Learning Center
Plans & pricing Sign in
Sign Out



									Application of an XML-Based Store-Forward System to Urban
P. Tarczy-Hornoch, MD1,2,3, T. Cannava, MHA3, W. Kramer2,3, S. Dyck, MHA3, R. Colven,
C. Haycox, MD, PhD4, R. Gray3, T. Norris, MD, MHA2,3,5, S. Fuller, PhD2,3
1Division of Neonatology, 2Division of Biomedical and Health Informatics, 3Telehealth

Program, 4Division of Dermatology, 5Department of Family Medicine, University of

The overall question the investigators               REQUIREMENTS: URBAN
sought to explore in this study revolves             TELEDERMATOLOGY STUDY
around the role of store-forward                     A key requirement for the dermatologists
teledermatology in an urban primary                  participating in the study was that the
care/tertiary care spoke/hub setting. The            system permit the creation of a UW-
University of Washington Physicians                  specific, store-forward, teledermatology-
Neighborhood (UWPN) Clinics is a                     specific, electronic consult request form. A
network of primary care clinics closely              related requirement was that the data on
affiliated with the University of                    this form be stored in a database for
Washington (UW). The UWPN clinics are                aggregate analysis of study data. The
located primarily in urban settings                  clinicians at the participating UWPN clinic
throughout the greater Seattle area1. The            required that the system integrate into the
bulk of tertiary care services and referrals         existing EPIC-based paperless clinic. In
for these clinics is handled by UW. Access           addition, the telemedicine investigators
to referral services is good but can be              required that the system have an
limited by temporal constraints (the long            acceptable level of security and be scalable
lead time in scheduling appointments with            to many UWPN sites and potentially to
certain subspecialists) and by convenience           rural sites (those in need of store-forward
(time and frustration costs of commuting in          teledermatology consultation and willing to
a metropolitan area with some of the worst           participate in a study).
traffic conditions in the Nation).
                                                     Specific to the store-forward study (as
In designing the store-forward                       opposed to completion of a teleconsult), the
teledermatology study in the context of the          dermatologists involved with the study had
Institute of Medicine report on evaluating           a number of questions they wanted to
telemedicine 2, the investigators identified         answer: What is the impact of
system requirements specific to a store-             teleconsultation on a primary care
forward teleconsultation system being used           provider’s diagnosis? Treatment? Level of
for an evaluation study. Existing systems            confidence? The telemedicine investigators
at the time of the initiation of the study did       also had process questions such as How
not meet the requirements; thus, a decision          much time is involved in preparing a
was made to build a Web-based store-                 teleconsultation? What are the technical
forward system. The system was designed              problems? Is there a contingency plan in
to be easily customizable for different              the absence of telemedicine services? What
clinical domains and different types of              is the ultimate disposition of the patient?
studies, since these requirements did not
appear unique to teledermatology and                 REQUIREMENTS: STORE-FORWARD
since there were other clinical domains              SYSTEM
that the UW Telehealth Program3 was                  Meeting the specific narrow requirements
interested in investigating.                         of the urban store-forward teledermatology

study (UW and UWPN) and then                        provider, specialty consultant, and system
generalizing these requirements to other            administrator all access the database via
clinical domains, other store-forward               the Web. Data extraction for analysis
teleconsultation studies, and other sites           purposes is done on the server side and not
(e.g., non-UW and nonurban) led to the              via the Web. The ASPs used are compatible
following requirements: (1) a “thin” client         with all major browsers, which permits
with no software installation (especially for       adding rural sites and users with no
heterogeneous rural sites), (2) use of the          remote software installation.
existing Internet infrastructure as a
communication backbone, (3) ability to              The system is designed to be table and
integrate into an existing clinic/referral          form driven. The administrator can add
workflow,                                           new specialists and/or specialties as well as
(4) adequate security measures, (5)                 change who is “on call” for
configurable consultation templates                 teleconsultations using the system’s Web-
(clinical domain- and/or consultation-type-         based administration tools. Both
specific), (6) storage of the data from the         consultation and study forms are encoded
consultations in a database for analysis,           in a simple XML notation that includes
(7) similar configurable evaluation-study-          tags for free-text entry, “pick one of list,”
specific templates                                  “pick multiple off list,” and “select value in
(preconsultation/postconsultation for the           range.” Any number of images (up to
primary care provider, postconsultation for         administrator-configurable
the consultant), (8) similar storage of the         MAX_IMAGES) can be attached to a
data from the study-specific templates in a         consultation. For maximum flexibility, the
database for analysis, and (9) open-source          system allows images in the common
design to permit adaptation of the system           standard image formats, permitting
for related uses by other telemedicine              primary care providers to use a variety of
investigators.                                      digital cameras.

RESULTS: IMPLEMENTATION OF SYSTEM                   At the UWPN Factoria study site (and
We built a system to meet the                       subsequently other UWPN sites for an
requirements outlined above. To minimize            expanded study), the templates were
the costs of replicating the system, the            customized to integrate with the EPIC
Windows NT and Microsoft BackOffice                 electronic medical record system.
software development environment was                Specifically, templates were developed for
chosen. The system is secured by applying           the EPIC system that used extracted
the latest patches, disabling all                   information from the EPIC system for
nonessential services, restricting privileges       incorporation into the store-forward
to the bare minimum necessary for users             teledermatology template developed by the
and applications, and using Tripwire to             dermatologists at UW. The cameras used
monitor for any intrusions. Secure Sockets          at the UWPN sites were Sony Mavica
Layer connections over the standard                 cameras, which recorded images on floppy
Internet are used to encrypt data                   disks for ease of use in the clinic setting
transmission. Authorization and                     (requiring only a standard floppy drive on
authentication are via encrypted user               the EPIC workstation computer rather
names and regularly updated passwords as            than a large number of camera interfaces
well as monitoring of usage logs.                   throughout the clinic).
Consultation and evaluation study data
are stored without patient identifiers in an        RESULTS: URBAN TELEDERMATOLOGY
SQL Server database. Access to the                  STUDY
database is via Active Server Pages (ASPs)          At the time of preparation of this abstract,
along with server side COM objects (third           accrual of patients into the evaluation of
party and locally developed). Primary care          urban store-forward teledermatology has
been completed, and final analysis of the          dermatologists) and similar preliminary
results is under way. Sixty-five consults          results from users in rural areas, it
were requested by five clinicians (including       appears that this model could be extended
a physician assistant) over a 22-month             to evaluation studies of store-forward
period (an average of three consults per           telemedicine in other clinical domains.
month). A preliminary analysis of the data         Preliminary analysis of the results of the
shows that the average time to prepare a           urban store-forward teledermatology study
teledermatology consultation was                   suggests that it plays a role as an adjunct
13 minutes. On an 11-point Likert scale            to in-person referral in an urban setting,
(0=Not Necessary to 10=Essential), the             but in a setting where in-person visits are
primary care providers rated the necessity         readily available, it does not obviate the
of the consultation as                             need for a significant number of in-person
7 on average. Prior to the teledermatology         referrals (approximately 40 percent of the
consultation, the primary care providers’          time).
levels of confidence in their diagnosis and
treatment plan were 5.5 and 5.7,                   ACKNOWLEDGMENTS
respectively, on average on a similar Likert       This project has been funded in part with
scale. There were issues with familiarity          Federal funds from the National Library of
with and experience in taking photographs          Medicine under Contract No. N01-LM-6-
of dermatological lesions despite training         3545. We thank the clinicians and staff at
and a “crib sheet,” but these issues               the Factoria UWPN Clinic for their
improved over time as the primary care             participation in this study.
providers and dermatologists interacted
with one another. The preliminary analysis         REFERENCES
of the dermatologists’ assessment of need          1. UW Physicians Online: Map of Practice
for additional in-person contact with the          Locations,
patient suggests a bimodal distribution, ,
                                                   accessed 2000.
with approximately
                                                   2. Institute of Medicine. Telemedicine: A
40 percent of the time feeling an in-person
                                                   Guide to Assessing Telecommunications for
visit was necessary despite the                    Health Care, 1996; Institute of Medicine.
teleconsultation. The consultation raised          3. University of Washington Telehealth
the primary care providers’ confidence in          Program,
the diagnosis and treatment plan to 7.9  ,
and 8.2, respectively. Overall satisfaction        accessed 2000.
with the process was 8.7 on a Likert scale
(ranging from 0=Very Dissatisfied to
10=Very Satisfied). On a scale from
0=Much Worse to 10=Much Better, the
average quality of the teleconsultation
process, compared with an in-person visit,
was rated 6.5.

The XML-based, configurable, store-
forward telemedicine system developed at
UW was successfully applied to a real-
world evaluation study of the potential role
and utility of store-forward
teledermatology as an adjunct to in-person
referral in an urban setting. On the basis
of the high user acceptance of the tool (by
both primary care providers and

To top