<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta>
      <journal-title-group>
        <journal-title>IWSG</journal-title>
      </journal-title-group>
    </journal-meta>
    <article-meta>
      <title-group>
        <article-title>REMEDI - Tactics of Expanding a Science Gateway Community</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <string-name>Claire Stirm, MS</string-name>
          <email>cfrist@purdue.edu</email>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Rich Zink, MBA</string-name>
          <email>zinkr@purdue.edu</email>
          <xref ref-type="aff" rid="aff1">1</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Sandra Gesing, PhD</string-name>
          <email>sandra.gesing@nd.edu</email>
          <xref ref-type="aff" rid="aff2">2</xref>
        </contrib>
        <aff id="aff0">
          <label>0</label>
          <institution>Purdue University, Information Technology Research</institution>
          ,
          <addr-line>Computing, West Lafayette, IN</addr-line>
          ,
          <country country="US">USA</country>
        </aff>
        <aff id="aff1">
          <label>1</label>
          <institution>Purdue University, Regenstrief Center for Healthcare, Engineering</institution>
          ,
          <addr-line>West Lafayette, IN</addr-line>
          ,
          <country country="US">USA</country>
        </aff>
        <aff id="aff2">
          <label>2</label>
          <institution>University of Notre Dame, Center for Research Computing</institution>
          ,
          <addr-line>Notre Dame, IN</addr-line>
          ,
          <country country="US">USA</country>
        </aff>
      </contrib-group>
      <pub-date>
        <year>2018</year>
      </pub-date>
      <volume>13</volume>
      <fpage>13</fpage>
      <lpage>15</lpage>
      <abstract>
        <p>-The Regenstrief Center for Healthcare Engineering (RCHE), an interdisciplinary research center located at Purdue University, facilitated the development of a scientific gateway for medical device informatics, known as Regenstrief National Center for Medical Device Informatics (REMEDI). Initially formed in 2009, REMEDI is an evidencebased community of practice, focused on smart pump technology and infusion therapy safety. REMEDI is a virtual community of pharmacists, nurses, other clinicians, infusion pump vendors, researchers, and national organizations such as the American Association of Medical Instrumentation (AAMI) and the American Society of Health-System Pharmacists (ASHP). In 2017, in response to a need identified by AAMI, REMEDI expanded its scope to include new applications, databases, and community engagement spaces for respiratory therapists, nurses, and ventilator vendors.</p>
      </abstract>
      <kwd-group>
        <kwd>Science Gateways</kwd>
        <kwd>patient safety</kwd>
        <kwd>hospital selfanalysis</kwd>
        <kwd>evidence-based community</kwd>
        <kwd>applications</kwd>
      </kwd-group>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>I. INTRODUCTION</title>
      <p>
        Regenstrief National Center for Medical Device Informatics
(REMEDI) was created to address patient safety issues associated
with the administration of medications using infusion pumps.
According to the US Food &amp; Drug Administration (FDA), in a
5year period 2005 – 2009, there were 56,000 adverse events and
numerous deaths associated with infusion devices [
        <xref ref-type="bibr" rid="ref1">1</xref>
        ]. Smart
infusion pumps, those infusion pumps containing a library of
medications with dosing guidelines are intended to ensure safe
intravenous medication use by preventing over and under dosing,
were in use by 72.9% of hospitals in 2013 [
        <xref ref-type="bibr" rid="ref2">2</xref>
        ].
      </p>
      <p>Since many hospital patients receive infusions, the REMEDI
project was initiated in response to the national priority to improve
patient safety for infusion pumps and to a request from the
Indianapolis Coalition for Patient Safety. The mission and primary
activity of the collaborative are to conduct activities that improve
patient safety and the quality of healthcare delivery.</p>
      <p>Organizing 34.8 million alerts and compliance data representing
nearly 136 million infusions in a database, the REMEDI science
gateway is used for benchmarking and provides pharmacists, and
other clinicians, evidence to make more informed decisions related
to infusion pump medication administration therapy. Community
members drive REMEDI operations. Through regular meetings,
representatives of the member hospitals share their knowledge in a
collaborative environment, fostering knowledge development to
advance medication administration practice and improve patient
safety</p>
      <p>Membership in REMEDI includes hospitals of many types (e.g.,
critical access hospitals, university teaching/research hospitals,
safety net hospitals, etc.) and is provided at no cost to members that
are willing to share their data and knowledge. Current membership
includes approximately 50 systems, representing 300 facilities in 24
states, plus 1 hospital in Costa Rica. More than 120 clinicians have
used REMEDI analytics tools to generate over 60,000 reports since
2009.</p>
      <p>During this gateway expansion, REMEDI and the teams
supporting the platform have developed best practices for
onboarding new community members, designed a new system for
accepting hospital contracts, introduced new applicationsfor new
community members, and initiated outreach plans targeting future
members. Through all of this work, REMEDI has developed a few
tactics for handling growth and has set future goals for continual
growth.</p>
      <p>II.</p>
      <p>BACKGROUND</p>
      <p>
        In 2009, medication errors topped the list of adverse events
occurring in hospitals [
        <xref ref-type="bibr" rid="ref3">3</xref>
        ]. Realizing that alarm safety was an
emerging concern among hospital administrators and national
regulatory organizations, Regenstrief Center for Healthcare
Engineering (RCHE) focused attention on this escalating patient
safety challenge in 2010 before it reached the level of high national
awareness [
        <xref ref-type="bibr" rid="ref3">3</xref>
        ]. RCHE is exponentially expanding to enable an even
greater impact on the safety and quality of healthcare delivery
nationwide. We are actively working on projects to 1) significantly
increase the capability of the current system to allow more users and
move users away from using a local system that does not encourage
data sharing to a collaborative platform; 2) capture all infusion
pump alarms, not just those generated against the drug limit
libraries; 3) dramatically increase the capabilities to include
physiological sensing monitor data; and most important 4) linking
smart pump and physiological sensing monitor data to anonymized
patient data to better measure impact. Forming new communities
around these additional medical devices will provide fertile ground
for Purdue researchers and our partners to examine not only the use
of these devices but how these communities advance patient safety
and quality.
      </p>
      <p>Currently, REMEDI communicates with their active
community through their online collaborative platform offering
forums, applications to interact with data records, and ability to
publish findings with the community through publications and
research results. Users of the REMEDI platform can perform
thousands of unique analyses by selecting different
combinations of hospital, date range, profile or care area,
facility, type of limit (hard/soft), infusion duration, drug or fluid,
action taken by clinician, alert type, and field limit type. Once
we have enough information, the goal is to provide a standard
shared alert level based on patient type and other values in order
to provide hospitals with enough information to make informed
decisions. REMEDI also sends direct email announcements and
shares marketing information through the Purdue University
Research social media accounts on Facebook, Twitter,
YouTube, and Google+. No patient or sensitive data is shared
outside the REMEDI platform. REMEDI also hosts regular
community webinar meetings and an annual community
conference. During the webinars and conference, the REMEDI
team interacts with community members to analyze the scope
for new features or improvement plans for existing applications.</p>
      <p>REMEDI is funded by the Regenstrief Foundation, which is
why membership is currently free for the REMEDI community.
The Regenstrief Foundation was created as part of a $17 million
endowment left by Sam Regenstrief with the intention to
improve healthcare delivery.</p>
      <sec id="sec-1-1">
        <title>A. Competitors</title>
        <p>
          REMEDI has one direct competitor to their infusion pump
community, Bainbridge Health [
          <xref ref-type="bibr" rid="ref4">4</xref>
          ]. This company was formed
in 2017 and is also vendor-neutral, meaning they provide
applications that compare data between hospitals despite what
medical device company they buy their machines. REMEDI
is unique in the fact that they are a free resource, also
vendorneutral, identify hospitals by name, and provide a community
space on their online platform. Infusion pump vendors and
respiratory machine vendors are indirectly REMEDI’s
competitors because they also offer applications to view data.
These vendors may see REMEDI as a threat for they sell
consulting while REMEDI offers free community-driven
consulting. To our best knowledge, REMEDI is the only
project offering such services and is prepared to offer them
also at the international level.
        </p>
      </sec>
      <sec id="sec-1-2">
        <title>B. Publics</title>
        <p>REMEDI’s current audiences are pharmacists, nurses,
infusion pump vendors, national organization representatives,
and Purdue University researchers. These audiences are
primarily medical professionals actively working in hospitals.
All current members actively work with the infusion pump
applications. After the April 2017 conference, a survey was
sent to all attendees. Feedback from this survey has helped
pinpoint what the infusion pump community thinks about the
REMEDI platform, community interactions, and applications.</p>
        <p>
          The infusion pump community views REMEDI and their
support with a high satisfaction with room for improvement in
training (2.2/5.0) and documentation (2.0/5.0), see Figure 2, on
a five-point scale where 1 = Excellent and 5 = Terrible.
Conference participants also gave feedback on the benefits of
using REMEDI. One attendee stated that, “REMEDI has greatly
helped us identify areas for performance improvement with our
Plum360 infusion pump drug library alerts [
          <xref ref-type="bibr" rid="ref4">4</xref>
          ].” They also
found benchmarking data against other hospitals and
collaborating with hospitals to be a major benefit. Another
benefit that community members found was that REMEDI,
“simplified the reporting and investigative process for our site.
It takes away many of the challenges one is faced with in an
attempt to interpret data from the vendor's standard reporting
system [
          <xref ref-type="bibr" rid="ref5">5</xref>
          ].”
        </p>
        <p>
          Participants also gave input on where they would like to
see improvements with the platform. “I would like to be able
to pull more data when basic infusion mode is used - what care
profile, time of day, possibly what meds were programmed
before and after, etc. This would help me better understand
why basic infusion is being used so I can target my compliance
education where it is needed [
          <xref ref-type="bibr" rid="ref5">5</xref>
          ].” Others suggested more
documentation on using the different applications available on
the platform and more information on using the platform
overall.
        </p>
        <p>This feedback is in line with what participants said in the
overall satisfaction with REMEDI. From this feedback,
REMEDI has developed a plan to build out the documentation
for their infusion pump community. This information also has
helped us predict the needs of future communities.</p>
      </sec>
      <sec id="sec-1-3">
        <title>C. Platform</title>
        <p>
          Around 2011, the cyberinfranstructure team supporting
nanoHUB.org, a science gateway run by the Network for
Computational Nanotechnology (NCN) at Purdue University
with a mission to create, deploy, and operate a national resource
for theory, modeling, and simulation in nanotechnology,
realized that the web based solution that build nanoHUB is
needed by scientific disciplines [
          <xref ref-type="bibr" rid="ref6">6-7</xref>
          ]. The project HUBzero was
developed out of the Rosen Center for Advanced Computing
(RCAC) the computing research division of Information
Technology at Purdue (ITaP). The HUBzero team produced an
open source software platform that allows researchers from any
field or study to build a dynamic web site, called a Hub. On a
Hub, researchers can host analytical tools, publish data, share
resources, collaborate, and build communities [8].
        </p>
        <p>Every Hub is a self contained eco-system that supports any
user through their research process with the ability to host
analytical tools, publish data, share resources, collaborate, and
build communities in a single web-based ecosystem. The
HUBzero infrastructure includes a tool execution and delivery
mechanism based on Virtual Network Computing (VNC). Any
tool with a graphical user interface can be installed on a Hub
and deployed within a few hours. The jobs themselves can be
dispatched to the high performance computing resources. A
Hub can be enhanced by installing Jupyter notebooks, RStudio,
and other Web applications for easier tool development.
Anyone with a registered account can develop a tool through
these services and publish a finalized version on their Hub so
their peers can access the simulation or model they have
developed. Each Hub is built to be flexible. The creation of tools
and access to published tools can be controlled with
configurable permissions. Each tool is containerized and can
easily be deployed when a user requests to run the tool
simulation.</p>
        <p>Each Hub also comes with a guided system to help
registered users publish research products or data. Platform
managers can provide experts on the Hub with access to review
submitted materials before they are published. Every
publication can receive a digital object identifier (DOI) which
can be included in a white paper about the research or in a
journal publication so readers can access the data after the
completion of a study.</p>
        <p>Finally, every Hub comes with collaborative spaces for
research teams and peers to share ideas, track project progress,
and connect file services such as Google Drive, GitHub, or
Dropbox. Communities come with a system to manage new
memberships, share ideas through collections, discuss
important topics through forums, and brand their collaborative
space.</p>
        <p>REMEDI is hosted by HUBzero locally at Purdue
University in the United States. United States hospital patients
give their conceit that anonymized data can be collected and
used. Any patient data must remain on a secure system meeting
the Health Insurance Portability and Accountability Act of 1996
(HIPAA) compliance level. REMEDI is HIPAA compliant.
HUBzero has administrative safeguards including risk
management plans, risk analysis checks, sanction policy,
information system activity reviews, contingency plans,
security responsibility, information access management,
security awareness measures, security training, access control,
audit control, and other technical safeguards. All of these
policies and procedures come United States legislation to
provide data privacy and security provisions for medical
information.</p>
      </sec>
    </sec>
    <sec id="sec-2">
      <title>III. GROWING A GATEWAY COMMUNITY</title>
      <p>Existing REMEDI members, national healthcare priorities,
and national organizations have driven RCHE to explore
REMEDI’s expansion, capturing data from smart pumps,
increase data collection from monitors, and linking data.
Forming new communities around these additional medical
devices will provide fertile ground for Purdue researchers and
partners to examine not only the use of these devices but how
these communities advance patient safety and quality. Through
these relationships, REMEDI is in a unique position to continue
growing the entire REMEDI community to include many more
U.S. hospital members and additional international members.
REMEDI also can continue meeting the needs of current and
future community members through developing a space for
collaboration and build applications unique to their medical
devices and needs.</p>
      <sec id="sec-2-1">
        <title>A. Best practices for onboarding a new community member</title>
        <p>AAMI introduced two new communities as opportunities
for REMEDI’s growth. The ventilator application audience
would be made up of respiratory technicians, national
organization representatives, vendors, and community
administrators. The physiological monitor audience consists
of nurses, hospital leads, national organization
representatives, vendors, and community administrators. For
both of these communities, they required similar set-up as
how the infusion pump community was created on REMEDI.
The REMEDI team created applications that analyzed data
provided by each hospital and provided a community space
for each new group.</p>
      </sec>
      <sec id="sec-2-2">
        <title>B. Introducing new applications</title>
        <p>In October 2017, REMEDI soft-released a new set of
applications for respiratory therapy and physiology
monitoring. The Physiological Monitor Defaults application
was designed to address the answer of “where does one start
when monitoring patients?” The Physiological Monitor
Defaults application, see Figure 3, helps clinicians answer this
question by providing benchmarking tools to see where other
hospitals initially set their alarm thresholds for heart rate (HR),
peripheral capillary oxygen saturation (SpO2), systolic and
diastolic blood pressures (BP), end-tidal carbon dioxide
(EtCO2), and other alarm conditions. Additionally, clinicians
can use the application to compare detection priorities for
numerous alarm options for apnea, asystole, brady, vfib/vtach,
PVC, trigeminy, and more. The Ventilator Settings
application, allows respiratory technicians and other clinicians
the ability to examine the variation in the default settings
hospitals use for key ventilator alarms including positive
endexpiratory pressure (PEEP), peak inspiratory pressure (PIP),
tidal volume (VT), minute volume (MV), respiratory rate
(RR), apnea and disconnect priority. Users can benchmark
other hospitals by filtering on several hospital characteristics
(e.g., urban/rural, size, teaching/non-teaching, primary
population, care level, etc.).</p>
        <p>These two applications will be formally launched with
marketing campaigns from AAMI and the American
Association for Respiratory Care (AARC) in Spring 2018.</p>
      </sec>
      <sec id="sec-2-3">
        <title>C. Gathering feedback from potential Ventilator Device users</title>
        <p>Prior to the official launch of the Ventilator Settings
application, we contacted 18 medical professionals from
AAMI’s respiratory clinicians mailing list to gather early
feedback on their interest in community activities around this
application and the two new AAMI databases the REMEDI
team was developing. To understand this new audience, a short
five-minute survey was developed and sent out to members of
the AAMI mailing list. Through the survey we will be able to
understand who would make up the REMEDI ventilator
community and determine this group’s needs and customize the
platform to fit these needs. The survey is designed to meet the
following results
1. 35% of targeted participants will respond to the
survey and submit their feedback by November
17th.
2. 50% of respondents will express concern the
number of false alarms and increase in alarm
fatigue.
3. 85% of respondents will be interested in
participating in the ventilator community.</p>
        <p>We expect these results since these 18 medical
professionals expressed interest of acting as founding users for
the community when AAMI sent out information about the
possibility of this new community. The survey is an
anonymous online poll that allows text and radio button
answers. Some questions allow multiple choices to be
selected, while others enable participants only to choose one
response. Most questions allow participants to fill in an
“Other” text box if the options provided do not match their
answer. Concepts will be measured by how participants
structured their responses and key driving terms, such as
“alarm fatigue.”</p>
        <p>After the survey has closed, we reviewed the qualitative
and quantitative results through text answers and bar graphs.
Since the sample size is small, we have not use any additional
analysis tools to assist with analysis.</p>
      </sec>
      <sec id="sec-2-4">
        <title>D. Results and analysis</title>
        <p>The survey was sent out November 10th, 2017 and data
was collected on November 17th, 2017. There was a 39%
response rate. Participants answered what state they work in
and what primary work area or unit of the hospital they work
in, see Figure 4. From the survey, we were able to gather that
the new community would be made up of
receive alerts when new discussions occur at a daily, weekly, or
monthly digest rate. Members will be able to share documents
through a group shared file system, publish materials on the site
for public consumption related to patient safety best practices,
create wiki pages, blog posts, dynamic HTML pages for
community branding, and other activities to encourage members
respiratory technicians, nursing staff, patient safety officers,
hospital administrators, respiratory therapists, and respiratory
managers. These medical professionals expressed in the survey
that they are facing problems with productivity requirements
that are pushing therapists to cut corners such as, “1) limited
time to spend assessing and treating ventilated patients, 2)
limited time to spend orienting new employees, 3) limited time
to monitor quality.” Participants also stated that there are issues
with alarm fatigue/priority signals and volume of the signals.
Excessive alarms cause alarm fatigue for the medical
professionals monitoring the patients. The inconsistency with
alarms and what triggers an alarm is a need for almost all
community members from therapists to nurses. Thus, all of
these points should be discussion threads in the new community.</p>
        <p>Survey participants stated that they are interested in
participating in the following areas in order to improve patient
safety in the respiratory field:</p>
      </sec>
    </sec>
    <sec id="sec-3">
      <title>A. Sharing ventilator default settings B. Sharing ventilator alarm data C. Leveraging a database to analyze their alarms against other hospitals</title>
      <p>D. Sharing standards
E. Participating in online forums/discussion threads
F. Sharing tribal knowledge
G. Attending a virtual or in-person conference</p>
      <p>H. Speaking at a virtual or in-person conference</p>
      <p>Participants were able to select all that applied to them and
most showed interest in each area, see Figure 5.</p>
      <p>Sharing data, settings, and standards is important to this new
community. This enthusiasm to share information is promising.
Members will have access to applications to share hospital data
and machine settings once they have signed a contract with
REMEDI and officially registered as a member of REMEDI.
Members will be able to share standards and best practices
through an online forum connected to the community platform.
The platform is flexible enough for members to sign- up for to
to create a unique online space that fits their community’s needs.</p>
      <sec id="sec-3-1">
        <title>E. Accepting new hospital contracts</title>
        <p>As of December 2017, there are 301 facilities (hospitals
and infusion centers) connected with REMEDI.3 With a rate
of one hospital joining per week, onboarding new hospitals
used to take a lot of time. The REMEDI team would spend
around three hours per interested hospital answering questions
via email and providing a one-on-one webinar demonstrating
the community applications. If a hospital decided to sign-up
to be a member of REMEDI, they would have to sign a general
contract with REMEDI to agree with sharing data that will be
viewed by cohorts. No patient data is shared. Due to the
sensitivity of the shared data, REMEDI administers the data
with security and privacy requirements needed for HIPAA
data.</p>
        <p>
          To automate this process and expedite on-boarding
membership, REMEDI and HUBzero worked together to add
in a custom PDF with an electronic signature into the
REMEDI platform. This allows hospitals to sign-up to be a
member without going through an administrator of the website
first. We have also published online platform tutorials to help
with on-boarding and provide training documents to new
users. Streamlining this process has allowed a major
roadblock to be overcome and improved the onboarding
experience that will help REMEDI reach one of its goals to
continue expanding and include all 5,534 U.S. registered
hospitals and potentially international hospitals [
          <xref ref-type="bibr" rid="ref3">3</xref>
          ].
IV. PLANNING FOR THE FUTURE
        </p>
      </sec>
      <sec id="sec-3-2">
        <title>A. Gathering feedback</title>
        <p>The REMEDI team has realized the importance of
community feedback as the project has matured. The major
success of this gateway is the opportunity to tap into these
resources and gather feedback from both the larger group and
focused feedback from the steering committee, which is made
up of industry professionals who are aware of problems hospitals
and medical staff are facing with new devices. Use- cases and
design documents for new applications and databases are
developed in collaboration with the REMEDI steering
committee and the community during monthly meetings. The
community also provides feedback during the annual in-person
conference hosted by REMEDI.</p>
      </sec>
      <sec id="sec-3-3">
        <title>B. Outreach and marketing plans</title>
        <p>REMEDI needed to rebrand their current marketing
materials and the REMEDI gateway before the release to
make the portal more inclusive of new communities.</p>
        <p>The REMEDI and HUBzero team ran an analysis of the
platform’s audience and marketing materials to gauge where to
adjust the message, so it was more inclusive. Through this
analysis, the REMEDI team was able to discover several broken
communication paths and include new communities that have
recently joined in the marketing. This work helped the
engagement with AAMI since the materials and outlook for the
outreach campaigns AAMI was leading were already prepared
by us before meeting with the AAMI communication team.</p>
        <p>While there has been no formal communication launch,
AAMI and AARC reached out to community members to see if
they would be early adopters and help pre-populate data into
REMEDI’s new applications. AAMI and AARC are also
working on launching marketing materials for how to join
REMEDI, how to use both applications and capture data, and
finally how to access the new applications and databases.</p>
      </sec>
      <sec id="sec-3-4">
        <title>C. Addressing the international community</title>
        <p>REMEDI works exclusively with the US community and
US hospitals yet, analyses are underway in collaboration with
researchers working with hospitals in Australia and New
Zealand. While hospitals also on those continents may use the
same vendors for smart pumps, for example, there might be
diverse and additional vendors with different characteristics of
the instruments and available data.</p>
        <p>The main challenge is to address the health regulations in
the different countries and the sharing data policies by hospitals
[9]. The benefit for the users of sharing data via REMEDI seems
obvious. However, the procedure for hospitals participating in
REMEDI in the US might be not sufficient for further countries
or the sharing of data on an international level might need
various procedures. If the policies and regulations on an
international level will hamper applying the existing science
gateway in diverse countries, we anticipate to establish multiple
HUBzero instances with REMEDI services dedicated to diverse
countries. The long-term goal is to interchange data between
such instances as open as possible regarding regulations and
policies for each country. The more data can be exchanged, the
more the user communities can benefit from the services.</p>
        <p>The analyses for Australia and New Zealand are underway
and the next target for an international outreach are The
Netherlands since the HUBzero team closely collaborates with
researchers at the Amsterdam Medical Center.</p>
        <p>Setting up a European instance in The Netherlands might
open possibilities for further European countries. We are aware
of that each European country has its own health data regulations
but sharing inside of Europe is probably easier to achieve than
sharing on intercontinental level with the US, Australia or New
Zealand.</p>
      </sec>
    </sec>
    <sec id="sec-4">
      <title>V. CONCLUSIONS</title>
      <p>Through the expansion of the REMEDI platform to include
new communities, REMEDI was able to improve onboarding
new hospitals, work with national organizations to provide
relevant applications for their members and develop new
outreach campaigns with the help of AAMI and AARC.
REMEDI’s impact on patient safety in the United States has
been noted in several journal articles, and the collaborative was
recognized as impacting a significant clinical care problem when
it received the AAMI Foundation &amp; Institute of Technology in
Health Care’s Clinical Solution Award and the Institute for Safe
Medication Practices (ISMP) Cheers Award.</p>
      <p>We plan to continue expanding the alert applications for
the infusion pump community, discuss other medical device
technologies including bar code medication administration in
the next upcoming conference, and continue to onboard new
hospitals. We also plan to continue collaborating with medical
professionals internationally and share our lessons learned
with new science gateways. The most prominent takeaway
from this gateway expansion was to prioritize requirements
based on the community’s needs. Moving forward, our vision
is to continue improving the platform and adding in more
machines and continue working towards patient safety in
hospitals. In the future, we would like to integrate more
medical professionals into this community and open doors for
the public to also become involved. While there is always
more work, one can do, understanding the priority from a
community will keep work relevant and keep the gateway
thriving.</p>
    </sec>
    <sec id="sec-5">
      <title>ACKNOWLEDGMENT</title>
      <p>This work was supported by the REMEDI Project and by
HUBzero, both Purdue University affiliated projects.</p>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          [1]
          <string-name>
            <given-names>U.S. Food &amp; Drug</given-names>
            <surname>Administration</surname>
          </string-name>
          . (
          <year>2016</year>
          ). Medical Devices - Infusion
          <string-name>
            <surname>Pumps</surname>
          </string-name>
          .
          <article-title>Division of Industry and Consumer Education</article-title>
          .
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          [2]
          <string-name>
            <surname>Fox</surname>
            ,
            <given-names>B. I.</given-names>
          </string-name>
          , et al. (
          <year>2015</year>
          ).
          <article-title>“ASHP National Survey on Informatics: Assessment of the Adoption</article-title>
          and
          <article-title>Use of Pharmacy Informatics in U</article-title>
          .S. Hospitals--
          <year>2013</year>
          .
          <article-title>” American Journal of Health-System Pharmacy</article-title>
          , vol.
          <volume>72</volume>
          , no.
          <issue>8</issue>
          , pp.
          <fpage>636</fpage>
          -
          <lpage>655</lpage>
          ., doi:10.2146/ajhp140274.
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          [3]
          <string-name>
            <given-names>Purdue</given-names>
            <surname>Discovery Park</surname>
          </string-name>
          . “Regenstrief Center for Healthcare Engineering,”
          <year>2018</year>
          .
        </mixed-citation>
      </ref>
      <ref id="ref4">
        <mixed-citation>
          [4]
          <string-name>
            <given-names>BainBridge</given-names>
            <surname>Health.</surname>
          </string-name>
          (
          <year>2018</year>
          ). https://www.bainbridgehealth.com/.
        </mixed-citation>
      </ref>
      <ref id="ref5">
        <mixed-citation>
          [5]
          <string-name>
            <surname>Zink</surname>
            ,
            <given-names>R.</given-names>
          </string-name>
          (
          <year>2017</year>
          ).
          <article-title>“2017 REMEDI Pump Survey</article-title>
          .”
        </mixed-citation>
      </ref>
      <ref id="ref6">
        <mixed-citation>
          [6]
          <string-name>
            <given-names>M.S.</given-names>
            <surname>Lundstrom</surname>
          </string-name>
          and
          <string-name>
            <surname>G. Klimeck,</surname>
          </string-name>
          (
          <year>2006</year>
          ). “
          <string-name>
            <surname>The</surname>
            <given-names>NCN</given-names>
          </string-name>
          : [7]
          <string-name>
            <surname>Science</surname>
          </string-name>
          , Simulation, and Cyber Services,”
          <source>2006 IEEE Conference on Emerging Technologies - Nanoelectronics</source>
          ,
          <fpage>10</fpage>
          -
          <lpage>13</lpage>
          Jan. 2006 Page(s):
          <fpage>496</fpage>
          -
          <lpage>500</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref7">
        <mixed-citation>
          <string-name>
            <surname>Klimeck</surname>
          </string-name>
          , “
          <article-title>Hub- based simulation and graphics hardware accelerated visualization for nanotechnology applications</article-title>
          ,
          <source>” IEEE Trans Vis Comput Graph</source>
          ,
          <volume>12</volume>
          (
          <issue>5</issue>
          ):
          <fpage>1061</fpage>
          -
          <lpage>8</lpage>
          , Sept-Oct
          <year>2006</year>
          .
        </mixed-citation>
      </ref>
      <ref id="ref8">
        <mixed-citation>
          <string-name>
            <surname>HUBzero.</surname>
          </string-name>
          (
          <year>2018</year>
          ). http://hubzero.org/.
        </mixed-citation>
      </ref>
      <ref id="ref9">
        <mixed-citation>
          <string-name>
            <surname>Manataki</surname>
            ,
            <given-names>P.</given-names>
          </string-name>
          <string-name>
            <surname>Papapanagiotou</surname>
            ,
            <given-names>M.</given-names>
          </string-name>
          <string-name>
            <surname>Parsons</surname>
          </string-name>
          ,
          <article-title>Healthcare data safe havens: towards a logical architecture and experiment automation</article-title>
          ,
          <source>IET J</source>
          . (
          <year>2016</year>
          ).
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>