From: Subject: YMPOSIUM ROCEEDINGS Date: Tue, 21 Oct 2003 20:10:58 +0300 MIME-Version: 1.0 Content-Type: text/html; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Location: http://216.239.59.104/search?q=cache:XfXK8ZwisqUJ:www.nhlbi.nih.gov/health/prof/heart/mi/ha_tech.pdf+craig+feied&hl=en&ie=UTF-8 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1165 =EF=BB=BF YMPOSIUM ROCEEDINGS
This = is the html=20 version of the file http://www.nhlbi.nih.gov/health/prof/heart/mi/ha_tech.pdf.
G o = o g = l e=20 automatically generates html versions of documents as we = crawl the=20 web.
To link to or bookmark this page, use the following = url:=20 = http://www.google.com/search?q=3Dcache:XfXK8ZwisqUJ:www.nhlbi.nih.g= ov/health/prof/heart/mi/ha_tech.pdf+craig+feied&hl=3Den&ie=3DUTF-= 8

Google is not affiliated with the = authors=20 of this page nor responsible for its=20 content.
These=20 search terms have been highlighted:  craig  feied 
Page 1
N A T I = O N A=20 L
I N S = T I T U T=20 E S O F H E A L T H
N A T I = O N A=20 L
H E A = R T , L U=20 N G , A N D B L O O D I N S T I T U T E
 S YMPOSIUM P ROCEEDINGS N EW I NFORMATION T ECHNOLOGY A ND T HE N ATIONAL H EART A TTACK A LERT P ROGRAM
: =
S
ETTING=20 A
5-Y EAR A GENDA
APRIL = 14=E2=80=9315,=20 1998
LISTER = HILL=20 CENTER,
NATIONAL LIBRARY=20 OF MEDICINE,
BETHESDA,=20 MARYLAND
National Heart=20 Attack Alert Program
National Heart=20 Attack Alert Program
 S YMPOSIUM P ROCEEDINGS N EW I NFORMATION T ECHNOLOGY A ND T HE N ATIONAL H EART A TTACK A LERT P ROGRAM
: =
S
ETTING = A
5-Y EAR A GENDA
Cosponsored=20 by:
National=20 Heart, Lung,
and = Blood=20 Institute;
National=20 Library
of = Medicine;=20 Agency
for = Health=20 Care Policy
and = Research
NIH = Publication No. 99-4089
September=20 1999
National=20 Institutes of Health
National=20 Heart, Lung,
and = Blood=20 Institute
 iii
Cosponsoring=20 Agency Directors . . . . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . = .iv
Program=20 Planning Committee . . . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .=20 .v
Program=20 Planning Committee Task Forces . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . = .vii
Symposium=20 Proceedings Overview . . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . = .viii
Symposium=20 Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .=20 .viii
Introduction=20 to the Symposium . . . . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . = .1
Tutorial=20 Presentations . . . . . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . = . .=20 .3
The = National=20 Heart Attack Alert Program: Its Mission and Its Accomplishments . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . .3
Telecommunications and=20 the Next Generation Internet . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . .5
Confidentiality=20 and Medical Information . . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13
Diagnosis of=20 Patients with Acute Myocardial Infarction and Other Acute Coronary = Syndromes . .=20 . . . . . . . . . . . . . .15
Databases and=20 Data Standardization . . . . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17
Economic, Social,=20 and Psychological Obstacles to Achieving the
National Heart=20 Attack Alert Program=E2=80=99s Goals . . . . . . . . . . . . . . . . . . = . . . . . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . .19
Symposium=20 Presentations and Abstracts . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . = .21
Telehealth . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .=20 .21
Telemedicine in=20 the Department of Defense (DoD) . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . .22
Use of=20 Telemedicine Technology To Assist Small Rural Hospitals . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . .22
Ambulance-on-the-Web:=20 A Cellular Phone-Based, Real-Time
Video-Mobile=20 Telemedicine System for Ambulance Transport . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . .23
Behavioral Issues=20 in Initiating Use of Telemedicine Technology . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . .23
Electronic=20 Housecalls at Fort Gordon . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . = .24
Using = Televideo=20 To Improve Patient Outcomes . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . .25
Education of=20 Public, Patients, and Health Care Providers . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . .27
Heart = Attack=20 Rapid Early Action for Coronary Treatment (REACT): A Community-Based=20 Intervention
To = Reduce Patient=20 Delay in Seeking Emergency Medical Care for Heart Attack Symptoms . . . = . . . .=20 . . . . . . . . . .27
Using a = Comprehensive Web Site To Attract, Identify, and Educate People at Risk = for=20 Heart Attacks . . . . . . . . .28
Stimulating the=20 Use of Information Technology in Support of Information Seeking . . . . = . . . .=20 . . . . . . . . . . . . . . .29
Medical = Records=20 Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . = .30
Executive Summary=20 G-7 Cardio Project Subproject III: Improving
Prevention,=20 Diagnosis, and Treatment of Major Cardiovascular Diseases . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . .31
C ONTENTS iv
Accessing Medical=20 Records Using the Internet . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . .32
Portable=20 Information Storage Devices for Immediate Access to Cardiac Data at the = Point of=20 Care . . . . . . . . . . .33
Diagnosis and=20 Treatment Decision Support . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . .34
Use of=20 Interactive Decision Trees for Patients To Decide About the Need To Call = an=20 Ambulance . . . . . . . . . . . . .35
The = Promise and=20 Problems of Community-Based Measurement
of = Biochemical=20 Markers of Acute Coronary Syndromes . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . .35
Home = 12-Lead ECG=20 Acquisition in Patients at High Risk of AMI . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . .36
Access = to=20 Electronic Medical Records: Critical Data About the Right Patient at the = Right=20 Time . . . . . . . . . . . . . .37
Understanding=20 Cognitive Processes in Medical Decisionmaking . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . .38
The Use = of=20 Computerized Electrocardiograph To Provide Decision Support for the Use = of=20
Thrombolytic=20 Therapy as Well as To Record Retrospective Analysis of Performance for = Feedback=20 . . . . . . . . . . .39
Large-Scale=20 Databases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .=20 .41
Use of = the=20 Internet for Database Access . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42
Continuous=20 Quality Improvement in the Care of Patients With Acute Myocardial = Infarction:=20
The = National=20 Registry of Myocardial Infarction (NRMI) . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . .43
Validity and=20 Efficiency in Collecting Quality Data in Acute Myocardial Infarction = Patients:=20
A = Comparison of=20 the Cooperative Cardiovascular Project (CCP) and NRMI-2 . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . .44
The = Cochrane=20 Collaboration and the Agency for Health Care Policy and = Research=E2=80=99s
Evidence-Based=20 Practice Centers: The Benefit of Systematic Evidence Reviews in = Achieving=20
the = Mission of=20 the National Heart Attack Alert Program . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . .45
Use of = Clinical=20 Trial Data To Support Advances in Emergency Cardiac Research . . . . . . = . . . .=20 . . . . . . . . . . . . . .47
Appendix: List=20 of Participants . . . . . . . . . . . . . . . . . . . . . . . . . . = . . . .=20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . = .48
Claude = Lenfant,=20 M.D.
Director
National Heart,=20 Lung, and Blood Institute
National=20 Institutes of Health
Bethesda,=20 MD
Donald = Lindberg,=20 M.D.
Director
National Library=20 of Medicine
Bethesda, MD=20
John = M.=20 Eisenberg, M.D.
Administrator
Agency = for=20 Health Care Policy and Research
Rockville,=20 MD
C
OSPONSORING=20
A GENCY D IRECTORS v P
ROGRAM=20
P
LANNING=20
C OMMITTEE S
YMPOSIUM=20
P
ROGRAM=20
C HAIR:
Mark S. = Smith,=20 M.D.
Chairman,=20 Department of Emergency Medicine
Washington=20 Hospital Center
Clinical=20 Professor of Emergency Medicine
The = George=20 Washington University School
of = Medicine and=20 Health Sciences
Washington,=20 DC
M EMBERS:
James = M. Atkins,=20 M.D., F.A.C.C.
Medical = Director,=20 Emergency Medicine Education
Professor of=20 Internal Medicine
Division of=20 Cardiology
University of=20 Texas Southwestern Medical Center
at=20 Dallas
Dallas, = TX=20
J. = Robert Beck,=20 M.D.
Vice=20 President
Information=20 Technology Program
Baylor = College of=20 Medicine
Texas = Medical=20 Center
Houston, TX=20
Domenico=20 Cianflone, M.D., F.E.S.C.
Operational=20 Coordinator
G-7 = CARDIO=20 Project
Instituto di=20 Cardiologia
Universita=20 Cattolica del Sacro Cuore
Rome,=20 Italy
James = J. Cimino,=20 M.D.
Associate=20 Professor
Department of=20 Medical Informatics
Department of=20 Medicine
Columbia=20 University
Columbia-Presbyterian=20 Medical Center
New = York,=20 NY
Paul = Clayton,=20 Ph.D.
Professor and=20 Chair
Department of=20 Medical Informatics
Columbia=20 University
Director of=20 Clinical Information Services
Columbia-Presbyterian=20 Medical Center
New = York,=20 NY
William = Cordell,=20 M.D.
Director,=20 Emergency Medicine Research
Methodist=20 Hospital
Clarian = Health=20 Partners and Indiana University School
of = Medicine=20
Indianapolis, IN=20
Mary = Jo Deering,=20 Ph.D.
Director, Health=20 Communication and Telehealth
Office = of=20 Disease Prevention and Health Promotion
Public = Health=20 Service
Department of=20 Health and Human Services
Washington, DC=20
Craig Feied,=20 M.D.
Director of=20 Informatics
Department of=20 Emergency Medicine
Washington=20 Hospital Center
Associate=20 Clinical Professor of Emergency Medicine
The = George=20 Washington University School
of = Medicine and=20 Health Sciences
Washington, DC=20
Capt. = David W.=20 Ferguson, M.C., U.S.N.
Professor and=20 Associate Chairman of Medicine
Division of=20 Cardiology
National Naval=20 Medical Center
Bethesda,=20 MD
Reed = M. Gardner,=20 Ph.D.
Professor and=20 Chairman
Department of=20 Medical Informatics
The = University=20 of Utah
School = of=20 Medicine,
Salt = Lake City,=20 UT
Carole = A.=20 Gassert, Ph.D., R.N.
Acting = Deputy=20 Branch Chief
Division of=20 Nursing, Bureau of Health Policy
Health = Resources=20 and Services Administration
Rockville,=20 MD
Attilio Maseri,=20 M.D.
Director
Instituto di=20 Cardiologia
Universita=20 Cattolica del Sacro Cuore
Rome,=20 Italy
Clement J.=20 McDonald, M.D.
Distinguished=20 Professor of Medicine
Indiana=20 University School of Medicine
Director
Regenstrief=20 Institute for Health Care
Indianapolis, IN=20
Joseph = P.=20 Ornato, M.D., F.A.C.C., F.A.C.E.P.
Professor and=20 Chairman
Department of=20 Emergency Medicine
Virginia=20 Commonwealth University
Medical College=20 of Virginia
Richmond,=20 VA
Daniel = A.=20 Pollock, M.D.
Medical=20 Epidemiologist
National Center=20 for Injury Prevention and Control
Centers for=20 Disease Control and Prevention
Atlanta,=20 GA
vi
Charles = Safran,=20 M.D.
Director,=20 Informatics and Advanced Technology
Beth = Israel=20 Deaconess Medical Center
Boston, = MA
Jay H. = Sanders,=20 M.D.
President
American=20 Telemedicine Association
Adjunct = Professor=20 of Medicine
Johns = Hopkins=20 University School of Medicine
President and=20 CEO
The = Global=20 Telemedicine Group
McLean, = VA
Paul = Tang,=20 M.D.
Medical = Director
Information=20 Systems
Northwestern=20 Memorial Hospital
Chicago,=20 IL
William = Tierney,=20 M.D.
Professor of=20 Medicine
Indiana = University School of Medicine
Senior = Health=20 Research Scientist
Roudebushva=20 Medical Center
Senior = Research=20 Scientist
Regenstrief=20 Institute for Health Care
Indianapolis,=20 IN
AGENCY FOR=20 HEALTH CARE POLICY
AND = RESEARCH=20 STAFF
J. = Michael=20 Fitzmaurice, Ph.D.
Senior = Science=20 Advisor for Information Technology
Agency = for Health=20 Care Policy and Research
Rockville, MD=20
Kathleen=20 McCormick, Ph.D., R.N.
Senior = Medical=20 Advisor
Agency = for Health=20 Care Policy and Research
Center = for=20 Information Technology
Rockville,=20 MD
 N ATIONAL H
EART,=20
L
UNG, AND=20
B
LOOD=20
I NSTITUTE S TAFF
Glen = Bennett,=20 M.P.H.
Coordinator
Advanced=20 Technologies Applications in Health Education
Programs
Office = of=20 Prevention, Education, and Control
National Heart,=20 Lung, and Blood Institute
National=20 Institutes of Health
Bethesda, MD=20
Mary = M. Hand,=20 M.S.P.H., R.N.
Coordinator
National Heart=20 Attack Alert Program
Office = of=20 Prevention, Education, and Control
National Heart,=20 Lung, and Blood Institute
National=20 Institutes of Health
Bethesda, MD=20
Gregory J.=20 Morosco, Ph.D., M.P.H.
Director
Office = of=20 Prevention, Education, and Control
National Heart,=20 Lung, and Blood Institute
National=20 Institutes of Health
Bethesda,=20 MD
N ATIONAL L
IBRARY OF=20
M
EDICINE=20
S TAFF
Michael J.=20 Ackerman, Ph.D.
Assistant=20 Director for High Performance Computing and
Communications
National Library=20 of Medicine
Bethesda, MD=20
Patricia R.=20 Carson
Special=20 Assistant to the Director
National Library=20 of Medicine
Bethesda,=20 MD
Elliot = R.=20 Siegel, Ph.D.
Associate=20 Director for Health Information Programs
Development
National Library=20 of Medicine
Bethesda, MD=20
R.O.W.=20 S
CIENCES,=20
I
NC.,=20
C
ONTRACT=20
S TAFF
Dawn = M.=20 Apple
Conference=20 Representative
R.O.W. = Sciences,=20 Inc.
Rockville,=20 MD
John = Clinton=20 Bradley, M.S.
Program=20 Manager
R.O.W. = Sciences,=20 Inc.
Rockville, MD=20
Pamela = A.=20 Christian, R.N., M.P.A.
Program=20 Specialist
R.O.W. = Sciences,=20 Inc.
Rockville,=20 MD
vii P
ROGRAM=20
P
LANNING=20
C
OMMITTEE=20
T
ASK=20
F ORCES T ELEHEALTH
Carole = Gassert,=20 Ph.D., R.N. (Chair)
Rockville,=20 MD
Capt. = David W.=20 Ferguson, M.C., U.S.N. (Vice Chair)
Bethesda, MD=20
Michael = J.=20 Ackerman, Ph.D.
Bethesda,=20 MD
William = Cordell,=20 M.D.
Indianapolis,=20 IN
Charles = Safran,=20 M.D.
Boston, = MA
Jay H. = Sanders,=20 M.D.
McLean, = VA
Bruno = M. Vasta,=20 M.S.
Lusby,=20 Maryland
L ARGE- S
CALE=20
D ATABASES
William = Tierney,=20 M.D. (Chair)
Indianapolis,=20 IN
Costas = T.=20 Lambrew, M.D. (Vice Chair)
Portland, ME=20
Domenico=20 Cianflone, M.D., F.E.S.C.
Rome,=20 Italy
J. = Michael=20 Fitzmaurice, Ph.D.
Rockville,=20 MD
Clement = McDonald,=20 M.D.
Indianapolis,=20 IN
Daniel = A.=20 Pollock, M.D.
Atlanta,=20 GA
D
IAGNOSIS AND=20
T
REATMENT=20
D
ECISION=20
S UPPORT
James = J. Cimino,=20 M.D. (Chair)
New = York,=20 NY
Robert = J.=20 Zalenski, M.D., M.A. (Vice Chair)
Detroit, MI=20
J. = Robert Beck,=20 M.D.
Houston, TX=20
Craig Feied,=20 M.D.
Washington, D.C.=20
Kathleen=20 McCormick, Ph.D., R.N.
Rockville,=20 MD
M EDICAL R
ECORDS=20
A CCESS
Reed = Gardner,=20 Ph.D. (Chair)
Salt = Lake City,=20 UT
Joseph = Ornato,=20 M.D. (Vice Chair)
Richmond,=20 VA
Paul = Clayton,=20 M.D.
New = York,=20 NY
Lawrence D.=20 Jones, M.D.
Summit, MS=20
Paul = Tang,=20 M.D.
Chicago,=20 IL
E
DUCATION OF=20
P
UBLIC,=20
P
ATIENTS=20
AND =
H
EALTH=20
C
ARE =
P ROVIDERS
James = M. Atkins,=20 M.D., F.A.C.C. (Chair)
Dallas, TX=20
Mary = Jo Deering,=20 Ph.D. (Vice Chair)
Washington, D.C.=20
Glen = Bennett,=20 M.P.H.
Bethesda, MD=20
Gregory J.=20 Morosco, Ph.D., M.P.H.
Bethesda, MD=20
Elliot = R.=20 Siegel, Ph.D.
Bethesda,=20 MD
viii S
YMPOSIUM=20
P
ROCEEDINGS=20
O VERVIEW
This = book=20 contains the proceedings for the symposium, New
Information=20 Technology and the National Heart Attack Alert
Program:=20 Setting a 5-Year Agenda, which took place on April
14 and = 15, 1998,=20 in Bethesda, Maryland.
The = symposium was=20 cosponsored by the National Heart,
Lung, = and Blood=20 Institute; the National Library of Medicine;
and the = Agency=20 for Health Care Policy and Research. The
goal of = the=20 symposium was to develop an agenda of
innovative ways=20 in which the domain of information
technology and=20 medical informatics can be used to help fulfill
the = mission of=20 the National Heart Attack Alert Program
(NHAAP), a=20 program of the National Heart, Lung, and Blood
Institute of the=20 National Institutes of Health. The mission of
the = NHAAP is to=20 reduce morbidity and mortality from heart
attack = through=20 early recognition and treatment.
Symposium=20 attendees comprised NHAAP coordinating
committee=20 members; representatives from the National
Library = of=20 Medicine; from the National Heart, Lung, and
Blood = Institute;=20 and from the Agency for Health Care Policy
and = Research;=20 content experts in the field of information
technology; and=20 health professionals in emergency medicine,
cardiology,=20 nursing, and emergency medical services. A list of
participants can=20 be found in the appendix.
The = symposium=20 was divided into two sections:
1. A = series of=20 tutorial presentations on medical informatics
content and=20 NHAAP background to lay the foundation for
understanding=20 the proposals presented on the second day
of the = symposium
2. A = series of=20 presentations organized into one of five
informatics=20 content areas=E2=80=94telehealth; education of the
public,=20 patients, and health care providers; medical records
access;=20 diagnostic and treatment decision support; and
large-scale=20 database access and mining=E2=80=94that constitute
specific ideas=20 and proposals for the application of new
information=20 technology to the NHAAP.
The = abstracts of=20 the presentations are structured in a common
format: a=20 delineation of the core content of the proposal, a
review = of what=20 has already been done in the area, a discussion
of how = the=20 proposal specifically serves the mission of the
NHAAP, = a=20 statement of perceived obstacles to successful
implementation=20 of the proposal, and recommendations for
next = steps to be=20 taken.
S
YMPOSIUM=20
P ROGRAM T
UESDAY,=20
A
PRIL 14,=20 1998
Tutorial=20 Presentations on New Information Technology
and the=20 National Heart Attack Alert Program:
Setting a=20 5-Year Agenda
Lister = Hill=20 Center
National Library=20 of Medicine
Bethesda,=20 Maryland
1:30-5:00=20 p.m.
1:30=20 p.m.
Welcome and=20 Overview
Mark = S. Smith,=20 M.D.
Symposium=20 Program Chair
Chairman,=20 Department of Emergency
Medicine
Washington=20 Hospital Center
Clinical=20 Professor of Emergency Medicine
The = George=20 Washington University School of
Medicine and=20 Health Sciences
Washington, DC=20
1:40=20 p.m.
The = National=20 Heart Attack Alert Program:
Its = Mission=20 and Its Accomplishments
James = M. Atkins,=20 M.D., F.A.C.C.
Medical=20 Director
Emergency=20 Medicine Education
Professor of=20 Internal Medicine
Division of=20 Cardiology
University of=20 Texas, Southwestern Medical
Center = at=20 Dallas
Dallas,=20 TX
2:10=20 p.m.
Telecommunications=20 and
the = Next=20 Generation Internet
Craig Feied,=20 M.D.
Director of=20 Informatics
Department of=20 Emergency Medicine
Washington=20 Hospital Center
Associate=20 Clinical Professor
of = Emergency=20 Medicine
The = George=20 Washington University School of
Medicine and=20 Health Sciences
Washington,=20 DC
 ix
2:40=20 p.m.
Confidentiality=20 and Medical Information
Paul = Clayton,=20 Ph.D.
Professor and=20 Chair
Department of=20 Medical Informatics
Columbia=20 University
Director of=20 Clinical Information Services
Columbia=20 Presbyterian Medical Center
New = York,=20 NY
3:10=20 p.m.
BREAK
3:30=20 p.m.
Diagnosis of=20 Patients with Acute Myocardial
Infarction=20 and Other Acute Coronary
Syndromes
Robert J.=20 Zalenski, M.D., M.A.
Associate=20 Professor
Emergency=20 Medicine and Internal Medicine
Wayne = State=20 University
Detroit,=20 MI
4:00=20 p.m.
Databases=20 and Data Standardization
Clement J.=20 McDonald, M.D.
Distinguished=20 Professor of Medicine
Indiana=20 University School of Medicine
Director
Regenstrief=20 Institute for Health Care
Indianapolis,=20 IN
4:30=20 p.m.
Economic,=20 Social, and Psychological
Obstacles to=20 Achieving the National Heart
Attack Alert=20 Program=E2=80=99s Goals
Joseph P.=20 Ornato, M.D., F.A.C.C., F.A.C.E.P.
Professor and=20 Chairman
Department of=20 Emergency Medicine
Virginia=20 Commonwealth University
Medical College=20 of Virginia
Richmond,=20 VA
5:00=20 p.m.
Adjournment
Mark = S. Smith,=20 M.D.
5:00-
6:30=20 p.m.
Reception:=20 Lister Hill Center
Sponsored by=20 the Friends of the
National=20 Library of Medicine
W
EDNESDAY,=20
A
PRIL 15,=20 1998
New=20 Information Technology and the National Heart
Attack=20 Alert Program: Setting a 5-Year Agenda
Lister = Hill=20 Center
National Library=20 of Medicine
Bethesda,=20 Maryland
8:00 = a.m-5:00=20 p.m.
8:00=20 a.m.
Welcome and=20 Introduction
Donald=20 Lindberg, M.D.
Director
National=20 Library of Medicine
National=20 Institutes of Health
Bethesda,=20 MD
Claude Lenfant,=20 M.D.
Director
National Heart,=20 Lung, and Blood Institute
National=20 Institutes of Health
Bethesda,=20 MD
John = M.=20 Eisenberg, M.D.
Administrator
Agency for=20 Health Care Policy
and=20 Research
Rockville,=20 MD
Mark = S. Smith,=20 M.D.
Symposium=20 Program Chair
8:30-
10:00 a.m=20 Telehealth
Carole Gassert,=20 Ph.D., R.N.
Chair
Telehealth Task=20 Force
Capt. = David W.=20 Ferguson, M.C., U.S.N.
Vice=20 Chair
Telehealth Task=20 Force
8:30=20 a.m.
Telemedicine=20 in the Department
of = Defense=20 (DoD)
C. = Forrest=20 Faison III, M.D., C.D.R., M.C., U.S.N.
Director
Telemedicine=20 Program Office
U.S. = Department=20 of Defense
Bethesda,=20 MD
8:40=20 a.m.
Use of=20 Telemedicine Technology
to = Assist=20 Small Rural Hospitals
Napoleon=20 Knight, M.D.
Co-investigator
The = Carle Rural=20 Telemedicine Project
Medical=20 Director
Department of=20 Emergency Medicine
Carle = Foundation Hospital
Carle = Clinic=20 Association
Urbana,=20 IL
 x
8:50=20 a.m.
Ambulance-on-the=20 Web: A Cellular Phone-
Based,=20 Real-Time Video-Mobile Telemedicine
System for=20 Ambulance Transport
Marian P.=20 LaMonte, M.D., M.S.N.
Neurological=20 Program Director
The = Maryland=20 Brain Attack Center
University of=20 Maryland School of Medicine
Baltimore, MD=20
Panel on=20 Electronic Housecalls:
Jay = H. Sanders,=20 M.D.=E2=80=94Moderator
President
American=20 Telemedicine Association
Adjunct=20 Professor of Medicine
Johns = Hopkins=20 University School of Medicine
President and=20 CEO
The = Global=20 Telemedicine Group
McLean,=20 VA
9:00=20 a.m.
Behavioral=20 Issues in Initiating Use
of = Telemedicine Technology
Angelo A.=20 Alonzo, Ph.D.
Associate=20 Professor of Sociology
The = Ohio State=20 University
Columbus,=20 OH
9:10=20 a.m.
Electronic=20 Housecalls at Fort Gordon
Jack = A. Horner,=20 B.S.
Vice=20 President
Marketing and=20 Operations
Total = Access=20 Services, Inc.
Martinez,=20 GA
9:20=20 a.m.
Using=20 Televideo To Improve Patient
Outcomes
Linda = L.=20 Roman
President and=20 CEO
H.E.L.P.=20 Innovations, Inc.
Lawrence, KS=20
9:30=20 a.m.
Operational=20 Issues in Home Telemedicine
Charles Safran,=20 M.D.
Director,=20 Informatics and Advanced Technology
Beth = Israel=20 Deaconess Medical Center
Boston,=20 MA
9:40=20 a.m.
Audience=20 Discussion
10:00 a.m.=20 BREAK
10:20-
11:35=20 a.m. Education of Public, Patients,
and=20 Health Care Providers
James = M.=20 Atkins, M.D., F.A.C.C.
Chair
Education of=20 Public, Patients, and Health Care
Providers Task=20 Force
Mary = Jo=20 Deering, Ph.D.
Vice=20 Chair
Education of=20 Public, Patients, and Health Care
Providers Task=20 Force
10:20=20 a.m. Heart Attack Rapid Early Action
for Coronary=20 Treatment (REACT):
A=20 Community-Based Intervention
To = Reduce=20 Patient Delay in Seeking
Emergency=20 Care for Heart Attack Symptoms
Deborah J. K.=20 Alexander, M.A.
Program=20 Coordinator
Heart = Attack=20 REACT/Midwest
Division of=20 Epidemiology
School of=20 Public Health
University of=20 Minnesota
Minneapolis,=20 MN
10:35=20 a.m. Using a Comprehensive Web Site
To = Attract,=20 Identify, and Educate People
at = Risk for=20 Heart Attacks
Fran = Carl,=20 M.P.H.
Health=20 Education Manager
National=20 Member, Technology Group
Kaiser=20 Permanente
Oakland,=20 CA
10:50=20 a.m. Using New Technology To Deliver Consumer
Health=20 Information: Focus on Information
Kiosks
Victor=20 Strecher, Ph.D., M.P.H.
Professor and=20 Associate Director
Health Media=20 Research Laboratory
Comprehensive=20 Cancer Center
University of=20 Michigan
Ann = Arbor,=20 MI
11:05=20 a.m. Stimulating the Use of Information
Technology=20 in Support of Information
Seeking=20
J. = David=20 Johnson, Ph.D., M.A.
Dean
College of=20 Communications and Information
Studies
University of=20 Kentucky
Lexington,=20 KY
11:15=20 a.m. Audience Discussion
xi
11:35=20 a.m.-
12:45 p.m.=20 Medical Records Access
Reed = M.=20 Gardner, Ph.D.
Chair
Medical Records=20 Access Task Force
Joseph P.=20 Ornato, M.D.
Vice=20 Chair
Medical Records=20 Access Task Force
11:35 a.m.=20 Executive Summary G-7 CARDIO Project
Subproject=20 III: Improving Prevention,
Diagnosis,=20 and Treatment of Major
Cardiovascular=20 Diseases
Domenico=20 Cianflone, M.D., F.E.S.C. (Remote)
Operational=20 Coordinator
G-7 = CARDIO=20 Project
Instituto di=20 Cardiologia
Universita=20 Cattolica del Sacro Cuore
Rome, = Italy
Attilio Maseri,=20 M.D.
Director
Instituto di=20 Cardiologia
Universita=20 Cattolica del Sacro Cuore
Rome, = Italy=20
Enrico=20 Carlin
European Health=20 Telematics
Development=20 Manager
Medical=20 Customer Services Division
Hewlett Packard=20 Europe
Cernusco,=20 Italy
12:05 p.m.=20 Accessing Medical Records Using
the=20 Internet
John = D.=20 Halamka, M.D., M.S.
Associate=20 Professor of Medicine
Beth = Israel=20 Deaconess Medical Center
Executive=20 Director
Center for=20 Quality and Value
Harvard Medical=20 School
Boston,=20 MA
12:15 p.m.=20 Portable Information Storage Devices for
Immediate=20 Access to Cardiac Data at the
Point of=20 Care
Aziz = Boxwala,=20 M.B.B.S., Ph.D.
Instructor
Decision=20 Systems Group
Harvard Medical=20 School
Brigham and=20 Women=E2=80=99s Hospital
Boston,=20 MA
12:25 p.m.=20 Audience Discussion
12:45 p.m.=20 LUNCH
1:30-
2:50=20 p.m.
Diagnosis=20 and Treatment Decision Support
James = J.=20 Cimino, M.D.
Chair
Diagnosis and=20 Treatment Decision Support Task
Force
Robert J.=20 Zalenski, M.D., M.A.
Vice=20 Chair
Diagnosis and=20 Treatment Decision Support Task
Force
1:30=20 p.m.
Use of=20 Interactive Decision Trees
for Patients=20 To Decide About
the Need To=20 Call an Ambulance
J. = Robert Beck,=20 M.D.
Vice=20 President
Information=20 Technology Program
Baylor College=20 of Medicine
Texas = Medical=20 Center
Houston,=20 TX
1:40=20 p.m.
The Promise=20 and Problems
of = Community-Based Measurement
of = Biochemical Markers of Acute
Coronary=20 Syndromes
Robert H.=20 Christenson, Ph.D.
Professor of=20 Pathology
University of=20 Maryland School of Medicine
Baltimore,=20 MD
1:50=20 p.m.
Home 12-Lead=20 ECG Acquisition in Patients
at = High Risk=20 of Acute Myocardial Infarction
Robert J.=20 Zalenski, M.D., M.A.
Associate=20 Professor
Emergency=20 Medicine and Internal Medicine
Wayne = State=20 University
Detroit,=20 MI
2:00=20 p.m.
Access to=20 Electronic Medical Records
Zak = Kohane,=20 M.D.
Director of=20 Medical Informatics Program
Division of=20 Endocrinology
Children=E2=80=99s=20 Hospital
Assistant=20 Professor
Harvard Medical=20 School
Department of=20 Pediatrics
Boston,=20 MA
xii
2:10=20 p.m.
Understanding=20 Cognitive Processes
in = Medical=20 Decision Making
Vimla = L. Patel,=20 Ph.D.
Professor and=20 Director
Centre for=20 Medical Science Education
McGill=20 University
Montreal,=20 Quebec, Canada
2:20=20 p.m.
The Use of=20 Computerized
Electrocardiograph=20 To Provide Decision
Support for=20 the Use of Thrombolytic
Therapy as=20 Well as to Record Retrospective
Analysis=20
of = Performance for Feedback
Harry = P.=20 Selker, M.D., M.S.P.H.
Chief, Division=20 of Clinical Care Research
Director,=20 Center for Cardiovascular
Health Services=20 Research
Associate=20 Professor of Medicine
New = England=20 Medical Center
Tufts = University School of Medicine
Boston,=20 MA
2:30=20 p.m.
Audience=20 Discussion
2:50=20 p.m.
STRETCH=20 BREAK
3:00-
4:10=20 p.m.
Large-Scale=20 Databases
Dr. = William=20 Tierney
Chair
Large-Scale=20 Databases=E2=80=94Entry, Mining,
Research Task=20 Force
Dr. = Costas=20 Lambrew
Vice=20 Chair
Large-Scale=20 Databases=E2=80=94Entry, Mining,
Research Task=20 Force
3:00=20 p.m.
Use of the=20 Internet for Database Access
James = J.=20 Cimino, M.D.
Associate=20 Professor
Department of=20 Medical Informatics
Department of=20 Medicine
Columbia=20 University
Columbia-Presbyterian=20 Medical Center
New = York,=20 NY
3:10=20 p.m.
Continuous=20 Quality Improvement
in = the Care=20 of Patients With Acute
Myocardial=20 Infarction: The National
Registry of=20 Myocardial Infarction (NRMI)
Costas T.=20 Lambrew, M.D.
Director,=20 Division of Cardiology
Maine = Medical=20 Center
Portland,=20 ME
Professor of=20 Medicine
University of=20 Vermont
Burlington,=20 VT
3:20=20 p.m.
Validity and=20 Efficiency in Collecting Quality
Data in=20 Acute Myocardial Infarction
Patients: A=20 Comparison of the Cooperative
Cardiovascular=20 Project (CCP) and NRMI-2
Nathan R.=20 Every, M.D., M.P.H.
Assistant=20 Professor of Medicine
Division of=20 Cardiology
University of=20 Washington
Seattle,=20 WA
3:30=20 p.m.
The Cochrane=20 Collaboration and the Agency
for Health=20 Care Policy and Research=E2=80=99s
Evidence-Based=20 Practice Centers: The
Benefit of=20 Systematic Evidence Reviews in
Achieving=20 the Mission of the National Heart
Attack Alert=20 Program
Jane = D. Scott,=20 Sc.D., M.S.N, R.N.
Health Sciences=20 Researcher
National Study=20 Center for Trauma and EMS
University of=20 Maryland School of Medicine
Baltimore,=20 MD
3:40=20 p.m.
Use of=20 Clinical Trial Data To Support
Advances in=20 Emergency Cardiac Research
Harry = P.=20 Selker, M.D., M.S.P.H.
Chief, Division=20 of Clinical Care Research
Director,=20 Center for Cardiovascular Health
Services=20 Research
Associate=20 Professor of Medicine
New = England=20 Medical Center
Tufts = University School of Medicine
Boston,=20 MA
3:50=20 p.m.
Audience=20 Discussion
4:10=20 p.m.
Symposium=20 Assessment/Next Steps
Mark = S. Smith,=20 M.D.
Audience
4:45=20 p.m.
Adjournment
Donald=20 Lindberg, M.D.
Mark = S. Smith,=20 M.D.
1
Mark S.=20 Smith, M.D.
Chair, = Department of Emergency Medicine
Washington=20 Hospital Center
Clinical=20 Professor of Emergency Medicine
The = George=20 Washington University School
of = Medicine and=20 Health Sciences
Washington,=20 DC
This opening=20 welcome highlights the unique focus
of = the=20 symposium=E2=80=94information technology and its
application=20 to early recognition and treatment of patients
with=20 potential heart attacks=E2=80=94and the sponsoring = agencies
that made the=20 symposium possible.
Welcome to the=20 symposium, =E2=80=9CNew Information Technology
and = the National=20 Heart Attack Alert Program: Setting a
5-Year = Agenda.=E2=80=9D=20 This is a unique symposium for several
reasons. First=20 and foremost, the symposium is unique
because of its=20 subject: the intersection of the National Heart
Attack = Alert=20 Program (NHAAP) with the discipline of
medical=20 informatics.
The = NHAAP is=20 part of the National Heart, Lung, and Blood
Institute=20 (NHLBI). The mission of the NHAAP is to reduce
morbidity and=20 mortality from acute myocardial infarction
(i.e., = heart=20 attack) through early recognition and treatment of
these = patients.=20 The scientific basis underpinning the NHAAP
is = that if time=20 is reduced from the onset of symptoms of a
heart = attack to=20 delivery of the right therapy, lives can be
saved = and=20 substantial long-term cardiac damage can be
avoided. The=20 NHAAP has been in existence for 6 years and
has = been very=20 successful at improving emergency
department (ED)=20 management of the patient with symptoms
and = signs of a=20 heart attack.
The = NHAAP faces=20 three distinct challenges in the future:
how to = shorten=20 the access time to definitive care for the
patient with=20 symptoms and signs of a heart attack, how to
help = ensure that=20 the correct care is rendered once the patient
enters = the=20 health care system, and how to better measure the
outcome of these=20 new programs and interventions. Stated
succinctly, the=20 NHAAP must promote improved access,
improved=20 decisionmaking, and improved outcomes
measurement. The=20 agenda of this symposium is how
medical=20 informatics and new information technology can
help = further and=20 fulfill the mission of the NHAAP.
The = idea behind=20 the symposium is to shine the laser beam of
medical=20 informatics onto the challenges faced by the
NHAAP = and see=20 what can result. This is not a generic
emergency=20 cardiology conference nor a generic medical
informatics=20 conference. This is a marriage of two
disciplines, the=20 intersection of two worlds that are
orthogonal to=20 each other. The purpose of the symposium is
to = explore what=20 happens at the crossing point.
The = goal of the=20 symposium is to develop an agenda of
innovative ways=20 in which the domains of information
technology and=20 medical informatics can be used to help
fulfill the=20 NHAAP=E2=80=99s mission, and proposals for new
programs and=20 projects that could and should be undertaken
and = completed=20 in the next 5 years. The proposals must meet
only = two=20 criteria: they have to involve an application of new
information=20 technology and they have to be in the service of
the = mission and=20 goals of the NHAAP.
This = is also a=20 unique symposium because it brings together
an=20 unprecedented collaborative effort of three respected
Federal=20 agencies as cosponsors of the event: NHLBI; the
National=20 Library of Medicine; and the Agency for Health
Care = Policy and=20 Research. It is relatively unusual to have
dual = Federal=20 agency cosponsorship of a program such as
this, = much less=20 tri-agency cosponsorship.
Finally, the=20 symposium is unique because it utilized a novel
method for=20 generating ideas for presentation. The
symposium=20 issued an =E2=80=9CRFI,=E2=80=9D a =E2=80=9CRequest for Ideas,=E2=80=9D = over the
Internet. The=20 call for contributions was posted on the World
Wide = Web and on=20 various mailing lists and professional
societies. The=20 RFI garnered 96 responses, several of which
were = chosen for=20 presentation at the symposium.
The = symposium=20 is structured in two parts. The first part is a
series of six=20 tutorial presentations=E2=80=94three covering issues of
the = NHAAP and=20 three covering topics in informatics. The
second part=20 consists of presentations of specific ideas and
proposals for=20 how information technology can further the
mission of the=20 NHAAP. The purpose of these tutorials is to
give = everyone,=20 regardless of his or her background and
discipline, a=20 similar core level of knowledge. Those familiar
with = the NHAAP=20 need to be conversant with key issues in
medical=20 informatics, and those who are informaticians need
a = comprehensive=20 overview of the issues facing the NHAAP.
James = Atkins,=20 M.D., will present an overview of the
NHAAP=E2=80=94its=20 goals, missions, and accomplishments. Craig
Feied, M.D., will host a tour of the world of = the=20 next
generation=20 Internet. Paul Clayton, Ph.D., will speak about
the = thorny=20 issues of privacy, confidentiality, and security.
Robert=20 Zalenski, M.D., will cover the emergency
physician=E2=80=99s=20 diagnostic decisionmaking process. Clement
McDonald, M.D.,=20 will speak about the very important topic
of = databases=20 and data standardization. Joe Ornato, M.D.,
will = close the=20 tutorial section of the program with an
exposition=20 about the real, social, psychological, and
behavioral=20 obstacles to the NHAAP=E2=80=99s achievement of
success.
The = second part=20 of the symposium comprises idea
presentations.=20 There will be 24 presentations of ideas about
how = new=20 information technology can further the mission of
the = NHAAP. The=20 presentations are organized around five
informatics=20 domains: telehealth; education; medical records
access;=20 diagnostic and therapeutic decision support; and
large-scale=20 database entry, access, and mining.
I NTRODUCTION 2
The = specific=20 proposals that were selected for presentation
were = chosen by=20 five task forces comprised of experts in
informatics and=20 emergency cardiac care. Each task force
was = responsible=20 for the presentations in one of the five
informatics=20 domains. Each of the task force chairs and vice-
chairs = solicited=20 topics and speakers to present their visions
and = ideas for=20 the future.
There = are=20 certain premises underlying this symposium. The
first = is that=20 the single most important and influential
scientific=20 advance and major force that will change clinical
medical care in=20 the United States in the next 10 years is the
application of=20 new information technologies and the
discipline of=20 medical informatics. The application of new
information=20 technology to the practice of medicine will
transform the=20 landscape of the practice of medicine, more so
than = any=20 particular diagnostic or therapeutic advance. In
general, the=20 field of medicine has lagged behind other
disciplines and=20 industries in the application of that
technology. The=20 next 10 years will be dramatic and
dizzying=E2=80=94part=20 catch-up and part envelope-pushing.
This = symposium=20 is also borne out of the belief that
sometimes the=20 best way to change things is not to tackle a
problem head on,=20 but rather to approach it obliquely, to
think = laterally.=20 This is especially true when talking about
changing human=20 behavior, which is really at the core of the
challenge facing=20 the NHAAP. People must be able to decide
to = become=20 patients and seek care within minutes and not
within = hours of=20 their experiencing the symptoms of a heart
attack.
What = information=20 technology can do is change the
landscape and=20 effect changes in behavior that otherwise
would = take=20 countless frustrating hours of trying to reshape
old = patterns=20 with little success. New information technology
provides the=20 tools to approach human behavior problems
obliquely. Human=20 beings are complex systems. It is neither
desirable nor=20 possible to control 100 percent of the behavior
of = such a=20 complex system. The key is to find the 15-percent
leverage point=20 and let the remaining 85 percent self-organize
into = a=20 desirable configuration.
Permit me a=20 personal story that illustrates why knowledge
and = education=20 alone are probably not sufficient to change a
patient=E2=80=99s=20 behavior of not seeking care immediately after
symptoms of a=20 heart attack occur, and why a technological
solution may=20 well be one answer. Several months ago, I
awoke = at around=20 1:00 a.m. with terrible substernal burning.
By = all=20 measures, I constitute an educated consumer. I know
the = symptoms of=20 a heart attack. I teach the symptoms of a
heart = attack=20 and what to do if one experiences them. If
anyone had the=20 symptoms that I was having and asked me
what = to do, I=20 would tell them to either be driven to the
hospital by=20 their spouse or to call 9-1-1. What did I do? I
stumbled into=20 the bathroom, took some antacid, and went
back = to bed,=20 hoping that what I was having was the
heartburn of=20 esophagitis and not the pain of a heart attack. I
had a = fair=20 amount of anxiety at that time and was planning
what = would=20 happen if I was wrong. But I shied away from
the=20 inconvenience, embarrassment, and disruption of calling
an = ambulance=20 and going to a hospital ED.
What = would I=20 have liked at that moment? I would have
liked = to have=20 had a home ECG machine that I could have
strapped onto=20 myself, taken a tracing that would have been
compared=20 automatically to my previous ECG that was
stored, and=20 then be informed by the machine whether or not
ST-segment=20 elevation or some other change from my
previous=20 tracing was present. This is an information
technology=20 solution to what is perceived as a human
behavior=20 problem.
The = goal of=20 this symposium is to venture into uncharted
territory by=20 using new information technology to develop
innovative and=20 different approaches in the service of the
mission of the=20 NHAAP.
3 T UTORIAL P RESENTATIONS=E2=80=94 T
UESDAY,=20
A
PRIL = 14,=20 1998
T
HE=20
N ATIONAL H
EART=20
A
TTACK=20
A
LERT=20
P
ROGRAM:=20
I
TS =
M
ISSION=20
AND =
I
TS =
A CCOMPLISHMENTS
James M.=20 Atkins, M.D., F.A.C.C.
Medical=20 Director, Emergency Medicine Education
Professor of=20 Internal Medicine
Division of=20 Cardiology
University of=20 Texas, Southwestern Medical Center
Dallas,=20 TX
This tutorial=20 session provides an overview of the history and
goals of the=20 National Heart Attack Alert Program (NHAAP).
During = the past=20 50 years, the National Heart, Lung, and
Blood = Institute=20 (NHLBI) has made significant contributions
to = improving the=20 quality of care of people with heart
disease, lung=20 disease, and other blood vessel disease. There
have = been=20 significant reductions in mortality rates from
cardiovascular=20 disease. The number of deaths per 100,000
population=20 peaked in 1964 and 1965.
1
Since = that time=20 the
population has=20 significantly increased and the number of
individuals over=20 65 years of age has dramatically increased;
both = factors=20 would be anticipated to increase the number of
deaths = from=20 cardiovascular disease. In spite of the increased
population and=20 older age, the absolute number of deaths
decreased from a=20 high of 1.06 million to 918,000 in 1992.
Since = 1992, the=20 absolute number of deaths from
cardiovascular=20 disease has increased to the 940,000 to
955,000 range.=20 This represents an absolute decline of more
than = 100,000=20 deaths since 1972.
1
Such = a decline=20 in mortality
is = particularly=20 impressive considering that during this period
the = percentage=20 of the population over age 65 has
quadrupled. In=20 addition, the average life expectancy has
increased by 6=20 years, with 4 of those years most likely due
to = improvements=20 in cardiovascular care.
The = results from=20 a study done in Minneapolis/St. Paul,
Minnesota, 2
show = that from=20 1970 to 1990, both in-hospital
and=20 out-of-hospital mortality rates for men due to
cardiovascular=20 disease declined significantly. Mortality data
for = women also=20 indicated a decline but at a lower rate than
that = for men. In=20 1984, the total mortality for women equaled
that = for men for=20 the first time. Currently, more women than
men = die from all=20 cardiovascular disease.
3 Therefore,
addressing=20 cardiovascular disease in women will be one of
the = challenges=20 for the NHAAP in the future.
Time = is an=20 essential element in treating cardiovascular
disease, and a=20 similar course of action should be applied to
heart = attack=20 patients as is currently applied to trauma
patients. The=20 phrase, =E2=80=9Cthe golden hour,=E2=80=9D has been adopted = by
surgeons to=20 highlight the importance of providing treatment
quickly to=20 trauma victims. Prompt treatment is equally
critical for=20 cardiac arrest victims. There is a 0 to 2 percent
chance = that=20 cardiac arrest victims will survive if an
ambulance=20 arrives and paramedics defibrillate at
approximately=20 10 minutes. If a bystander administers
cardiopulmonary=20 resuscitation (CPR) within 10 minutes,
approximately 2=20 to 8 percent will survive. In addition, if
CPR = is=20 administered and the victim is defibrillated within
7 = minutes, 20=20 percent will survive. The survival rate
increases to 30=20 percent if CPR is administered within
2 = minutes and=20 defibrillation occurs within 4 minutes.
4
The = same trends=20 occur in patients with acute myocardial
infarction=20 (AMI) and acute coronary syndromes. With
medical=20 advances such as thrombolytic therapy and acute
interventional=20 angioplasty, lives can be saved. The earlier
the=20 intervention, the greater the number of lives that can be
saved. Data=20 from the Myocardial Infarction, Triage, and
Intervention=20 trial demonstrate a reduction in mortality from
12 = percent to=20 less than 1 percent if interventions are begun
within an hour=20 of the onset of symptoms. The average
amount of=20 muscle loss can be reduced from approximately
15 to = 18=20 percent, to less than 5 percent in 80 percent of the
patients. 5
Many = studies=20 have addressed the issue of delayed treatment
and = have=20 calculated mean and median times. Mean times are
not = as valuable=20 an indicator because they are skewed by
extremes (e.g.,=20 one person who waits 38 hours before
deciding to=20 seek treatment may offset many who decide
within an hour=20 of onset of symptoms). Although median
times = appear to=20 have improved, studies have indicated that
the = median time=20 is at least 2 hours. Transport times and time
spent = in the=20 emergency department before treatment are
considerations.=20 All of these times must be decreased in order
to = reduce=20 mortality rates.
Initiation=20 of the NHAAP
In = 1990, the=20 NHLBI sponsored a symposium that addressed
the = need for a=20 national education program focusing on
reducing=20 morbidity and mortality from AMI. The structure
for = this=20 program was developed at this symposium and the
NHAAP = was=20 launched in 1991. The initial goal of the
program was to=20 reduce morbidity and mortality from AMI,
including=20 sudden cardiac death, through rapid identification
and = treatment=20 of individuals with symptoms and signs of
AMI, = thereby=20 heightening the potential for an improved
quality of life=20 for AMI patients and those around them. This
goal = has since=20 been broadened to include patients with acute
coronary=20 syndromes.
At = the 1990=20 symposium, three phases of the treatment
process were=20 identified: patient/bystander recognition and
action,=20 prehospital action, and hospital action. The NHAAP
first = addressed=20 the prehospital and hospital phases because
implementing an=20 effective medical response system
appeared to be=20 a logical first step. For example, an effective
9-1-1 = system=20 should be in place nationwide before a public
campaign=20 encouraging patients to use that resource is
implemented.=20 Methods have not yet been identified for
intervening in=20 the patient/bystander phase.
4
Hospital=20 Phase
One of = the first=20 undertakings of the NHAAP was the 60
Minutes to=20 Treatment Working Group report headed by Drs.
Mark = Smith and=20 Costas Lambrew. The report, =E2=80=9CEmergency
Department:=20 Rapid Identification and Treatment of Patients
with = Acute=20 Myocardial Infarction,=E2=80=9D was published in 1994
in the = Annals=20 of Emergency Medicine and as a Government
Printing Office=20 (GPO) document by the same name in 1993.
The=20 recommendations from this report have been widely
quoted = and=20 accepted. In this report, critical timepoints were
identified at=20 varying stages where delays from door-to-drug
time = are likely=20 to occur in the emergency department. The
report = provides=20 recommendations on how to reduce delay at
these = stages=20 utilizing a continuous quality improvement
process. Each=20 step of this complex process requires
intervention.=20
Several studies=20 have demonstrated a substantial reduction in
the = time that=20 elapses from the patient=E2=80=99s arrival at the
emergency=20 department until thrombolytic therapy is
administered. In=20 1988, the average door-to-drug time was
144 = minutes; in=20 1991, 72 minutes; in 1993, 46 minutes; and
in = 1994, 21=20 minutes. Some hospitals have reported times as
low as = 15 or 16=20 minutes (verbal presentations).
Patient/Bystander=20 Recognition and Action Phase
A = report from=20 the NHAAP targeting the patient/bystander
recognition and=20 action phase was published in 1993 as a
GPO = document=20 titled =E2=80=9CPatient/Bystander Recognition and
Action: Rapid=20 Identification and Treatment of Patients with
Acute = Myocardial=20 Infarction.=E2=80=9D This report identified a
number = of key=20 factors that influence the time that elapses
before = a patient=20 decides to seek treatment. Factors such as
lay = consultation=20 with a spouse or friend, medical
consultation,=20 and self-treatment may increase the delay time.
Lay = consultation=20 occurs 93.2 percent of the time, with the
greatest delay=20 being caused by consultation with a spouse,
followed by=20 consultation with a friend, a coworker, and a
stranger. If a=20 spouse is consulted, denial on the part of both
spouses and an=20 unwillingness for them to confront one
another may=20 contribute to a lengthy delay. Past medical
history, such as=20 a history of diabetes mellitus or stable
angina, has been=20 shown to be a factor in increased delay.
With = the latter,=20 the patient might have difficulty determining
whether the=20 symptoms were indicative of angina or a more
acute = condition.=20 Demographic factors that have been shown
to = increase the=20 delay time include older age, being female,
African-American=20 race, low socioeconomic status, and a
low = educational=20 level. Also, daytime onset or being at home
during = onset of=20 symptoms may contribute to delay.
Certain=20 conditions have been shown to decrease delay time;
for = instance,=20 whether the patient recognizes the symptoms as
cardiac in=20 origin, whether the pain is severe, or whether the
patient is=20 hemodynamically unstable. Being at work during
the = onset of=20 symptoms also decreases the delay time. Factors
that = do not=20 appear to have an effect on the delay time
include=20 personality type, a prior cardiac event, and having
heart = failure,=20 hypertension, or known coronary artery
disease.=20 Education about cardiovascular disease hopefully
decreases the=20 delay time, but results to date on this aspect
have = been=20 mixed. The results of the NHLBI-funded Rapid
Early = Action=20 for Coronary Treatment trial should provide
more=20 guidance.
Prehospital=20 Phase
The = emergency=20 medical services (EMS) system is a
complex system=20 with many components. These components
include=20 discovery, the bystander=E2=80=99s reaction, access into = the
EMS = system,=20 dispatching, on-scene activities (e.g.,
assessing,=20 resuscitating, and extricating the patient), and
transportation=20 to the hospital.
In = 1995, the=20 NHAAP published three articles addressing
prehospital=20 areas including 9-1-1, staffing and equipping
EMS = systems,=20 and emergency medical dispatching.
Currently, 86=20 percent of the U.S. population is covered by
9-1-1 = systems.=20 Although many larger cities and suburbs
have = these=20 systems, they are less likely to be found in rural
areas. In fact,=20 a significant portion of the country does not
have = complete=20 9-1-1 coverage. When the NHAAP started in
1991, = only 2=20 States (Rhode Island and Delaware) had
border-to-border=20 coverage; today that number has been
expanded to 20=20 States. But 30 States still do not have
statewide 9-1-1=20 services and most 9-1-1 centers do not have
enhanced 9-1-1.=20 This technology enables a caller=E2=80=99s phone
number to=20 appear on a center=E2=80=99s computer screen as the call
is = answered,=20 followed shortly by the caller=E2=80=99s address that is
supplied from a=20 database. This enhancement overcomes the
problem of=20 dispatchers not being able to locate an
emergency=20 caller who is unable to give a correct address, or
callers who=20 become unconscious.
Another problem=20 is that the majority of EMS systems do not
have=20 dispatchers who are trained in medical protocols.
Although=20 computer-assisted systems are available, most
cities do not=20 use them. Recommendations on this and on
ambulance=20 staffing and equipment were made in the
NHAAP = paper=20 addressing staffing and equipping of EMS
systems. One of=20 the recommendations=E2=80=94automated external
defibrillation=E2=80=94has=20 been widely adopted. Other NHAAP
publications=20 related to EMS include an article on the need
for = community=20 planning, a 1997 article on diagnostic
technologies,=20 and several versions of a paper on identifying
and = treating=20 high-risk patients.
Challenges
Many = challenges=20 remain for the NHAAP, posed by the
complexity of=20 the emergency medical care process and the
intervention=20 options. The process not only involves many
steps = but also=20 many components, such as patients and
bystanders, the=20 EMS system, and the hospital. These
components each=20 involve technology, but the technologies
are = not always=20 integrated into a systems approach. The
5
intervention=20 involves both psychological and sociological
considerations.=20 One-third of patients with an AMI do not
display even=20 moderately classical symptoms, and the
symptoms are=20 difficult for patients to differentiate from
indigestion.=20 There is not a straightforward, simple message
for = the patient,=20 family, and bystanders; rather, the goal is to
enable = the=20 patient to recognize a range of symptoms and
then = initiate a=20 chain of events that involves multiple players.
Health = maintenance organizations (HMOs) and preferred
provider=20 organizations (PPOs) are concerned about the costs
to the = health=20 care system for patients with no disease, and
there = also are=20 psychological costs to these patients. There
also = are=20 concerns about how to frame messages given to
women = to=20 encompass their increasing risk as they age. And
to = target=20 high-risk patients, one must be able to identify
them; = one-third=20 of people who experience an AMI or sudden
cardiac arrest=20 do not know that they have cardiovascular
disease prior to=20 the event. Other challenges include
countermeasures=20 from many HMOs and PPOs (such as
calling the=20 nurse/screener first) and areas that need
improvement,=20 including the wider use of enhanced 9-1-1
systems,=20 training of EMS dispatchers, and improved triage
in = emergency=20 departments.
Future=20 Goals
In = 1996, upon=20 completion of its first 5 years, the NHAAP
held a = retreat=20 meeting at which goals were developed for the
future. These=20 goals included the following: to use evidence-
based = evaluation=20 of diagnostic technologies, strategies, and
protocols to=20 identify quickly patients with acute coronary
syndromes in the=20 emergency department; to develop a health
care = system and=20 community-planning approach to integrate
the = messages of=20 the NHAAP; to apply new information
technologies to=20 the goals of the NHAAP; and to improve the
education of=20 professionals, high-risk patients, other patients,
and = bystanders.=20 In addition, the NHAAP plans to address the
problem of=20 patients discharged from emergency departments
who = have been=20 ruled out for an AMI but who may have
future = cardiac=20 events. Presentation to the emergency
department=20 provides an opportunity to train patients to
recognize=20 symptoms and to encourage them to call for help
if = these=20 symptoms occur.
Hopefully,=20 achieving these goals will help further reduce the
time = from onset=20 of symptoms to treatment and, therefore,
reduce = the loss=20 of lives. Technology can contribute to these
goals = by=20 facilitating the development of an organized,
integrated=20 system that will provide the fastest, most effective
assistance to=20 patients.
References
1. = National=20 Heart, Lung, and Blood Institute. Fact Book Fiscal
Year=20 1997. Bethesda, MD; February 1998.
2. = McGovern PG,=20 Pankow JS, Shahar E, et al.: Recent trends in
acute = coronary=20 heart disease: Mortality, morbidity, medical
care, = and risk=20 factors. N Engl J Med 1996;334:1884.
3. = Heart=20 Association Fact Book, 1998.
4. = Cummins RO=20 (ed). Textbook of Advanced Cardiac Life
Support.=20 Dallas: American Heart Association, 1994,
p=20 4.2.
5. = Weaver WD,=20 Cerqueira M, Hallstrom AP, Litwin PE, Martin
JS, = Kudenchuk=20 PJ, Eisenberg M, for Myocardial Infarction
Triage and=20 Intervention Project Group. Prehospital-initiated vs.
hospital-initiated=20 thrombolytic therapy: the Myocardial
Infarction=20 Triage and Intervention Trial. JAMA
1993;270(10):1211. T ELECOMMUNICATIONS=20 AND THE N EXT G
ENERATION=20
I NTERNET
Craig Feied,=20 M.D.
Director of=20 Informatics
Department of=20 Emergency Medicine
Washington=20 Hospital Center
Associate=20 Clinical Professor of Emergency Medicine
The = George=20 Washington University School of Medicine and
Health=20 Sciences
Washington,=20 DC
This=20 tutorial presentation provides an overview of the past,
present, and=20 future of telecommunications=E2=80=94including the
next=20 generation Internet.
The=20 Evolution of Connectivity
Before the=20 advent of the telephone, information exchange
was, = with few=20 exceptions, limited by the need for face-to-
face=20 conversations or the delivery of written documents.
Lengthy delays=20 (a high latency) attended information
exchange over=20 any significant distance, and these delays
largely defined=20 the communal life of a country in which
news = of a=20 presidential election could take several weeks to
cross = the=20 continent. Technological innovations such as the
railroad and=20 the telegraph allowed a more speedy exchange
of = written=20 information, but it was not until the widespread
adoption of the=20 telephone that person-to-person interaction
at a = distance=20 became possible. The telephone eliminated
some = of the=20 natural constraints of distance and catapulted
the = world into=20 the first era of rich information exchange.
This = ability to=20 exchange information can be called
=E2=80=9Cconnectivity=E2=80=9D=E2=80=94the=20 ability to give to other people
information=20 that we possess, and to receive from other
people=20 information that we want. The more rapidly we can
exchange=20 information, the richer the types of information
that = we can=20 exchange, and the larger the universe of = people
6
with = whom we can=20 exchange information, the greater the
degree = of=20 connectivity we enjoy. Since the introduction of
the = telephone,=20 the maximum achievable degree of
connectivity has=20 grown dramatically, and this growth has
had a=20 transforming effect on a society that has become
increasingly=20 dependent on information transfers for the
economic,=20 political, business, and personal vitality of its
citizens.
Initially, the=20 telephone created small islands of connectivity
with = large gaps=20 between them. In the 1870=E2=80=99s, the earliest
telephone=20 installations connected up to 48 subscribers on a
single = exchange.=20 All subscribers had to be within a fairly
close = geographic=20 proximity, and there was no connection
between any two=20 exchanges. In larger cities, many different
companies=20 operated telephone exchanges, and calls could
not be = connected=20 between competing companies. End-to-end
voice=20 connectivity came about only when one telephone
company=20 succeeded in obtaining a monopoly, eventually
unifying the=20 entire country within a single proprietary
standard for=20 telephone connectivity. It is now difficult to
imagine a life=20 without telephones, but the public was
initially slow=20 to accept them. The first branch exchanges
were = installed=20 in 1877, and it was not until 50 years later
that = the number=20 of telephone subscribers reached 20 million.
In = contrast to=20 hard-wired telephony, mobile cellular
telephony was=20 adopted very rapidly. Once considered so
unlikely a=20 concept that they appeared only in science-fiction
novels, 20=20 million cellular telephones were installed in the
first = 9 years=20 after their introduction in 1985. At the
beginning of=20 1998 there were more than 50 million cellular
telephones in=20 use in the United States. Rapid growth in the
number = of=20 cellular telephones continues to be fueled by
portable units=20 that are ever smaller and ever cheaper, a trend
that = is=20 exemplified by the recent introduction of a cellular
telephone built=20 into a wristwatch.
Just = as=20 telephony created a tremendous network of voice
connectivity,=20 other devices have piggybacked onto the
telephone=20 network to add richness and complexity to that
connectivity.=20 The first of these was the facsimile machine,
permitting the=20 rapid transmission of a document between
two = standard=20 units that can be attached to any telephone
instrument=20 anywhere in the world. When facsimile
transmission=20 first became available, few people recognized
its = value and=20 importance, but the speed and convenience of
facsimile=20 transmission eventually became widely
recognized.=20 Today, facsimile equipment is nearly ubiquitous,
adding = significantly to the richness of global connectivity as
compared to=20 voice-only connections.
During = the=20 1970=E2=80=99s and 1980=E2=80=99s, computer systems were = also
being=20 piggybacked onto the voice communications network.
With a = computer=20 at each end of the line, not only could a
facsimile of a=20 document be transmitted over great distances,
but = the editable=20 content of the document could also be sent
as a = computer=20 file. Large numbers of computer enthusiasts
subscribed to a=20 tremendous network of public-access file
exchange=20 systems known as =E2=80=9Cbulletin board systems.=E2=80=9D = The
system was=20 effective, but cumbersome, because a direct
telephone=20 connection between the two machines was
needed. Like an=20 isolated telephone branch exchange, a
bulletin board=20 system was an island of connectivity because
it = could serve=20 only a few people in an isolated geographic
area = and could=20 not route files or messages to people using
other = systems=20 in other areas.
In = the 1990=E2=80=99s=20 the widespread adoption of TCP/IP, a standard
for = exchanging=20 messages between computers from different
manufacturers,=20 facilitated the development of local-area
networks (LANs)=20 of interconnected computers and then of
connections=20 between networks to form =E2=80=9Cinternetworks.=E2=80=9D = For
the = first time,=20 information could be freely exchanged
between any two=20 computers connected to any part of a
far-flung=20 network of networks, known collectively as =E2=80=9Cthe
Internet.=E2=80=9D
The = number of=20 Internet-connected computers grew from
0 to = 20 million=20 in the first 4 years of its existence. As of July
1998, = there=20 were nearly 37 million host computers
registered as=20 accessible in the Internet domain name system
(DNS), with=20 many more host systems connected to the
Internet but=20 not registered for public access. This level of
connectivity is=20 without precedent: for the first time, it has
become possible=20 to obtain information on virtually any topic
at = exactly the=20 place and time at which the information is
needed. If the=20 Internet continues to grow at this same
phenomenal=20 rate, by the year 2002 there will be one
computer=20 connected to the Internet for every man, woman,
and = child in=20 the United States.
Underlying=20 Communications Technologies
Whether by=20 voice, by facsimile, or by computer connection,
modern=20 communication depends upon the underlying
telecommunications=20 technologies that link together
buildings,=20 cities, states, and countries. The power of a
particular=20 mechanism for information exchange depends
partly on the=20 highest speed with which a very small amount
of = information=20 can be sent (the latency) and partly on the
amount of=20 information that can be sent per unit time (the
bandwidth). An=20 overnight delivery of a shipping crate full of
documents has a=20 high latency (it takes an entire day for a
single bit of=20 information to be delivered) but also has a high
bandwidth (a=20 huge amount of information can be transferred
per = day). A=20 flashlight used by a Girl Scout to transmit Morse
code = has a low=20 latency (the first bit of information is
delivered at=20 the speed of light) but also has a low bandwidth
(very = little=20 information can be transferred per day). With the
exception of=20 satellite connections, modern telecommuni-
cations=20 modalities typically have a low latency, but the
amount of=20 bandwidth varies greatly from modality to
modality. 7
Voice Origins=20 of Digital Communications Circuits
Today = many=20 different kinds of information are delivered
through=20 telecommunications links, but these links are not
optimized for=20 the traffic they carry because they were
originally=20 developed solely to carry human voices. The
microphone=20 inside a telephone handset converts a sound
wave = into a=20 continuously varying voltage (an analog signal).
The = sound waves=20 from a normal human voice oscillate up to
4,000 = times per=20 second; thus, an analog voltage signal
representing=20 that voice will have up to 4,000 cycles per
second, or 4=20 kilohertz (kHz). This 4 kHz signal can be sent
through the=20 wires directly as an analog (continuously
varying)=20 voltage, or it can be digitized and sent through the
wires = as a=20 series of high and low voltages representing
digital ones and=20 zeros (a digital signal). End-to-end analog
connections are=20 rare today: in most cases an analog voice
signal = goes from=20 the telephone handset through a =E2=80=9Clocal
loop=E2=80=9D to the=20 telephone company branch exchange, where it
is = converted=20 from an analog to a digital signal and passed
along = to a=20 destination branch exchange. At the destination
exchange, the=20 signal is converted back into analog form and
sent = through the=20 =E2=80=9Clocal loop=E2=80=9D wires to the telephone handset
at the = other=20 end.
When a = 4 kHz=20 human voice is converted from an analog to
a = digital=20 signal, it requires two digital samples per cycle and
eight = levels of=20 measurement per sample; thus, 4 x 2 x 8 =3D
64,000 = bits of=20 information per second (64K bps). A single
64K = bps channel=20 is often referred to as a =E2=80=9Cbasic channel=E2=80=9D = or
=E2=80=9CB-channel,=E2=80=9D and=20 most high-bandwidth connections are
defined as=20 multiples of this channel. For example, an
Integrated=20 Services Digital Network (ISDN) line carries two
digital=20 B-channels, or a total of 128K bps. A leased T-1 line
offers = 24=20 B-channels plus one-eighth of a channel used for
timekeeping, for=20 a total of 1.544 million bits per second.
When = an=20 =E2=80=9Cend-to-end=E2=80=9D digital circuit is available to = connect
two = computers,=20 all of the potential bandwidth of the circuit
is = available for=20 sending and receiving information. When
some = part of the=20 circuit must carry an analog signal,
however, some of=20 the potential bandwidth is lost=E2=80=94the
computer=E2=80=99s=20 digital signal is converted from digital (at the
sending end) to=20 analog (at the telephone connection) to
digital (at the=20 sending telephone company branch exchange),
to = analog (at=20 the receiving telephone company branch
exchange) and=20 finally back to digital (at the receiving
computer). A=20 conversion from digital data to an analog
signal = and back=20 to digital again is known as modulation-
demodulation,=20 hence the term =E2=80=9Cmodem.=E2=80=9D Similarly, when an
analog = voice=20 signal is converted to a digital signal and back
again, = the term=20 =E2=80=9Ccoder-decoder=E2=80=9D or =E2=80=9Ccodec=E2=80=9D is = used.
With a = pure=20 digital connection, the 64K bps of a B-channel
can = carry a full=20 64K bps of data, but the same channel used
with a = modem at=20 each end carries much less data. Modem
technology=20 originally permitted only 120 bps through a
B-channel, but=20 evolved over a 30-year period to support
higher = transmission rates over the same basic channel.
Today = a=20 standard B-channel modem can transmit or receive
38,400 bps when=20 connected to another modem, and data
rates = can go as=20 high as 56,000 bps in special situations
where = one end=20 of the circuit is digitally terminated.
High-Speed=20 Digital Landlines
Under = the right=20 conditions, a four-wire copper wire circuit
can = be made to=20 carry as many as 1 billion bps (gigabit, or
1G = bps), but=20 this type of transmission requires high-quality
lines = and=20 expensive line conditioners and can run only short
distances=20 before signal regeneration is necessary. In contrast,
light = passing=20 along a single fiber of spun glass can carry
2.4 = billion bps=20 and the fiber can run 200 miles between
regenerators.=20 For this reason, high-speed network traffic
today = is=20 carried almost exclusively over fiberoptic cable.
Compared to new=20 copper wires, fiberoptic cable is very
cost-effective.=20 When fiber cable is made, many fibers are
bound = together=20 in a single bundle. Only a single fiber strand
is = needed to=20 make a high-bandwidth connection, and there is
a = large amount=20 of still-unused capacity (known as =E2=80=9Cdark
fiber=E2=80=9D)=20 remaining in most cable runs. Advances in a
technique known=20 as =E2=80=9Cwave-division multiplexing=E2=80=9D have
increased=20 per-strand bandwidth by making it possible to
send = and=20 receive data encoded into many different
wavelengths of=20 light through the same fiber strand all at the
same = time. The=20 amount of data that can be carried by a
single fiber=20 strand has been increasing rapidly, and is likely
to = continue to=20 increase over the next few decades.
With = the same=20 amount of data being squeezed into narrower
and = narrower=20 parts of the visible spectrum, there is no
theoretical=20 limit to the potential future data-carrying
capacity of=20 fiber connections. Today, commercially installed
synchronous=20 optical network (SONET) systems carry voice
and = data=20 traffic in a digital format at channel speeds up to
2.4G = bps.=20 Laboratory prototype SONET channels now
handle up to a=20 trillion bps (1T bps).
In = the United=20 States, fiberoptic cables now connect every
main = telephone=20 company exchange as well as many local
branch=20 exchanges and commercial buildings. Unfortunately,
it = has not=20 proven cost-effective to run new glass fiberoptic
cables the=20 =E2=80=9Clast mile=E2=80=9D into every home and office. For = this
reason, there=20 remains a great interest in new technologies
that = can=20 improve data transmission over existing copper
wires. The=20 principal =E2=80=9Clast mile=E2=80=9D solutions being = deployed
today = are=20 =E2=80=9Ccable modems=E2=80=9D that can transmit data over = the
coaxial cable=20 used for cable television and =E2=80=9Cdigital
subscriber=20 lines=E2=80=9D (DSL) that use special conditioners to
boost = the=20 digital speed of existing copper loops belonging to
the = local=20 telephone system. Under the proper conditions,
each = of these=20 can be a cost-effective solution carrying
anywhere from=20 750K bps to 64M bps of data.
Computer=20 Connectivity Over Cellular Telephony
When = digital=20 data are sent over wireless radiotelephone
circuits, there=20 is another added layer of complexity.
8
The = radio=20 frequency spectrum is a scarce resource, and
current cellular=20 technologies use a variety of tricks to
squeeze every=20 possible bit of voice traffic into the limited
bandwidth. The=20 first trick used is just to throw away data: so
long = as the=20 speaker is recognizable, voice quality can be
sacrificed to=20 save bandwidth. A pure analog cellular
telephone voice=20 channel can carry only about 48K bps,
rather = than the=20 64K bps carried by an analog landline. Other
digital and=20 hybrid systems encode and compress a voice
signal = using=20 algorithms that can deliver a recognizable voice
using = much less=20 bandwidth. The resulting cellular channels
are = adequate for=20 voice transmission, but when used with a
computer and a=20 modem, they provide much less bandwidth
than a = voice-quality landline. Many applications that work
well = when a=20 laptop is connected through a modem and a
landline will=20 fail miserably when the same laptop is
connected=20 through a modem and a cellular phone
connection. This=20 bandwidth problem severely limits the
connectivity of=20 mobile systems such as those in ambulances
and = helicopters.=20 Of all the digital systems, only Code
Division=20 Multiple Access (CDMA) systems can deliver
usable = data=20 rates over a single channel. Unfortunately,
CDMA = is not a=20 good final solution for mixed voice and data
transmission,=20 because adequate data bandwidth is only
obtained at the=20 expense of other cellular users who share the
channel. Figure=20 1. shows the maximum amount of digital
data = that can be=20 delivered over different types of cellular
channels=20 compared to the amount that can be delivered over
a=20 landline.
Satellite=20 Telecommunications
Communications=20 satellites may be grouped into three broad
categories=20 according to the height at which they orbit the
Earth. = The=20 period of an orbit depends upon its distance from
the = Earth: the=20 greater the distance from the Earth, the slower
the = orbit. At a=20 distance of 22,241 miles above the Earth=E2=80=99s
surface, the=20 orbital period is precisely 24 hours, and the
orbit = is called=20 =E2=80=9Cgeosynchronous.=E2=80=9D If a satellite in a
geosynchronous=20 orbit travels in exactly the same direction
as = the Earth=E2=80=99s=20 own rotation, the satellite can remain above
the = same spot=20 on the Earth at all times. Such an orbit is
called=20 =E2=80=9Cgeostationary.=E2=80=9D Because they are so high above = the
Earth, a single=20 geostationary Earth orbit (GEO) satellite can
transmit to=20 nearly an entire hemisphere. Despite this
advantage, GEO=20 satellites are poorly suited to voice
communications=20 and other real-time applications because
data=20 transmission, signaling, and receipt acknowledgment
over = this great=20 distance introduces an overall round-trip
latency of=20 approximately one-half second. The high power
and = large=20 satellite dish required to send a signal from the
earth = up to a=20 receiver at so great a distance also makes GEO
satellites a=20 poor choice for two-way communications. GEO
systems have=20 found their principal niche in applications that
require one-way=20 broadcasts of information, such as digital
television and=20 digital data downlinks.
Medium Earth=20 orbit (MEO) satellites orbit at 6,000=E2=80=9312,000
miles = above the=20 Earth, and although they move with respect
to = the ground,=20 each satellite can cover a lot of the earth and
can = stay within=20 view of a single ground-receiving station for
a = long time. A=20 moderately large satellite dish and a lot of
power = are=20 required to transmit from the ground to an MEO
satellite, but=20 round-trip signal latency is acceptable for voice
communications.=20 Inmarsat and TRW mobile communi-
cations systems=20 use MEO satellites both for voice and for
private network=20 data transmissions at up to 9,600 bps over a
standard=20 channel.
Low = Earth orbit=20 (LEO) satellites have a short signal latency
and = are close=20 enough (100=E2=80=93600 miles above the Earth) that
hand-held=20 cellular telephones can connect to them using a
low-power=20 transmitter and a standard antenna. Because they
are = so close to=20 the Earth, LEO satellites must travel about
17,000 miles=20 per hour to remain in orbit, circling the globe
about = every 90=20 minutes. As of this writing, there are no
global=20 commercial telephony services operating via LEO
satellites, but=20 the first such system (=E2=80=9CIridium=E2=80=9D) has 66 = active
LEO = satellites=20 already in orbit 420 miles high and is
scheduled to=20 offer cellular satellite telephone service
Standard=20 landline:
=E2=80=A2 56K bps with=20 standard modem in optimum configuration
=E2=80=A2 64K bps if=20 digital
Analog=20 cellular connection:
=E2=80=A2 33.4K bps with=20 typical modem
=E2=80=A2 48K bps=20 maximum with a custom-designed modem
NA-TDMA
(North=20 American Time Division Multiplexing):
=E2=80=A2 8K bps if a=20 standard modem is used with a cell phone
containing a=20 codec
=E2=80=A2 16K bps if the=20 system permits a digital signal
Figure 1.=20 Basic Channel Bandwidth
 EU-TDMA
(European=20 Time Division Multiplexing):
=E2=80=A2 13K bps if a=20 standard modem is used with a cell phone
containing a=20 codec
=E2=80=A2 33K bps if=20 the system permits a digital signal
CDMA (Code=20 Division Multiple Access):
=E2=80=A2 13K bps if a=20 standard modem is used with a cell phone
containing a=20 codec
=E2=80=A2 48K bps if=20 the system permits a digital signal
 9
worldwide=20 commencing in the last quarter of 1998.
Prototype=20 satellites have also been launched for another
system = (=E2=80=9CTeledesic=E2=80=9D) that proposes to place 288 = satellites
into = orbit 339=20 miles above the ground, linking them together
to = create an=20 =E2=80=9CInternet in the sky.=E2=80=9D Teledesic ground = stations
will = use small=20 roof-mounted satellite dishes that will give
each = home or=20 office a high-speed wireless connection
carrying 64M=20 bps, a capacity 2,000 times greater than a
current analog=20 modem and 6 times greater than most
hardwired LAN=20 connections.
Figure = 2. shows=20 the time it would take to download the
16 = megabyte=20 Netscape 4.0 Web browser via several different
types = of=20 connections.
Growth and=20 Congestion of the Internet
The = power of the=20 Internet to transform society cannot be
overstated. In a=20 few short years, this network of networks that
allows = an=20 immediate exchange of data between any two
computers nearly=20 anywhere in the world has become not only
ubiquitous, but=20 essential to the functioning of many industries.
The = Internet as=20 it exists was never planned nor developed by
any = formal=20 group. It grew upon the scaffolding of existing
telephone and=20 computer connectivity, and came into
existence mostly=20 as the result of the independent efforts of
scattered=20 visionaries working in laboratories at universities
and = colleges,=20 in the Government and military, and in private
industry. Two=20 decades of underlying work was largely
funded by=20 military and government grants supporting
research in a=20 variety of seemingly unrelated areas, much of
it = aimed at=20 providing distant connections to supercomputers
and = other=20 shared resources across the country.
The = United=20 States has benefited greatly from a position of
leadership in=20 worldwide connectivity. Increased production
and = an improved=20 competitive position in the worldwide
economy are=20 just two of the many ways in which the
connectivity of=20 the Internet has made a difference.
Unfortunately,=20 some of the benefits of this connectivity are
being = eroded=20 away by the very success of the Internet. Other
countries are=20 increasing their presence on the Internet and
are = moving=20 forward in ways that may well erode the
commanding lead=20 the United States has enjoyed in
technology=20 fields. Private enterprise and governmental
agencies are=20 afraid that the country will lose its competitive
edge = now that=20 the rest of the world has become heavily
committed to=20 the Internet.
At = the same=20 time, the network bandwidth available for
research and=20 development has been nibbled away by private
and = commercial=20 users. University, Government, private lab,
and = military=20 researchers once enjoyed the exclusive use of a
network with=20 bandwidth more than ample to meet the = needs
Connection Speed
Download=20 Time
Modem = (circa=20 1975)
120 = bps=20
12 = days=20
Modem = (circa=20 1980)
300 = bps=20
5 = days=20
Modem = (circa=20 1984)
1,200 = bps=20
30 = hours=20
TRW = or Inmarsat=20 satellite modem
9,600 = bps=20
4 = hours=20
Digital cell=20 phone with modem
13K = bps=20
2.7 = hours=20
Modem = (circa=20 1992)
28.8K = bps=20
75 = minutes=20
V.34 = modem=20 (circa 1994)
38.4K = bps=20
56 = minutes=20
V.90 = modem=20 (circa 1998)
56K = bps=20
38 = minutes=20
Dual = ISDN=20
128K = bps=20
17 = minutes=20
Cable = TV modem=20
750K = bps=20
3 = minutes=20
T-1 = leased line=20
1.544M bps=20
83 = seconds=20
Ethernet LAN=20
10M = bps=20
13 = seconds=20
T-3 = leased line=20
45M = bps=20
2.8 = seconds=20
VDSL = digital=20 subscriber line
54M = bps=20
2.3 = seconds=20
Teledesic=20 satellite connection
64M = bps=20
2 = seconds=20
Fast = ethernet=20 LAN
100M = bps=20
1.3 = seconds=20
Gigabit=20 ethernet LAN
1G = bps=20
0.13 = second=20
SONET = OC-48=20 fiber-optic channel
2.4G = bps=20
0.05=20 second
Figure 2.=20 Time to Download 16-megabyte File
 10
of any = achievable project. Today, new applications demand
more = and more=20 bandwidth as research capabilities continue
to = improve, yet=20 researchers find that their high-speed
connections to=20 important resources are slowed to a crawl as
Internet traffic=20 continues to rise beyond all expectations. At
the = current rate=20 of growth, by the time the next generation of
applications are=20 released there will be so much Internet
traffic=20 congestion that the system would hardly be functional
at=20 all.
Next=20 Generation Internet and Internet 2
The = recognition=20 of certain shortcomings in the current
Internet and the=20 desire for continued growth and leadership
in = technology=20 have sparked several initiatives to underwrite
the = development=20 of a =E2=80=9Cnew Internet.=E2=80=9D Just as the current
Internet has led=20 to profound and unforeseen changes in our
ways = of life, a=20 next generation version might permit the
development of=20 as-yet unimagined applications that could
continue to=20 redefine modern life as the current Internet does
today. = Two main=20 initiatives are underway in support of
many = different=20 groups working to bring this future vision to
life. = The first=20 is known as the =E2=80=9CNext Generation Internet=E2=80=9D
(NGI)=20 initiative, and the second is known as the Internet 2,
or I2, = initiative. The two share many of the same goals but
have = different=20 constituencies and complementary approaches
to the = goals.
NGI
The = NGI=20 initiative (http://www.ngi.gov/) is a multi-agency
Federal research=20 and development program with three goals:
to = connect=20 national laboratories and universities with
high-speed=20 networks that are 1,000 times faster than today=E2=80=99s
Internet, to=20 promote experimentation with the next
generation of=20 networking technologies, and to demonstrate
new = applications=20 that require advanced networking to meet
important=20 national goals and missions.
In an = effort to=20 replicate the dynamics that led to the creation
of the = current=20 Internet, a 1998 Federal investment of
$100 = million was=20 allocated as a catalyst for additional
investment by=20 universities and by the private sector. The
initiative is=20 directed by the National Coordination Office for
Computing,=20 Information, and Communications
(http://www.ccic.gov/),=20 and the principal agencies involved
have = been the=20 National Science Foundation (NSF), the
Defense Advanced=20 Research Projects Agency, the
Department of=20 Energy, the National Aeronautics and Space
Administration,=20 the National Institute of Standards and
Technology, the=20 National Institutes of Health, and the
National Library=20 of Medicine. Most of the project focus is
on = building a=20 new, very-high-speed network infrastructure to
support national=20 goals. The NGI initiative is a logical
outgrowth of the=20 existing High Performance Computing and
Communications=20 initiative, and all high-performance
mission-specific=20 networks now operated by Federal agencies
are = elements of=20 the NGI initiative, which builds upon such
networks as the=20 NSF=E2=80=99s very-high performance Backbone
Network Service=20 (vBNS).
Internet=20 2
The = Internet 2=20 (http://www.internet2.edu/) initiative is a
project of the=20 University Corporation for Advanced Internet
Development=20 (UCAID) (http://www.ucaid.edu/), a
collaborative=20 effort by more than 120 universities together
with = their=20 research partners in industry and government. The
goal = of the=20 project is to establish a distributed knowledge
system for=20 achieving ongoing innovations in research,
teaching, and=20 learning=E2=80=94to develop advanced Internet
technologies=20 and new applications that will support the
future research=20 and education needs of institutions of higher
education.=20
One = Internet 2=20 goal held in common with the NGI initiative
is = the effort=20 to connect universities, national laboratories,
and = private=20 partners with high-speed networks that are
1,000 = times=20 faster than today=E2=80=99s Internet, networks that can
handle at least=20 1G bps of connectivity for any user or
application=20 that needs it. One often-quoted aim is to be able
to = transmit and=20 receive the contents of the entire
Encyclopaedia=20 Britannica in under 1 second. Consortium
members have=20 established extremely high-capacity network
links = to a=20 separate national network with an infrastructure
parallel to the=20 commodity Internet, operated by many of the
same = common=20 carriers. Members connect to the high-speed
backbone=20 through special =E2=80=9Cpoint-of-presence=E2=80=9D service
providers that=20 are known as =E2=80=9Cgigapops=E2=80=9D because they can
deliver 1G bps=20 of data flow to each connection. Today it
costs = about=20 $500,000 in commitments and upgrades for a
university to=20 join UCAID and become connected to an
Internet 2=20 gigapop.
The = project=20 will demonstrate prototypes for a new
generation of=20 high-performance applications that support
scientific=20 research, national security, distance education,
environmental=20 monitoring, and health care, and that need
more = speed,=20 more bandwidth, more reliability, and more
security than=20 can be obtained using today=E2=80=99s Internet. It is
recognized that=20 most of today=E2=80=99s applications would perform
better if the=20 existing Internet worked smoothly and reliably
within its=20 current design parameters, but I2 is not an effort
to = improve the=20 speed of today=E2=80=99s Internet connections or to
resolve=20 shortcomings in the existing Internet. Instead, the
Internet 2=20 project aims to support the development of new
classes of=20 applications that by their very nature need
substantially=20 higher application-dedicated bandwidth,
bandwidth=20 reservation, improved security, guaranteed
quality of=20 service, and other advanced features of a new
generation=20 network.
Problems=20 With the Current Internet
Improving the=20 current Internet is not a goal of the I2 project,
but = in order to=20 support advanced =E2=80=9Cnext generation=E2=80=9D
applications,=20 Internet 2 will address many recognized
problems with=20 the current Internet, including problems with
network=20 management, end-to-end control, security,
reliability,=20 and the need for embedded functions to = support
11
distributed=20 systems. The lack of directory services and of
provisions for=20 inter-realm authentication will also be
addressed. A new=20 addressing scheme (=E2=80=9CIpv6=E2=80=9D) will be
implemented,=20 since the current Internet protocol (IP) address
space = is too=20 small to support the number of devices that now
need = to be=20 connected.
In = building the=20 NGI, the most important function that will be
added = to IP is a=20 set of =E2=80=9Cinternetworking=E2=80=9D protocols for
managing the=20 end-to-end connection. Central management
was=20 intentionally designed out of IP, because IP is meant to
be = deliberately=20 nondeterministic in the sense that each
information=20 packet traversing the network =E2=80=9Cfinds its own
way=E2=80=9D to its=20 destination based on the local conditions it
encounters along=20 the way. This makes the network
self-healing, as=20 individual packets can be independently
routed = around=20 broken connections or highly congested areas.
Unfortunately,=20 because there is no mechanism for sharing
information=20 about global conditions along an entire path,
each = packet can=20 be re-routed only after a problem is
encountered.=20 Internet 2 will improve the sharing of
knowledge about=20 distant conditions along a desired route
and = will permit=20 more sophisticated route management.
The = ability to=20 guarantee delivery is another important
function missing=20 from the current Internet, which uses a
delivery service=20 called =E2=80=9Cbest-effort=E2=80=9D delivery. = =E2=80=9CBest-effort=E2=80=9D
means = that all=20 packets are accepted from all senders, and are
sent = on their=20 way whenever possible. If there is too much
traffic at a=20 particular router, that router randomly discards
packets without=20 any concern for where they are coming
from = or where=20 they are going. Not only is there no guarantee
of = timely=20 delivery, there is no guarantee of delivery at all. If
an = expected=20 message or response fails to arrive, the sender
and = the receiver=20 will negotiate a resend of the incomplete
data, = in which=20 case even more packets have to be handled
by the = network.=20 This is satisfactory for the exchange of
e-mail = messages,=20 but not for a mission-critical system that
needs = guaranteed=20 delivery with a guaranteed maximum
latency.
Along = with=20 guaranteed delivery, Internet 2 aims to improve
the = end-to-end=20 quality of service of the network by
recognizing that=20 while dependability and reliability are
critical quality=20 parameters for all applications, different
applications=20 have different networking needs. A text file
transfer can=20 tolerate a 10-second delay in delivery, but a
command going to=20 a robotic control system in a nuclear
power = plant=20 cannot safely be delayed. Under Internet 2,
quality of=20 service (QOS) controls will permit applications to
reserve a=20 specific quality of service from the network and to
be = guaranteed=20 that the reserved latency and bandwidth will
be = available=20 when needed. Electronic commerce will be able
to = obtain secure=20 delivery guarantees. Real-time audio and
video = services=20 will be able to reserve bandwidth and to
obtain = guarantees of maximum packet loss rate and jitter.
Even = when an=20 application makes no quality-of-service
demands, new=20 =E2=80=9Ctype-of service=E2=80=9D (TOS) headers can help
routers to know=20 what type of material is included in each
packet. If a=20 Web page request and a fragment of a telephone
conversation=20 are competing for the same slot, the router can
make=20 intelligent decisions about prioritization. TOS can
actually reduce=20 congestion in absolute terms, because
preferential=20 delivery of an acknowledgment packet prevents
the = original=20 material from being resent by the sender.
One = implication=20 of the availability of quality-of-service
controls and=20 multicast capability expected in the next
generation=20 Internet is that Internet 2 will be far more
hospitable than=20 today=E2=80=99s Internet to connecting very large
numbers of=20 sensors. The capacity to make large amounts of
=E2=80=9Cpublic=E2=80=9D shared=20 sensor telemetry available to the Internet 2
community=20 represents an exciting opportunity to explore
new = classes of=20 applications. The ability of the network to
handle large=20 amounts of data that is irrelevant to most
people might be=20 seen as the very definition of today=E2=80=99s
Internet, but=20 in reality it is easy to swamp nearly any
real-life=20 network segment by attempting to pass real-time
telemetry data=20 for any complex event or series of simple
events. I2=20 promises to allow such data to exist on the
network without=20 interfering with other applications.
As an = example=20 of the potential for this kind of sensor-based
data, = consider=20 that today there are a few stores whose
inventory is=20 managed across the Internet, and there are a few
IP-enabled soft=20 drink vending machines whose contents can
be = queried by=20 their owners via the Internet. With this sort of
sensor data=20 widely available, it may soon be possible to
query = from any=20 public point of presence and learn in a few
seconds where=20 the nearest cola brand of your choice is
available=E2=80=94including=20 both stores and vending machines.
Another example=20 of IP-enabled equipment that is being
widely deployed=20 is vehicular traffic sensing units that report
local = traffic=20 conditions on a constant basis. When enough of
these = have been=20 deployed, it will be possible to receive a
constant stream=20 of sensor-provided data about traffic load at
most = busy=20 intersections across the country.
Prototype=20 Health Care Applications of the NGI
The = NLM has a=20 history of sponsoring health care activities
related to=20 high-performance computing and communi-
cations, and=20 now is funding the development of a variety of
health care=20 applications that will use and require the next
generation=20 Internet. Areas of particular interest to the NLM
include=20 advanced telemedicine, digital libraries, distance
learning, image=20 distribution, and other medical applications
that = require=20 the high-speed transfer of large amounts of data.
Other = funding=20 agencies are also underwriting the develop-
ment = of=20 advanced applications in health care that can serve
as=20 demonstration projects for the next generation Internet.
Medical=20 applications can qualify as NGI applications
without=20 involving large data sets or high-speed networking
if = they require=20 other features of the NGI, such as guaranteed
delivery,=20 guaranteed quality of service, or security of = private
12
medical=20 information that is to be transmitted over the public
network. There=20 are many examples of NGI and I2=E2=80=94
qualified=20 projects that have been funded in the past. Some
are = from fields=20 other than health care, but they all serve to
illustrate the=20 kinds of things that could or should be done in
the = design of=20 health care projects for NGI.
The = virtual=20 temporal bone (VTB) is a shared virtual reality
(VR) = simulation=20 based on an accurate 3-dimensional (3-D)
model = of the=20 temporal bone, designed as a teaching aid for
residents in=20 otolaryngology who are learning about the
anatomy of the=20 middle and inner ear. The VTB project was
developed by the=20 Virtual Reality in Medicine Laboratory,
the = Department=20 of Otolaryngology, and the Electronic
Visualization=20 Laboratory at the University of Illinois at
Chicago. This is=20 considered an Internet 2 project because the
quality of=20 service and end-to-end bandwidth made available
by = Internet 2=20 will be necessary to share this application
between=20 educational institutions. More information is
available at=20 http://www.sbhis.uic.edu/vrml/.
The=20 Collaborative Architectural Layout Via Immersive
Navigation=20 (CALVIN) is a testbed for applying VR in
architectural=20 design and collaborative visualization that uses
multiple=20 perspectives in an effort to take advantage of VR in
the = earlier,=20 more creative phases of the design process,
rather = than just=20 as a walkthrough of the final design.
CALVIN = is being=20 used to investigate issues in collaborative
VR, = including=20 network needs, avatar representations, audio
and = video=20 communication, and unusual user interfaces. This
is = considered an=20 Internet 2 project because for participants
to = collaborate=20 effectively, virtual environments require high
bandwidth and=20 low latency not possible on the current
Internet. CALVIN=20 is another project of the Electronic
Visualization=20 Laboratory at the University of Illinois at
Chicago. More=20 information is available at
http://www.evl.uic.edi/spiff/calvin.
Some = of the most=20 important projects in health care are those
requiring shared=20 access to extremely large data sets that
cannot = be=20 accessed in a reasonable amount of time without
the=20 high-bandwidth connections promised by the NGI. The
most = well-known=20 of the large data sets in health care is the
NLM=E2=80=99s =E2=80=9CVisible=20 Human=E2=80=9D data set, a complete, anatomically
detailed, 3-D=20 digital representation of the male and female
human = body.=20 Transverse computed tomograms, magnetic
resonance=20 images, and cryosection images of representative
male = and female=20 cadavers were acquired at 1 millimeter and
0.33 = millimeter=20 intervals to make this data set. The female
data = set is=20 approximately 40 gigabytes in size, and would
take = 97 days to=20 download through a standard modem in
1998. = More than=20 50 projects and products have been based
on the = visible=20 human data set, including the 3-D Virtual
Colonoscopy=20 project, demonstrating static images and fly-
through=20 animations from the Departments of Radiology and
Computer Science=20 at SUNY Stony Brook, the Interactive
Knee = Program=20 from the University of Pennsylvania Medical
Center, and=20 =E2=80=9CCathSim,=E2=80=9D an intravenous training system = that
includes a=20 tactile feedback device.
Another example=20 of large shared data sets is the Human
Genome Project=20 of the National Center for Biotechnology
Information=20 (NCBI). These are the most extensive databases
on = DNA and=20 protein sequence data in existence, and they
are = very=20 heavily used by researchers around the world. The
NCBI = Web site=20 attracts more than 2 million hits a day.
While = the basic=20 search engines serving these data do not
require NGI=20 attributes, such attributes may be needed by
other = applications using large portions of the data set, or by
those = using it=20 for visualization or rendering.
The = =E2=80=9CTerabyte=20 Challenge=E2=80=9D is a distributed testbed for
research into=20 the management and mining of truly massive
data = sets. Data=20 mining is the automatic discovery of
patterns,=20 associations, changes, and anomalies in large data
sets. = Developing scaleable data mining algorithms, software
tools, and=20 applications is a fundamental scientific challenge
with = important=20 implications for health care as well as for
many = other=20 scientific and business disciplines. The terabyte
challenge=20 project facilitates prototyping of new ideas and
software tools=20 and allows them to be tested against large
data = sets, and=20 against distributed data sets. This is
considered an=20 Internet 2 project because the mining of
geographically=20 distributed massive data sets cannot be done
without=20 high-performance links. The terabyte challenge is a
collaborative=20 project of the University of Illinois at Chicago,
the = University=20 of Pennsylvania, and the University of
Maryland. More=20 information is available at
http://www.nscp.uic.edu.
The = term=20 =E2=80=9Ccollaboratory=E2=80=9D is a neologism describing a
distributed=20 collaborative laboratory. Collaboratories often
support=20 teleconferencing as a basic collaboration tool, but
more=20 importantly they also permit remote access and
collaborative=20 (simultaneous) access to physical resources at
a = distant site.=20 Collaboratories need not deal with scientific
data: = an=20 existing music collaboratory (=E2=80=9Cjam session = site=E2=80=9D)
actually allows=20 two or more musicians to practice together
over = the=20 Internet. At least one CD has been released
featuring music=20 that was created and recorded through
collaborative=20 telepresence. Another existing collaboratory is
a=20 meteorological site that allows multiple users to
communicate in=20 real-time while simultaneously displaying
weather maps=20 and other streams of meteorological data that
are = germane to=20 a discussion.
Other = interesting wide-area projects provide access to and
control of=20 special imaging techniques. Examples have
ranged from=20 single-photon imaging in cultured cerebellar
granule cells=20 to radio astronomy synthesis imaging, a
process in=20 which 27 radio astronomy antennas track the
galactic center=20 while the data are recorded, merged, and
processed in=20 real-time to show a visible simulation of the
Fourier=20 transforms. CMDA, the =E2=80=9CCollaboratory = for
13
Microscopic=20 Digital Anatomy,=E2=80=9D makes a unique research
tool = available=20 over Internet 2 connections: An intermediate-
high-voltage=20 transmission electron microscope at the
National Center=20 for Microscopy and Imaging Research at
the = University=20 of California, San Diego, has been modified
to be = like a=20 computed tomography scanner. The specimen is
tilted = to a=20 series of different angles for a series of scans, and
the = resulting=20 data are tomographically reconstructed to give
3-D = information=20 about its internal structure. Researchers can
control the=20 electron microscope lenses and the 4-axis
specimen stage=20 by remote control. The tomographic recon-
struction is=20 very computationally intensive, so as the scans
are = performed=20 the data are transparently sent out (across
Internet 2) to=20 high-performance supercomputers for
calculation. The=20 resulting 3-D images are returned to the
remote = researchers, who possess neither microscope nor
supercomputer,=20 but who have enjoyed the use of both to
capture 3-D=20 structural images that would otherwise have
been = impossible=20 to obtain.
One of = the most=20 exciting tools for investigating the
possibilities of=20 NGI is a visualization tool called the CAVE.
Created by=20 scientists at the University of Illinois=E2=80=99 = Electronic
Visualization=20 Laboratory in 1992, the CAVE is a multi-
person, high=20 resolution, 3-D graphics video and audio
environment. It=20 is a 10x10x10-foot structure that sits in a
35x25x13-foot=20 darkened room and has rear-projected screen
walls = and a=20 front-projected floor. Inside the CAVE, a user
wearing 3-D=20 shutter glasses is fully immersed in a virtual
environment.=20 Objects appear to float in space, and the
computer tracks=20 a user=E2=80=99s position to keep the perspectives
correct while=20 the user =E2=80=9Cwalks=E2=80=9D around objects. A = =E2=80=9Cdesktop=E2=80=9D
version of such=20 a display also exists, called an
=E2=80=9CImmersadesk.=E2=80=9D=20 These visualization capabilities are already
available to=20 those who can afford the supercomputing power
needed = for=20 real-time rendering=E2=80=94in fact, General Motors
designed its=20 1997 and 1998 automobiles using an
automobile=20 simulator that uses the CAVE principles. NGI
and I2 = offer the=20 promise that CAVE environments could
offload their=20 real-time rendering to supercomputing centers
connected via=20 the next generation of network connections.
Summary
From = the=20 earliest days of small telephone exchanges until
the = present=20 time, connectivity has been a driving force for
change. The=20 Internet and the World Wide Web arose out of
the = existing=20 base of telephone connections and computer
networks and=20 inherited their bandwidth restrictions, but
nonetheless have=20 brought a previously unimagined richness
to = connectivity,=20 changing the world dramatically as a result.
The = NGI is being=20 specifically designed to provide high-
bandwidth=20 connectivity on a scale so vast that is difficult to
comprehend. It=20 is our task to imagine into existence the
richness and=20 complexity of the applications that will exploit
this = new=20 connectivity.
C ONFIDENTIALITY=20 AND M EDICAL I NFORMATION
Paul = Clayton,=20 Ph.D.
Professor and=20 Chair
Department of=20 Medical Informatics
Columbia=20 University
Director of=20 Clinical Information Services
Columbia=20 Presbyterian Medical Center
New = York,=20 NY
This=20 tutorial presentation discusses the thorny issues of
privacy,=20 confidentiality, and security in the digital age.
Social problems=20 related to medical informatics potentially
could = be=20 greater than technological ones. Much of this talk
draws = from a=20 report on protecting electronic health
information=20 that was funded by the National Library of
Medicine and=20 Massachusetts Healthdata Consortium in 1995
and = conducted=20 by the Computer Science and
Telecommunications=20 Board of the National Research Council. 1
An = eclectic=20 committee=E2=80=94including sociologists,
computer=20 scientists, a dean of nursing, an attorney, a
pharmacist, and=20 a clinical researcher=E2=80=94examined practical
aspects of=20 technology such as the status of current
technology,=20 related problems, and strategies to address these
problems.=20
Several key=20 terms relate to confidentiality issues: (1) privacy
is = the desire=20 of individuals to limit the disclosure of personal
information;=20 (2) confidentiality is a condition in which
information is=20 shared or released in a controlled manner
(e.g., the=20 release of information to a physician to benefit the
patient); and=20 (3) security involves measures to protect the
integrity,=20 confidentiality, and availability of information
(e.g., if=20 security is breached and an information system is
disrupted or=20 data are changed, the data are unavailable
to=20 users).
Advantages=20 of Electronic Records
There = are=20 numerous advantages to using electronic records.
First, with=20 worldwide connectivity, electronic records are
readily=20 available from virtually any location. Electronic
records also=20 facilitate organizing the information in medical
records. Based=20 on the data that are available and organized
in = the records,=20 reminders, suggestions, alerts, and critical
pathways can be=20 generated electronically. Electronic records
also = make it=20 possible to link to appropriate medical literature
and = to analyze=20 the health care delivery process, a task that
would = have been=20 too cumbersome with paper records.
Threats to=20 Confidentiality
For = these=20 reasons, paper records are being replaced with
electronic=20 ones. However, serious issues of privacy and
confidentiality=20 are associated with the improved access
afforded by=20 electronic records. One of the most difficult
challenges is=20 how to balance accessibility and
confidentiality. One=20 of the main reasons to use electronic
records is to=20 improve access to data; unlike paper records
14
that = can be=20 viewed in only one location; electronic records
provide access=20 to multiple users in multiple locations.
However,=20 increased access also means increased
opportunities=20 for breaches of confidentiality. To determine
appropriate=20 protection measures it first is necessary to
identify the=20 threats to confidentiality. The study group felt
that = insider=20 abuse accounts for 80 percent of the
confidentiality=20 infractions in the United States. Often these
abuses = are=20 perpetrated by individuals with access to the
information=20 system and are motivated by curiosity. Other
reasons include=20 vindictive retaliation, humiliation, and
blackmail.=20 Malevolent hackers, on the other hand, generally
are = not=20 motivated by personal reasons; their purposes more
likely = involve=20 the denial of service, blackmail, or the
destruction of=20 data integrity. Investigators also pose a threat;
for a = modest=20 fee, a determined investigator usually can find
anybody=E2=80=99s=20 electronic or paper medical history. No absolute
security exists=20 with either paper or electronic records. The
committee=20 strongly felt that the biggest threat to ordinary
individuals=20 concerned the way in which our health records
could = be used to=20 cause economic harm.
Increased access=20 is a factor as data are shared among various
users = and=20 organizations. Because data no longer are
controlled by a=20 single organization, even if some organiza-
tions = maintain=20 security, information loses confidentiality
once = it leaves=20 the originating organization=E2=80=99s control. Many of
the = systemic=20 threats are related to economic gain. If
companies have=20 access to potential employees=E2=80=99 medical
records, they=20 can greatly reduce their health care costs and
improve their=20 profitability by not hiring individuals with
certain medical=20 conditions. In addition, under the
Employment,=20 Retirement, and Income Security (ERISA)
legislation,=20 self-paying insurers (health insurance companies
merely = manage=20 the claims) have a right to learn what drugs
their = employees=20 are taking because they are paying for those
prescriptions.=20 This information could affect personnel
decisions.=20 Unprotected information on genetic susceptibility
to = disease also=20 could have broad economic implications. In
order = to prevent=20 fraud, the Medical Information Bureau
tracks = and=20 shares information among life insurance
companies on=20 everyone who applies for life insurance. In
addition,=20 records of test and control patients from a major
medical center=20 were subpoenaed in recent litigation related
to = breast=20 implants.
The = lack of=20 consistent policy on confidentiality issues
confounds the=20 problem. Currently, no Federal policies exist to
govern = the flow=20 of medical information, and there is no
widely = accepted=20 definition of what constitutes fair information
practice.=20 Although some Federal and State legislation exists, it
is not = sufficient to ensure confidentiality of medical records.
The=20 Kennedy-Kassebaum bill mandated that Congress would
enact=20 legislation on the privacy of medical information by
August = 1999.=20 Otherwise, by February 2000, the U.S.
Secretary of=20 Health and Human Services is obligated by law
to = institute=20 regulations on privacy and confidentiality.
Technical=20 Remedies
The = following=20 are practical, cost-effective technical
remedies to=20 confidentiality problems:
=E2=80=A2=20 Authentication: It is important to accurately = identify
system users.=20 Strong authentication involves techniques
such = as iris=20 recognition, retinal scans, and thumbprint
recognition;=20 weak authentication includes using passwords
and = individual=20 logon identification.
=E2=80=A2=20 Authorization: Within an institution and across the
system, someone=20 must establish who has access to what
data. = This is=20 probably the most difficult task and has little
to do = with=20 technology.
=E2=80=A2 Audit=20 trails: Audit trails track users and the data they have
accessed.=20 Although unauthorized users can be detected
through audit=20 trails, legislation is needed to ensure that
sanctions are=20 imposed when security violators are
identified.
=E2=80=A2 Physical=20 security: Do not allow unauthorized users access
to = the system=20 or the data centers.
=E2=80=A2 Control of=20 external communication links: In addition to
building=20 firewalls, control the circumstances under which
modems and PC=20 emulators may be used.
=E2=80=A2 Software=20 discipline: If you have a library of structured
query = language=20 (SQL) calls that allow anyone to query a
database, those=20 routines must conform to a security
architecture.
=E2=80=A2 System=20 backup/disaster recovery: Routinely back up data
on=20 tapes.
=E2=80=A2=20 Self-assessment: Hire experts to determine if the = system
security can be=20 violated.
Disincentives and=20 Challenges
Encryption=20 security software for transmitting data over the
Internet is=20 available currently but is used rarely. The Data
encryption=20 standard uses a single key, which is shared
between the=20 sender and the recipient. However, this key
tends = to get=20 widely distributed. With an asymmetric key
method, a=20 public key is used to encrypt the message, but a
private key is=20 used to decrypt the message. This method,
however, is=20 considerably slower and requires a certificate
authority to=20 ensure the identity of the user. The secure
socket layer=20 protocol allows a practical approach to
encryption and=20 provides security on a nonsecure network.
This = process=20 entails the use of a one-time session key and a
public key.=20 Newer applications can use this technology
seamlessly and=20 painlessly. However, in an information
system with=20 many applications, it is likely that only pieces
of a = security=20 solution are implemented rather than an entire
system.=20
Development of=20 policies can pose an even greater challenge
than = the=20 technical measures, because people often do = not
15
think = about=20 policy until there is a crisis. However, it is
important for an=20 organization to establish a policymaking
body, = to=20 designate an information security officer, and to
establish=20 sanctions. Because there often are extenuating
circumstances=20 related to security violations, specifying
sanctions after=20 the fact may be difficult. It also may be
difficult for=20 Congress to pass legislation that satisfies both
privacy=20 advocates and organizations that=E2=80=94for increased
profit = and/or=20 improved quality of care=E2=80=94are funding systems
which = increase=20 access to medical information.
There = are=20 several disincentives for organizational solutions,
including cost,=20 a lack of external incentives, and a lack of
focus = on=20 information technology. If cumbersome and
lengthy sign-on=20 procedures discourage users from logging
off = when=20 finished, then users will circumvent the intended
process.=20 Therefore, there is substantial difficulty in
balancing the=20 need for access with the need to limit
legitimate=20 access, which again is related to policy issues.
Conclusion
Overall,=20 widespread flow of information poses the greatest
threat = to=20 confidentiality of medical information. This flow
may = cause=20 economic harm to an individual. The develop-
ment = of medical=20 informatics to support the National Heart
Attack = Alert=20 Program could involve policy issues that are
more = daunting=20 than the technological issues. Until social
policy = on=20 confidentiality is established and confidentiality
issues = are=20 resolved, the advancement of technology will be
difficult.=20
Reference
1. = National=20 Research Council, Computer Science and
Telecommunications=20 Board. For the Record: Protecting
Electronic=20 Health Information. Washington, DC: National
Academy Press,=20 1997.
D
IAGNOSIS OF=20
P
ATIENTS WITH=20
A CUTE M YOCARDIAL I
NFARCTION=20 AND
O
THER=20
A CUTE C ORONARY S YNDROMES
Robert = J.=20 Zalenski, M.D., M.A.
Associate=20 Professor
Emergency=20 Medicine and Internal Medicine
Wayne = State=20 University
Detroit,=20 MI
This=20 presentation describes the pathophysiology of acute
myocardial=20 infarction (AMI) and other acute coronary
syndromes and=20 examines diagnostic challenges for the
emergency=20 physician, including options for diagnostic
testing and=20 techniques used in making appropriate treatment
decisions.
Acute=20 Myocardial Infarction
As = the body=20 ages, plaques develop in the coronary arteries
with = a=20 lipid-rich core covered by a fibrous cap. When the
plaque=20 ruptures, coronary thrombus, or clot, forms. People
often = are not=20 aware of this process in their bodies. However,
when = the=20 process reaches the dynamic phase, the artery is
blocked by=20 coronary thrombus and no blood passes. If this
condition=20 persists, the patient will experience pain,
electrocardiogram=20 (ECG) changes, heart attack, and, finally, death. 1
In = the=20 asymptomatic phase, the plaque ruptures, thrombus
forms, and=20 blood flow to the heart is reduced (myocardial
ischemia). When=20 myocardial ischemia occurs, the phase of
the = syndrome is=20 called unstable angina, characterized by
attacks of pain=20 that are unpredictable or suddenly increase in
severity or=20 frequency. If the thrombus remains and this state
persists, it=20 leads to myocardial necrosis.
Importance=20 of Reducing Cardiovascular Disease
AMI = is an=20 extremely prevalent phenomenon; it is expensive,
and = it is=20 lethal. There are nearly 6 million hospitalizations
per = year for=20 cardiovascular disease and over 1 million first
or = second=20 discharge diagnoses of AMI per year in the
United States.=20 Each year there are 500,000 mortalities due to
coronary artery=20 disease including 300,000 cardiac arrests.
Out = of 100=20 patients with AMI, there are 25 deaths per
month. 2
In = addition to=20 being associated with $33 billion per
year = in medical=20 costs, coronary artery disease is a leading
cause = of=20 malpractice payouts.
Strategies=20
Early = treatment=20 is essential to reducing mortality from AMI.
If = patients=20 with elevated ST-segment myocardial infarction
(MI) = (the=20 actual stage of heart attack) are identified and
treated early,=20 mortality can be greatly reduced. However, if
this = condition=20 is not identified, it can be lethal. The GISSI
study = found=20 that when heart attack patients were treated
within the=20 first hour of the occurrence of symptoms, the
21-day=20 mortality rate was reduced by nearly 50 = percent.
3
In
the = MITI=20 project, patients who received early treatment with
thrombolytic=20 therapy in the ambulance or hospital had a
1 = percent=20 mortality rate, compared to 8 percent for those who
received later=20 treatment.
4
McCarthy found=20 that 10 percent of
patients=20 discharged with unrecognized AMI died and that the
mortality rate=20 was 25 percent.
5
In = 1996, there=20 were approximately 10 million ambulatory
visits to the=20 hospital for chest pain,
6
and = the number=20 most
likely has=20 increased. Most of these patients need to be
considered for=20 diagnoses of AMI or ischemia, as well as other
diagnoses, but=20 unfortunately the diagnostic tools currently
available are=20 limited. An ideal diagnostic tool would be
applied easily=20 in the emergency department (ED), and it
would = have high=20 sensitivity and specificity, enabling accurate
diagnosis. In=20 addition, it would have low interpretation error,
carry = minimal=20 risk, and be cost- effective.
16
Existing=20 Diagnostic Tools
Existing=20 diagnostic tools include the physical examination,
the = ECG,=20 biochemical markers, and imaging tests. Because
most = patients do=20 not present with obvious symptoms, the
physical=20 examination is useful for diagnosing AMI and other
acute = coronary=20 syndromes only in certain patients.
Therefore,=20 diagnostic technologies are essential.
The=20 electrocardiogram
The = ECG is the=20 most important diagnostic tool. If patients
have = chest pain=20 and shortness of breath, an ECG is
indicated. A=20 finding of ST-segment elevation on the ECG is
95 = percent=20 specific for an AMI with a thrombus in a large
artery, a=20 condition that should be treated quickly. A finding
of ST=20 depression, indicating ischemia, has moderate
sensitivity and=20 specificity; ST depression is not as accurate a
test. = Other ECG=20 findings are of lower sensitivity and
specificity.=20 Therefore, a patient with a normal ECG or one
with = flattened T=20 waves still might be experiencing an acute
coronary=20 syndrome and not be identified through the ECG.
The = American=20 College of Cardiology/American Heart
Association=20 guidelines state that ST-segment elevation is
strong = evidence=20 for thrombotic occlusion and makes the
patient a=20 candidate for immediate reperfusion therapy.
According to the=20 National Registry of Myocardial
Infarction,=20 about 45 percent of patients have this finding,
and a = variety of=20 studies have found that approximately
50 = percent of=20 patients with AMI have ST-segment elevation.
With a = finding=20 of ST depression, indicating a clot in the
coronary artery,=20 it is not clear whether the clot will resolve
(with = no heart=20 damage) or will persist and result in a
non-Q-wave MI.=20 With a non-Q-wave MI, damage occurs to
only = part of the=20 wall of the myocardium, as opposed to a
Q-wave = MI that=20 suggests myocardial necrosis transmurally
(i.e., = through=20 the wall).
Cardiac=20 markers
If = sufficient=20 information is not obtained from the ECG, other
diagnostic=20 modalities are used. Cardiac markers detect
myocardial=20 necrosis in the blood and can indicate whether
the = patient has=20 experienced a heart attack; if any part of the
heart = is dying,=20 the patient needs additional treatment and
study. = One of=20 the markers, myoglobin, is found in both
skeletal and=20 cardiac muscle, which makes it difficult to
identify the=20 source without additional indicators. Creatine
kinase = (CK) and=20 CK-MB, which is relatively specific for
heart = muscle, is=20 another cardiac enzyme. However, because
CK-MB = is=20 primarily present in the heart rather than skeletal
muscle, it=20 requires analysis to determine if the heart is the
most = likely=20 source. The third cardiac marker, troponin, is
nearly = perfectly=20 specific for cardiac muscle. The major
advantage of=20 troponin is that it can indicate small quantities
of = heart damage,=20 so that small infarcts=E2=80=94not just large
ones=E2=80=94can be=20 detected.
A = study by=20 DeWinter of 307 ED patients who complained of
chest = pain=20 examined the cumulative proportion of patients
who = rose above=20 the positive cutoff zone for 4 cardiac
markers:=20 myoglobin, CK mass, CK activity, and = troponin.
7
Myoglobin rose=20 faster and reached 100 percent sensitivity
(at 6 = hours)=20 more quickly than did the other three markers.
The = study=20 indicated that all the markers could be effective.
However,=20 troponin may prove to be particularly useful
because of its=20 ability to detect small amounts of heart
damage.
Imaging tests=20
Some = patients=20 may have a complete occlusion of the artery
that = is not=20 indicated by ECGs or biochemical markers.
Imaging tests=20 examine the state of perfusion in the
myocardium and=20 help identify areas that are not perfused,
even = though the=20 nonperfusion may not last long enough to
kill = the heart=20 muscle.
Technetium (Tc)=20 99m sestamibi is a radiopharmaceutical
ideally suited=20 for the ED. When injected with this agent,
the = ischemic=20 area appears as a =E2=80=9Ccold spot=E2=80=9D within about
15 = minutes, and=20 imaging can take place up to 4 hours later.
Even = after the=20 patient=E2=80=99s pain has subsided, a positive scan
may = be=20 obtained. Imaging tests add the additional dimension
of = capturing=20 patients who have unstable angina or acute
cardiac=20 ischemia without infarction. In addition, patients
who = need=20 further treatment and study, even those with AMI,
can = be=20 identified earlier than with markers, which take
between 6 and=20 12 hours to reach 100-percent sensitivity.
An = ECG stress=20 test is an additional option for diagnosis of
cardiovascular=20 disease. Rather than waiting for a
reoccurrence of=20 chest pain, the physician can perform ECG
stress testing=20 after an initial period of survey and
observation=20 once acute symptoms have ceased. Ischemia
may = be detected=20 through the patient=E2=80=99s inability to meet the
additional=20 oxygen demands of a stress test.
Selecting=20 Appropriate Treatment
The = number one=20 priority in the ED is to identify reperfusion
candidates. All=20 patients with chest discomfort receive an
ECG, = and this=20 screening applied widely and early (including
in = the field)=20 translates into earlier treatment. If ST-segment
elevation or=20 depression is not present, the next step is to
exclude MI,=20 which can be accomplished with the
biochemical=20 markers or with Tc 99m sestamibi. The latter
also = is useful=20 for detecting acute ischemia without
infarction.=20
Risk=20 stratification
Once = patients=20 with coronary artery disease have been
identified,=20 they undergo risk stratification according to their
place = on the=20 spectrum of acute coronary syndromes. A
patient with=20 chest pain or dyspnea with ST-segment
elevation,=20 indicating coronary thrombosis, would undergo
immediate=20 reperfusion. If a patient=E2=80=99s ECG does not indicate
ST-segment=20 elevation, only a positive biochemical marker
17
would = confirm a=20 non-Q wave MI, in which case
thrombolytic=20 therapy would not be indicated and medical
therapy would be=20 indicated. A patient with positive rest
imaging, which=20 would indicate ischemia without necrosis,
also = would=20 receive medical therapy.
Decision=20 aids
A = variety of=20 decision aids and algorithms are available to
organize=20 diagnostic information and facilitate treatment
decisions. The=20 Acute Cardiac Ischemia Time-Insensitive
Predictive=20 Instrument may be particularly useful with
patients whose=20 diagnoses contain areas of uncertainty. This
instrument=20 collects some simple demographic data, such as
age = and gender;=20 determines whether the patient has chest
pain = and if so,=20 whether that is the chief complaint; and
collects=20 information on five ECG variables. It produces a
valid = and=20 reliable score of the risk of ischemia with a range
between 0 and=20 100. It initially was tested and validated in
5,800 = patients=20 and subsequently has been in trials with over
10,000 = patients.
8
In = summary,=20 various diagnostic tools and approaches are
used = to identify=20 patients with AMI and other acute coronary
syndromes. Their=20 usefulness will be substantially augmented
by = additional=20 biochemical markers that are being developed
for = detecting=20 cardiovascular disease in even earlier stages of
symptomatology.=20 With current and emerging diagnostic
tools, = ED=20 physicians are fortunate to have an opportunity to
identify=20 coronary artery disease before it leads to myocardial
damage,=20 arrythmias, or death.
References
1. = Theoroux P,=20 Fuster V. Acute coronary syndromes: unstable
angina and=20 non-Q-wave myocardial infarction. Circulation
1998;97(12):1195=E2=80=93206.
2. = Braunwald E,=20 Mark DB, Jones RH, Cheitlin MD, Fuster V,
McCauley KM,=20 Edwards C, Green LA, Mushlin AI, Swain JA,
Smith = EE, Cowan=20 M, Rose GC, Concannon CA, Grines CL,
Brown, L, Lytle=20 BW, Goldman L, Topol EJ, Willerson JT,
Brown = J,=20 Archibald N. Unstable Angina: Diagnosis and
Management.=20 Clinical Practice Guideline Number 10. AHCPR
Publication No.=20 94-0602. Rockville, MD: Agency for Health
Care = Policy and=20 Research and the National Heart, Lung, and
Blood = Institute, Public Health Service, U.S. Department of
Health and=20 Human Services. March 1994.
3. = Gruppo=20 Italiano per lo Studio della Streptochinasi = nell=E2=80=99Infarto
Miocardico=20 (GISSI). Long-term effects of intravenous
thrombolysis in=20 acute myocardial infarction: final report of the
GISSI = study.=20 Lancet 1987;2(8564):871=E2=80=934.
4. = Brouwer MA;=20 Martin JS; Maynard C; Wirkus M; Litwin PE;
Verheugt FW;=20 Weaver WD. Influence of early prehospital
thrombolysis on=20 mortality and event-free survival (the
Myocardial=20 Infarction Triage and Intervention [MITI]
Randomized=20 Trial). MITI Project Investigators. Am J Cardiol
1996;78(5):497=E2=80=93502.
5. = McCarthy BD;=20 Beshansky JR; D=E2=80=99Agostino RB; Selker HP.
Missed=20 diagnoses of acute myocardial infarction in the
emergency=20 department: results from a multicenter study [see
comments].=20 Ann Emerg Med 1993;22(3):579=E2=80=9382. Note:
Comment in:=20 Ann Emerg Med 1994;23(1):141=E2=80=932.
6. = Schappert=20 SM. Ambulatory Care Visits to Physician Offices,
Hospital=20 Outpatient Departments, and Emergency Departments:
United States,=20 1996. Vital Health Stat=E2=80=94Series 13: Data From
the = National=20 Health Survey. 1998; (134):1=E2=80=9337.
7. de = Winter=20 RJ, Koster RW, Sturk A, Sanders GT. Value of
myoglobin,=20 troponin T, and CK-MB
mass
in = ruling out=20 an acute
myocardial=20 infarction in the emergency room. Circulation
1995;92:3401=E2=80=937.
8. = Selker HP,=20 Griffith JL, D=E2=80=99Agostino RB. A time-insensitive
predictive=20 instrument for acute myocardial infarction mortality:
a = multicenter=20 study. Med Care = 1991;29(12):1196=E2=80=93211.
D
ATABASES AND=20
D ATA S TANDARDIZATION
Clement J.=20 McDonald, M.D.
Distinguished=20 Professor of Medicine
Indiana=20 University School of Medicine
Director
Regenstrief=20 Institute for Health Care
Indianapolis,=20 IN
This=20 tutorial presentation discusses access to databases and
data=20 standardization in general. It serves as background for
subsequent=20 symposium presentations on databases relevant
to = clinical=20 care and research of heart attack patients.
A = rich lode of=20 clinical data exists in laboratories,
pharmacies,=20 hospitals, and other clinical care settings.
Health=20 insurance companies, researchers, pharmaceutical
manufacturers,=20 and physicians require data in a variety of
forms, leading=20 to increased demand for data automation. In
addition,=20 cardiology procedure reporting systems, cardiology
outcome data,=20 and medical records are all becoming
available in=20 electronic format.
However, there=20 are several barriers to the pooling of these
data. = The data=20 are stored in many isolated locations; each
system has a=20 different internal system, conceptualizations,
and = internal=20 codes; and there is a radical mismatch in the
conceptualization of=20 the systems that generate data and
those = that use=20 them. Analysts tend to conceptualize data as a
flat = file,=20 where the case is one record, the fields contain the
individual=20 values of interest, and these values are often
derived (or=20 rolled up)=E2=80=94they are often conclusions rather
than = the raw=20 information available from a production
system. In the=20 flat file, the definitions of the questions are
buried in the=20 definitions of the fields. So what one sees in
the = data does=20 not really fully define the questions, making it
difficult to=20 merge these different kinds of data from many
different=20 sources.
18
In the = application conceptualization, each question has its
own = record, and=20 this record contains many attributes, e.g.,
the = name of the=20 variable (question), its value (answer), its
units, = and=20 normal range when applicable. Indeed, the
answers are=20 often in raw data form, and the questions are
fully = defined in=20 a separate master file. In comparison with
the = analyst=20 model, the application model has more records
and = greater=20 detail. Individuals who want to extract data from
systems for=20 administrative and research purposes should use
the = application=20 conceptualization standard and then convert
these = data as=20 needed into the analytic form in their own
systems. 1
Entropy leads to=20 different data structures and different
idiosyncratic=20 codes within different computer systems. We
need = to apply=20 standards
2
to = the outputs=20 of these systems in
order = to reduce=20 this entropy.
Several=20 different types of data standards are needed: security
standards,=20 message standards, code standards, and person
and = place=20 identifier standards. The technology exists to
address privacy=20 issues, but the challenges lie in developing
procedures for=20 implementing them. Likewise, the Internet
Engineering Task=20 Force and Health Level Seven (HL7=E2=80=99s)
security,=20 special interest group provide standard tools for
implementing=20 security measures.
Message=20 Standards
The = principal=20 clinical message standard is HL7
3
,=20 although
ASC=20 X12
4
provides claims=20 transmission and other
administrative=20 information, and the National Council for
Prescription=20 Drug Program
5
provides=20 message standards for
community=20 pharmacies. HL7 is the American National
Standards=20 Institute-approved clinical message standard and is
widely = used in=20 the United States and in a number of
countries=20 throughout the world. The Centers for Disease
Control and=20 Prevention (CDC) has implemented the Data
Elements for=20 Emergency Department Systems (DEEDS)
6
database using=20 HL7.
HL7 = was=20 established to develop standards for the electronic
interchange of=20 clinical, associated financial, and
administrative=20 information both within and among care
institutions. It=20 defines data types and structures for the
records within=20 the messages. Sixty-five percent of those
surveyed at a=20 Health Information Management Systems
Society=20 conference said that HL7 was the most important
standard in=20 their environment.
7
In = the HL7=20 standard, field
types = are=20 standardized; for example, dates are represented as
CCYYMMDD (e.g.,=20 Christmas of 1998 is 19981225). This
makes = HL7=20 automatically year-2000 compliant, and allows
birth = dates for=20 patients over 100 years of age.
Code=20 Standards
HL7 = has a coded=20 data type. It requires that users always send
a code = source=20 with the code so that there is never confusion
about = the source=20 book for the code. This will be very helpful
during = the=20 overlap period from updating the International
Classification=20 of Diseases, Ninth Revision, Clinical
Modification=20 (ICD-9-CM) to ICD-10-CM.
An=20 electrocardiograph is another source of patient
information.=20 Comit=C3=A9 Europ=C3=A9en de Normalisation, the
European=20 standardization organization, has developed a
standard for=20 communication protocol for computerized
electrocardiography. 8
It = provides=20 standards for representing
the = ECG tracing=20 itself, diagnostic codes, and the measure-
ments = extracted=20 from the ECG.
Many = domains of=20 code standards exist and not everything is
covered under=20 one umbrella code system. The Logical
Observation=20 Identifiers, Names and Codes (LOINC) system
provides=20 standard names and codes for identifying clinical
questions,=20 variables, and report headers.
9
It = provides=20 codes
for=20 representing the clinical questions, not the answers.
LOINC = names=20 consist of up to six parts: component,
property,=20 timing, system, scale, and method. LOINC has
been = widely=20 adopted in part due to its availability on the
Internet=20 (http://www.mcis.duke.edu/standards/termcode/
loinc.htm). It=20 is endorsed by the American Clinical
Laboratory=20 Association and used by the larger reference
laboratories.
Systematized=20 Nomenclature of Human and Veterinary
Medicine=20 (SNOMED)
10
is = the code=20 source for answers and
provides over=20 250,000 categories of answers. SNOMED is a
comprehensive,=20 multiaxial nomenclature classification
system created=20 for indexing the entire medical record,
including signs=20 and symptoms, diagnoses, and procedures.
SNOMED=20 International is rapidly being accepted worldwide
as = the standard=20 for indexing medical record information.
Person and=20 Place Code Standards
Though many=20 countries have universal patient identifiers, the
United States=20 does not. Consequently, each institution creates
its = own=20 registry and patient chart number. The Health Care
Insurance=20 Portability and Accountability Act requires that
provider and=20 patient identifiers be put in place, but there may
be a = reluctance=20 to develop patient identifier codes because of
the = privacy=20 implications. Provider identifiers also are
important to=20 outcomes analysis and health services research.
The = Health Care=20 Financing Administration is developing a
national=20 identifier system for health care providers. This
means = that we=20 will soon have a national provider code.
Public health=20 organizations have taken the lead in developing
standards for=20 patient data, especially the CDC. They have
developed a=20 series of standards-based projects, including
communicable=20 disease reporting, in which they use HL7 for
the = message,=20 LOINC for identifying the tests, and SNOMED
for = identifying=20 the answers. The CDC, in cooperation with
the = Council of=20 State and Territorial Epidemiologists (CSTE),
has = developed a=20 series of standards-based projects, including
communicable=20 disease reporting, in which they use HL7 for
the = message,=20 LOINC for identifying the tests, and SNOMED
for = identifying=20 the answers. The CSTE developed tables for
19
each = of the 50=20 States, thereby defining all the disease
requirements,=20 what lab tests should be performed, and the
threshold for=20 reporting a test result as positive. In addition,
the = CDC has the=20 emergency medical care DEEDS database,
which = is very=20 relevant to the work of the National Heart
Attack = Alert=20 Program.
The = major goal=20 of medical informatics should be to reduce
the = entropy of=20 the many sources of health information by
developing=20 standards for messages, codes, and for user and
component=20 interfaces.
References
1. = McDonald CJ,=20 Overhage JM, Dexter P, Takesue BY, Dwyer
DM. A = framework=20 for capturing clinical data sets from
computerized=20 sources. Ann Int Med 1997;127:675=E2=80=9382.
2. = McDonald CJ,=20 Dexter PR, Overhage JM, Takesue B. Health
Informatics=20 Standards: A View From Mid-America. IMIA
Yearbook of=20 Medical Informatics. Van Bemmel JH, Musen
MA, = eds.=20 Heidelberg/New York: Bohn Stafleu Van Loghum,
Houten, and=20 Springer-Verlag, 1997;67=E2=80=9374.
3. = Health Level=20 Seven. An application protocol for electronic data
exchange in=20 healthcare environments, Version 2.3. Ann Arbor,
MI: = Health=20 Level Seven, Inc.; 1997.
4. = Defense=20 Information Systems Agency. ASC X12 Standards for
Electronic=20 Document Interchange, 3072, Alexandria, VA: DISA
1997.
5. = National=20 Council for Prescription Drug Programs, Inc.
Telecommunication=20 Standard Format. Version 3.2. Phoenix,
AZ: = National=20 Council for Prescription Drug Programs, Inc.;
1992.
6. = Pollock D.=20 Data Elements for Emergency Department Systems
(DEEDS)=20 Release 1.0. Atlanta, GA: Division of Acute Care,
Rehabilitation=20 Research, and Disability Prevention, National
Center for=20 Injury Prevention and Control, Centers for Disease
Control and=20 Prevention, January 1997.
7.=20 Hewlett-Packard. Trends in Healthcare Computing. = Eighth
Annual=20 HIMSS/Hewlett- Packard Leadership Survey; 1997.
8. = Comit=C3=A9=20 Europ=C3=A9en de Normalisation. Standard = Communications
Protocol for=20 Computer- Assisted Electrocardiography. CEN
TC251 = WG3 ENV=20 1064. Brussels, Belgium. Comit=C3=A9 Europ=C3=A9en
de=20 Normalisation; 1993.
9. = Forrey AW,=20 McDonald CJ, DeMoor G, Huff SM, Leavelle D,
Leland D, Fiers=20 T, Charles L, Griffin B, Stalling F, Tullis A,
Hutchins K,=20 Baenziger J. The logical observation identifier
names = and codes=20 (LOINC) database: A public use set of codes
and = names for=20 electronic reporting of clinical laboratory test
results.=20 Clin Chem 1996;42:81=E2=80=9390.
10. = American=20 College of Pathology. SNOMED: Systematized
Nomenclature=20 of Medicine. 2nd ed., vs. 1 and 2. Skokie, IL:
American=20 College Pathology; 1996.
E
CONOMIC,=20
S
OCIAL, AND=20
P SYCHOLOGICAL O
BSTACLES TO=20
A
CHIEVING THE=20
N ATIONAL H
EART=20
A
TTACK=20
A
LERT=20
P
ROGRAM=E2=80=99S=20
G OALS
Joseph P.=20 Ornato, M.D., F.A.C.C., F.A.C.E.P.
Professor and=20 Chairman
Department of=20 Emergency Medicine
Virginia=20 Commonwealth University
Medical College=20 of Virginia
Richmond,=20 VA
This=20 presentation addresses some of the major challenges of
the National=20 Heart Attack Alert Program (NHAAP) and
examines=20 specific economic, social, and psychological
barriers to=20 achieving the goals of the program. Terms that
might be=20 confusing to a lay audience are defined. Lessons
learned from=20 everyday experiences and observations related
to = the care=20 of patients with acute cardiac problems are
presented.=20 Opportunities for creating a technological
partnership=20 within the emergency medical services (EMS)
system are=20 described, and methods for bridging emergency
medicine=20 (including cardiology) and medical informatics are
explored.
The=20 Challenges of Phase I of the NHAAP
Medically, a=20 heart attack is an occlusion or blockage in an
artery=20 nourishing the heart, with the potentially resultant
death = of heart=20 muscle tissue. This occlusion is what causes
the = chest pain=20 and other symptoms. The lay press often
incorrectly=20 defines =E2=80=9Cheart attack=E2=80=9D as a sudden cardiac = arrest.
In = cardiac=20 arrest, the heart stops beating suddenly and
unexpectedly;=20 within seconds, the brain and other vital
organs are=20 deprived of necessary blood and oxygen, and the
patient loses=20 consciousness. Unless cardiopulmonary
resuscitation,=20 defibrillation, or other interventions are
applied=20 quickly, the patient will die. Sudden cardiac arrest
occurs as a=20 complication of a heart attack, or an acute
myocardial=20 infarction (AMI).
Most = deaths=20 from AMI are caused by cardiac arrest, usually
within 1 or 2=20 hours of the onset of the coronary artery
occlusion. This=20 narrow timeframe for prevention of cardiac
arrest provides=20 the major challenge for the NHAAP: to not
only = increase=20 awareness and education of patients, potential
patients, and=20 those around them, but more importantly, to
influence their=20 behavior. Patients must not only recognize
AMI = symptoms,=20 but also take the essential next step=E2=80=94seek
emergency=20 care.
Facilitating=20 this initial link between awareness and action is
a = major concern=20 of the NHAAP and the foundation of the
Rapid = Early=20 Action for Coronary Treatment (REACT)
project. The=20 findings of the REACT project may indicate
that = public=20 education has some measurable benefit but, by
itself, is not=20 enough to substantially reduce patient delays in
seeking care.=20 The existing system for emergency cardiac
care = is not=20 user-friendly and it is not focused on the = patient=E2=80=99s
20
psychological,=20 emotional, economic, and sociologic needs.
As a = result, the=20 patient undergoes a complicated reasoning
process that=20 delays his/her decision about whether to seek
emergency=20 medical care. For these reasons, reducing the
time = for=20 patient/bystander recognition and action, the focus
of = Phase I of=20 the NHAAP, remains the greatest challenge. In
the = other phases=20 of the program substantial progress has
been = made,=20 particularly through the work of the 60-minutes-
to-treatment=20 working group in Phase III (the hospital phase).
Varying=20 Perspectives on Emergency Cardiac Care
Patients
The = process of=20 emergency cardiac care can be explored from
the = perspectives=20 of patients, EMS providers, and emergency
department (ED)=20 personnel. When a patient experiences
potential AMI=20 symptoms, he/she must decide whether to
consult a doctor=20 or spouse, take antacids, or, as advised by
numerous health=20 organizations, call 9-1-1. That decision is
probably one of=20 the most difficult and serious in a person=E2=80=99s
lifetime.=20 However, there are significant barriers to calling for
emergency care.=20 In fact, patients who have experienced
previous=20 coronary events often delay calling longer than those
with = no prior=20 events. Interviews conducted through the
NHAAP = and the=20 American Heart Association revealed that
patients=E2=80=99=20 experiences had caused them to fear loss of control,
embarrassment,=20 inconvenience, and costs. Psychological
aspects of=20 patient delay need further study.
Emergency=20 medical services providers
One of = the most=20 difficult problems for paramedics or
firefighters is=20 identification of the patient. Often, important
medical=20 information is unavailable because the patient is
unresponsive,=20 there is no identification on the person, or no
one is = accompanying the patient. Lack of medical information
about = the=20 patient makes it difficult to make appropriate
decisions=20 concerning medications and treatment.
Although 12-lead=20 electrocardiograms (ECGs) are essential for
quickly=20 assessing the patient=E2=80=99s condition, only a handful = of
EMS = systems are=20 using them in the field. In addition, most
EMS = ECG systems=20 are unable to communicate with hospital
ECG = systems to=20 compare information. Therefore, when
paramedic units=20 in most cities are dispatched to a patient with
chest = pain,=20 transportation of the patient to the hospital is
virtually=20 required to gain access to the hospital=E2=80=99s ECG = system.
Although useful=20 information may exist, it cannot always be
accessed in the=20 most helpful way. For example, most
paramedics=20 document data manually. Although there are some
rudimentary=20 documentation systems, they cannot integrate
information from=20 monitors into a time-accurate, comp-
rehensive=20 database from the field. The Data Elements for
Emergency=20 Department Systems project, sponsored by the
Centers for=20 Disease Control and Prevention, is a possible
model = of a=20 database structure that might allow development
of = systems that=20 could provide patient information to be shared
across = EDs.
Emergency=20 department personnel
The = ED staff=20 experience many of the same problems as the
EMS = providers.=20 In many cases, no prior medical information
is = available.=20 In the optimum situation, a 12-lead ECG is
used = upon the=20 patient=E2=80=99s arrival at the hospital, results from a
prior = ECG at=20 the hospital are readily available electronically,
and a = comparison can be made. More often, the ED staff
must = try to get=20 information from doctors=E2=80=99 offices or other
hospitals by=20 fax. They also have problems with multiple
data = sources,=20 particularly because they have individual
timepieces set=20 at varying times. In time-dependent events
like = a cardiac=20 resuscitation, a dissimilarity of plus or minus
2 or = 3 minutes=20 can make a tremendous difference. A study in
the = April 1998=20 issue of the Annals of Emergency Medicine
examined=20 various time-keeping devices used in EMS
situations,=20 including: wall clocks; watches on nurses,
doctors, and=20 firefighters; and timepieces on defibrillators
and = fire=20 trucks.
1
When = this=20 information was combined with a
simulation of=20 5,000 cardiac arrests, 93 percent of cardiac
arrest cases=20 would contain a documentation error of
=E2=89=A5 2 minutes and 41 percent of cases would contain=20 a
documen-tation=20 error of
=E2=89=A5 5 minutes. A month = after
synchronizing=20 all timepieces, the 2-minute documentation
error = rate was=20 cut in half and the 5-minute documentation
error = rate was=20 cut by three-fourths. However, after 2 more
months with no=20 additional synchronization, the original
levels of=20 inaccuracy returned.
The = problem of=20 manual documentation in the ED also is
similar to the=20 situation in the field. Although the ED has
more = electronic=20 devices, most of them cannot communicate
with = one=20 another=E2=80=94or if they can, they do not share the same
time-base=20 language, which adds to the confusion.
Future=20 Solutions
There = are=20 several potential technological strategies for
overcoming=20 current barriers to improving emergency
medical care.=20 These include: (1) identifying all patients
through devices=20 similar to those currently used for
fingerprint or=20 retinal technology; (2) making online
universal=20 medical records available to EMS and ED medical
personnel; (3)=20 linking devices with a common language and
data-synchronizing=20 them to the atomic clock; and (4) linking
EMS = vehicles to=20 the hospitals through prehospital
telemedicine to=20 facilitate evaluation of selected patients in
the = field.=20 Assuming that these technological advances could
be=20 accomplished, EMS practices could undergo related
changes.=20 Several station wagons or sports utility vehicles
could = be added=20 to the fire trucks and ambulances used as
emergency=20 units. In addition to being staffed with
paramedics,=20 they could be equipped with a 12-lead ECG and
a = portable=20 telemedicine system. Trained paramedics at the
9-1-1 = dispatch=20 centers could use specially constructed
protocols to=20 stratify patients according to risk. A paramedic
would = be=20 dispatched quickly=E2=80=94without lights and = sirens=E2=80=94to
those = patients=20 assessed as relatively low risk. The paramedic
would = evaluate=20 the patient, do a 12-lead ECG, and send it to
21
the = physician at=20 the hospital via a telemedicine link. Further
followup could=20 occur the next day. This system would offer
the = public=20 another possible outcome of a 9-1-1 call: prompt
medical=20 assistance at relatively low cost and with no
embarrassment,=20 as well as expert assistance in deciding
whether=20 hospitalization is necessary.
In = this new=20 system, medical record prehospital data would
be = available,=20 and prehospital ECGs would become standard
throughout the=20 country. Telemedicine consultations would
be = available in=20 many areas, with the ability to link data with
the = patient=E2=80=99s=20 primary care physician. All medical devices
would = contribute=20 time-synchronized data on a wireless
network (e.g,=20 via satellite). The data captured by various
devices would be=20 combined in one unified, time-sensitive
electronic=20 record, and the system would be integrated into a
quality=20 improvement followup mechanism. Within a system
such = as this,=20 the NHAAP could achieve its goals.
To = overcome=20 current obstacles, it will take a vision of the
future; a=20 partnership among the EMS system, emergency
medicine, and=20 medical informatics; as well as a favorable
climate of=20 research support for implementing these ideas.
With = all these=20 components in place, it will be possible to
develop an=20 integrated, readily available information system
for = emergency=20 medical care.
Reference
1. = Ornato JP,=20 Doctor ML, Harbour LF, et al. Synchronization of
timepieces to=20 the atomic clock in an urban emergency medical
services=20 system. Ann Emerg Med = 1998;31:483=E2=80=9387.
Presentations=20 are organized around five informatics domains: telehealth; education; = medical=20 records access; diagnostic and
therapeutic=20 decision support; and large-scale databases. For each domain, there is = an=20 overview of the presentations in that area,
followed by=20 the abstracts of those = presentations.
Telehealth
Carole Gassert,=20 Ph.D., R.N. (Chair),
CAPT = David W.=20 Ferguson, M.C., U.S.N.
(Vice = Chair)
During the=20 telehealth portion of the
symposium, Dr.=20 Forrest Faison III reviewed
efforts to=20 expand the infrastructure for
telemedicine=20 programs in the U.S.
Department of=20 Defense using as many
standardized=20 technologies as possible. Future
efforts will=20 expand the Department of
Defense=20 telehealth infrastructure.
Dr. = Napoleon=20 Knight discussed how the
Carle = Rural=20 Telemedicine Project enables
physician=20 assistants at three rural hospitals in
Illinois to=20 administer thrombolytic therapy to
patients. These=20 clinicians are linked through
telehealth=20 technologies to physicians who
provide=20 diagnoses and recommend
interventions=20 from the hub site. The project
hopes = to expand=20 these same services to
additional=20 rural sites and to involve nurses in
providing=20 thrombolytic therapy. Dr. Marian
LaMonte=20 presented a project that allows
thrombolytic=20 therapy to be administered
during=20 ambulance transport to patients who
suffer an acute=20 ischemic stroke. A similar
project could=20 be used to administer
thrombolytic=20 therapy to patients who
experience a=20 myocardial infarction.
Dr. = Jay=20 Sanders, Dr. Angelo Alonzo, Mr.
Jack = Horner,=20 Ms. Linda Roman, and Dr.
Charles Safran=20 formed a five-member panel
to = present=20 information about an application
of = telemedicine=20 termed =E2=80=9Celectronic house
calls.=E2=80=9D The=20 panel discussed existing
telehealth=20 projects and how those or similar
efforts could=20 be used to meet National Heart
Attack Alert=20 Program objectives. Behavioral
and=20 organizational issues of using tech-
nology were=20 presented. Telehealth
technologies=20 are increasingly available to
health care=20 clinicians and patients, but
speakers=20 reminded the audience that many of
these = individuals still have technophobia and
will = need=20 assistance in adopting telehealth
technologies.=20 It was suggested that both
clinicians and=20 patients receive training to use
the = technology=20 before a crisis event occurs,
not = on a=20 just-in-time basis.
Speakers noted=20 that telehealth tech-
nologies may=20 change clinician-patient
relationships.=20 Placing telehealth technologies
in = the home=20 puts the clinician in the role of
guest = while the=20 patient maintains control of
the = encounter.=20 Presenters also felt that
technology=20 increases the frequency of
clinician-patient=20 contacts. Patients with home
technologies=20 expect to contact clinicians
more = easily for=20 follow-up advice than they
would = through=20 routine office visits.
Discussions=20 emphasized that in order to
be = successfully=20 implemented, telehealth
technologies=20 need to be flexible concerning
patient=20 confidentiality issues, and secondly,
they = need to be=20 easily integrated into the
workflow of=20 clinicians. A further need was
recognized for=20 improving the infrastructure
to = link=20 telehealth technologies with other
information=20 technologies used in health care.
It = also was=20 suggested that telehealth
solutions be=20 considered from the perspective
of = whether to=20 pursue a low-cost and broad
approach or a=20 higher-cost and more targeted
approach to=20 change behavioral reactions
surrounding the=20 onset of heart attack
symptoms. It=20 was further suggested that
targeting=20 high-risk patients or those who
have = already=20 made a 9-1-1 call, for example,
might = result in=20 a more effective intervention
for = patients=20 responding to acute ischemic
symptoms. When=20 developing telehealth
solutions,=20 limited resources=E2=80=94particularly in
the = rural=20 areas=E2=80=94may require that lower-end
technologies be=20 considered whenever
possible. S
YMPOSIUM=20
P
RESENTATIONS=20 AND
A BSTRACTS=E2=80=94 W
EDNESDAY, APRIL=20 15, 1998
22 T
ELEMEDICINE=20 IN THE
D
EPARTMENT=20
OF=20
D
EFENSE=20
(DoD)
C. = Forrest=20 Faison III, M.D., CDR, MC, USN
Director
Telemedicine=20 Program Office
Department of=20 Defense
Bethesda,=20 MD
How = this idea=20 can help the
National=20 Heart Attack Alert Program
The = use of=20 telemedicine technologies in the prehospital
setting can=20 enhance survival and treatment en route to
definitive care.=20 The use of asynchronous communications,
whether for=20 electrocardiogram interpretation or synchronous
communications=20 to allow specialty treatment through an
extended care=20 provider, enhances specialty access to
cardiologic=20 care, improves early diagnosis, and, through
well-defined=20 metrics analysis, yields valuable data to both
refine = care and=20 plan system expansion.
What=20 currently is being done in this area
Telemedicine is=20 a $325 million investment for the DoD,
which = has=20 contributed approximately 58 percent of the total
Government=20 (Federal and State) investment in this
technology.=20 There are currently 197 ongoing telemedicine
projects in the=20 DoD. Most notable for early cardiac care is a
cooperative=20 venture in the greater San Antonio area where
ambulances are=20 being equipped with synchronous
communications=20 that are transmitted through an embedded
fiberoptic=20 network used for traffic control. This synchronous
communications=20 system allows cardiologists to direct care in
the = field while=20 minimizing communications costs through
use of = intrinsic=20 communications systems.
Obstacles to=20 successful implementation of the idea
For = the DoD, the=20 cost of use in areas outside the continental
United = States is=20 a primary obstacle. Standardized
methodologies=20 for providing uniform care and creating
nonstandard=20 technical solutions are not yet realized, and
thus = remain=20 primary obstacles.
Steps=20 necessary to implement the idea
Telemedicine is=20 a mature technology, and use of commercial
off-the-shelf=20 products is the most cost-effective alternative.
Cooperative=20 partnerships for cost-sharing and controlled
projects to=20 develop standard technical and clinical solutions
are = the most=20 logical methodologies for implementing
standardized,=20 cost-effective telemedicine in the United
States. U
SE = OF=20
T
ELEMEDICINE=20
T ECHNOLOGY
TO =
A
SSIST=20
S MALL R URAL H OSPITALS
Napoleon=20 Knight, M.D.
Co-investigator
The = Carle Rural=20 Telemedicine Project
Medical=20 Director
Department of=20 Emergency Medicine
Carle = Foundation Hospital
Carle = Clinic=20 Association
Urbana,=20 IL
How this=20 idea can help the
National=20 Heart Attack Alert Program
This = type of=20 program can assist small rural hospitals in their
ability to=20 rapidly diagnose and treat patients that present for
care = of acute=20 myocardial infarction at a small rural hospital.
As = more and=20 more rural hospitals are faced with financial
difficulties,=20 using the type of arrangement that we have with
John = and Mary=20 Kirby Hospital may enable them to continue
to = provide high=20 quality care in their communities.
What=20 currently is being done in this area
Since = 1994 we=20 have been using Telemedicine Technology
Vtel = units with=20 full T-1 lines to link the emergency
department (ED)=20 at Carle Foundation Hospital with the ED
at = Kirby=20 Hospital. Kirby is the smallest hospital in Illinois,
and = at the time=20 that we started our program, they were faced
with = the=20 possibility of having to close their ED due to the
financial=20 difficulty that they were having trying to keep it
open. = We worked=20 with them to develop treatment protocols
that = the=20 physician assistants (PAs) used to treat and evaluate
patients. This=20 is a low-volume ED, seeing 8-12 patients per
day. = The=20 majority of patients that are seen are low acuity,
and = so=20 utilizing the PAs with experience in emergency care,
and = with=20 advanced trauma life support, advanced cardiac life
support, and=20 pediatric advanced life support training has
worked quite=20 well. For patients who do not fit the =E2=80=9Csee,
evaluate,=20 treat, and discharge protocols,=E2=80=9D the PA links with
us in = our ED,=20 and we help with