive storage and links to a vast selection of
supporting resources enabling users to retrieve,
collaborate, and learn, anytime and place.
NMKB Consortium Focus
Multimedia-based medical information
services three key markets.
Medical Texts
Medical Images
Medical Sounds
Medical Videos
Virtual Medical
Environments
• Digital Content
Acquisition
• Validation
• Data Integrity
Health Care Delivery
Parallel
Object/Relational
Database
(TOR)
National Medical Knowledge Bank
Features
• Web-based Architecture
• Targeted to healthcare practitioners
• Integrated applications
– Virtual Medical Conferences
– Nursing Training
– Case-Based Retrieval for Diagnosis/Treatment
Determination
– Disease Domain-specific Literature Search
• Offers Continuing Education credits
• Designed to use an object relational database as
a scalable, parallel data warehouse
What is an Object Relational
Database?
Traditional Data
+
Alphanumeric
• Integer
• Character
• Date
• Float
New Data
System Defined Types (SDTs)
• Audio
• Image
• Video
• Geospatial
• Document/Text
System/User Defined Functions (SDFs/UDFs)
• Audio/Video matching; Voice Recognition
• Iris Scan; Tumor Classification
• Playback/Fast-forward; Retrieval
• Map-overlay; Distance computations
• Word-Recognition/Translation
Object Relational Table
Alphanumeric Attributes
“Object” Columns (SDTs)
Char (n) Integer Float Image Audio Video Point
Account
Name
Patient
Name
Account
ID
Patient
Age
Balance
Account
Balance
Face
Print
MRI
Scans
Voice
Print
Doctor
Comm.
Text
ATM
Home
Transcript
tnx
Location Last Call
Angio- Work
Transcribed
gram
Location Dr. Comm
Virtual Healthcare Events:
Virtual Conferences
Grand Rounds
Brain Attack Conference - Video
with Slide Indexing
•
Asynchronous,
discretionary viewing
•
Lower cost option
(travel, time)
•
Concept searching
•
CME credit,
including JE/JIT
•
Indexed presentation
outlines for fast
navigation
•
Streaming
video/audio with
synchronized slides
Primary Care Grand Rounds Contents and Presentation
Intelligent Agent-based Active
Learning Framework
Intelligent Agent-based Continuing
Education - Activity Menu
•
Delivers personalized,
active education
experience
•
Problem-based learning
•
At convenience of student
schedule; can complete in
multiple sessions
•
Certified CNE credit
•
Reusable framework;
lowers production cost
•
Agent technology - lesson
planning; student model;
tutor; ontology
•
Applicable to other
domains
Intelligent Agent-based Continuing
Education - HPI Video Interview
•
Patient interview
video vignettes
(streaming video)
•
Student selects
questions and
orders them
•
Student evaluated
on question choice
and order
Intelligent Agent-based Continuing
Education – Review Patient Chart
Case Finder (Case-based Reasoning)
Case Finder - Retrieval of
Similar Cases
• Uses Case-Based
Reasoning to Find
“Similar” Cases
• Eases clinician’s
burden of reading and
recalling cases
• Saves clinician time
• Web-based; works
with any ODBC
database
• Matches complex
data
Medical Ontology
Case Finder uses the Unified Medical
Language System (UMLS) with >700K
concepts and 1.5M concept names
– allows matching at the concept level
• cva = stroke = brain attack
– identifies relationships between concepts
• embolic stroke is a kind of stroke
Evidence Finder
Evidence Finder: Query
Key Disorders
Diabetes
Lipid Disorders
Stroke
Asthma
Osteoporosis
AIDS
Pneumonia
Subtopic
Study Method
Epidemiology
Randomized
Prevention
Trial
Case Study
Diagnosis
Screening
Population Study
Treatment
Guidelines
Non-pharm Rx
Review Article
Outcomes
Meta-Analysis
Cost-Effectiveness
Heart Failure
Hypertension
Cost Benefit
Analysis
•
•
•
•
•
“Essence of the
essence” of recent
published literature:
abstracts and “pearls”
Evidence-based;
categorized by
evidence quality
Reduces need for
colleague consultation
Accessible, intuitive,
relevant
“Think with me”
functionality
Preop Consults
Enter words or phrases, separated by commas:
glucophage
Submit Query
Clear
Evidence Finder: Results
You matched 6 out of 121 documents
1 Efficacy of metformin in
patients with non-insulindependent diabetes mellitus.
The Multicenter Metformin
Study Group
In obese patients with poorly
controlled type 2 diabetes on diet
alone, use of metformin titrated to
maximal dosage or fasting
plasma glucose of <140 mg/dl
resulted in significantly improved
average glycemic control,
compared to placebo.
…
Presented by James Gavin, M.D.
6 Minority Issues in the
at MCP Hahnemann University
Management of Diabetes
Mellitus - Recommendations on 3/3/99 [0 CME credits]
Evidence Finder: Results
EVIDENCE BASED MEDICINE INDEXED DOCUMENT RETRIEVAL
Treatment Pharmacologic (Randomized Control-Certified)
Pearl
In obese patients with poorly controlled type 2 diabetes on diet alone, use of
<metformin> titrated to maximal dosage or fasting plasma glucose of <140 mg/dl
resulted in significantly improved average glycemic control, compared to placebo
(hemoglobin A1C 7.1% vs. 8.6%, respectively). In a second study, obese patients
with poorly controlled type 2 diabetes despite maximal glyburide therapy had
improved glycemic control on glyburide plus <metformin> (titrated as above)
compared to those remaining on glyburide alone (hemoglobin A1C 7.1% vs. 8.7%,
respectively). Patients changed from maximal glyburide to titrated dose
<metformin> had minimal improvement in average glycemic control.
Abstract
BACKGROUND: Sulfonylurea drugs have been the only oral therapy available for
patients with non-insulin-dependent diabetes mellitus (NIDDM) in the United
States. Recently, however, <metformin> has been approved for the treatment of
NIDDM. …
Evidence Finder: Results
MULTIMEDIA DOCUMENT RETRIEVAL, WITH CME
CREDIT, INDEXED BY PRESENTATION SEGMENT
What the Warehouse Holds
• NMKB Content Index for all current and potential
content for all applications - based on Dublin Core
• Evidence-based medicine abstracts and “pearls”
• All exemplar cases for CaseFinder application
• Digital video and slides/images for all virtual events
• All content for Active Learning Framework training
modules
• Unified Medical Language System ontology
• Student history
• User data
• Potential content for the NMKB
Summary and Conclusions
• The NMKB project sponsored by NIST ATP was
successfully completed
• The NMKB will support innovative interactive
multimedia-enabled medical applications.
• Goal is to commercialize the NMKB or pieces of it.
• The data warehouse of a commercial NMKB must
be supported by object relational database
technology
• Scalable growth is required as content base grows.
• Parallel database operation is required for
sophisticated CBR searching and concurrent
access by large numbers of users.