Transcript for 2008 VeHU Session #182
VistA Imaging Coordinators: I Have to Do What?
Brent: Good morning everyone, welcome to VeHU 2008 Session 182 VistA Imaging
Coordinators' "I Have to Do What?" This is a very apropos title and I'll tell you why in
just a minute. My name is Brent Fugett and I am assistant facility chief information
officer at the VA in Huntington, WV. I have a history since 2000 of working with VistA
Imaging and at this point in time I oversee VistA Imaging as one of several projects from
oversight capacity. Co-presenting with me is Michele Krajewski from the Philadelphia
VA; she'll be introducing herself later on in the session. Well I'd like to welcome you
guys and welcome to a very exciting project; I can safely say that imaging is the most
rewarding project I've ever been involved in and I say that without hesitation. My
background is that of IT, information technology. But before I get in to all of that, I
should disclaim that I am an employee of the Department of Veteran Affairs so any
mention of any third party products during this presentation is purely for referential
purposes and should not be considered an endorsement of that product.
The reason I bring up my background of IT is that VistA Imaging, one of the exciting
things about VistA Imaging is it brings a lot of disciplines to the table. I like to say that
it's a jewel of many facets. There is the biomedical component, there is the IT
component, there's the administrative component, the clinical component, the project
management component; it's a vast entity that touches a lot of aspects of a medical center.
So before I go any further, how many IT people do I have in the room? Wow, cool
probably half. Any biomed people? One (Laughter) that's the first time that's happened;
thank you Larry for volunteering that. How many clinicians do we have or - a couple?
Any clinical coordinators? Okay probably six or seven, okay great. Administrators or
any - okay one administrator, couple of administrators. So the reason I want to bring that
up is because I want to bring the position - this type of a discipline, this project to your
doorstep and hopefully approach it to you in such a way that you can see it from your
perspective. The session outline involves some terms I want to get out of the way early
so that when I throw them around you're a little bit familiar with them. The various
VeHU 182.doc 1
Transcript for 2008 VeHU Session #182
configurations that VistA Imaging can take; some basics about DICOM and I'll explain
that term in just a little bit, that's pretty fundamental to what we do. And then hence the
DICOM gateways, how they function and what they're for and then also I'll talk about
image quality. And then at that point I'll turn it over to Michelle and Michelle will talk to
you about adding new devices, preventative maintenance contingency plans, the
implementation work group, why that's important, getting support, how to become a test
site and then we'll both jump up on our soap boxes and preach a little bit about customer
support which is very important in our discipline as with a lot of technical disciplines, but
I want to bring that to the forefront also.
Now what is VistA Imaging? Probably the best way to describe what VistA Imaging is
to describe its fundamental components; the first component is in its name, VistA and it's
an acronym that stands for Veterans Information System and Technology Architecture
and for any of you in the room that might not be a part of the VA, that is essentially our
hospital information system. It is the archive for all the text based information that we
keep on our patient records. It is text based so therefore it's sized to hold not as much
information as what medical imaging brings to the table; for that we need a little bit
higher capacity and that's where the imaging component comes in. Now imaging is a
parallel system that's linked to VistA and it's stored and the architecture is made so that it
can store lots and lots of information; actually terabytes of information. To give any of
you non IT folks a perspective, a terabyte - well it's just a bunch of data; if you go and
look at an entire DVD movie, a DVD movie will consume about five gigabytes and there
are over a thousand gigabytes in a terabyte; so all of that storage has to be used to archive
all that information. The information is managed and it's obviously arranged in such a
way that can be easily retrieved, what benefit would it be if you couldn't retrieve it. It
allows the images to be linked with various parts of the medical record; if you've ever
seen CPRS, which is the GUI portion or the front end part of VistA, its structured in such
a way, it's tabbed and it organizes the diagnoses and the imaging reports and notes and
things like that in such a way that you have a linkage from those structures to imaging so
that if there's an image associated with those structures, for example a note, you might
have a signature document or something to that effect, you can link that with that part of
VeHU 182.doc 2
Transcript for 2008 VeHU Session #182
the medical record. Now both systems are programmed in a language called MUMPS
and it’s run in a Cache database environment, more of that we'll touch on a little bit later.
Now the type of data that can be stored in VistA Imaging includes scanned documents,
rich text reports, EKG wave forms, patient photos, X-rays, which is typically the first big
thing that a site gets into when they implement VistA Imaging. CTs or CAT scans,
magnetic resonance images, MRIs, ultrasounds, dental images, ophthalmic and on and on
and on and on; it's just a very flexible system at storing and archiving and retrieving
patient images.
Now here's kind of the standard screen shot of the VistA Imaging component, the piece
that the clinicians actually see. You have up here, whoever this guy is, looks like a
president I guess, he's actually Demo Patient; we have a sagittal view of an MRI and I
have a pathology slide here, I have an EKG wave form in the background and you see
that it's associated with the text of VistA here and it brings those components together.
Also in CPRS, here's a little glimpse of CPRS for those of you who have never seen it;
here are the tabs along the bottom, here is some lab results I believe are some blood
results here and the images that can be associated with that data over on this side.
Now it's necessary to mention, and this will mean a lot more for the IT folks who are not
as familiar with medical devices, that VistA Imaging is classified 510K as an FDA
medical device. And what that basically means is you can't mess with it unless you're
authorized. In the IT world we're sort of used to just to swapping out components as we
see fit, just go out to CDW and find out what's the cheapest to buy and whatever seems to
work and you can just swap it out but that ain't happening, as we say in West Virginia.
With FDA classified devices there are certain realistic latitudes that can be performed; if
you're a biomedical person and you really know what you're doing you can extend
latitudes with that, but if you don't know what you're doing you shouldn't be swapping
components out of a medical device, that results in what's called an adulterated medical
VeHU 182.doc 3
Transcript for 2008 VeHU Session #182
device and it could impact patient care and you definitely don't want patients' care being
impacted just because you tried to save $200 on a monitor.
So some terms to know, the first one is PACS and we'll throw that around a little bit and
that stands for Picture Archival and Communications System and it's a system that's
generically designed to store and hold patient information. Now historically a PACS type
system is historically related to radiology, X-rays and such as that, but VistA Imaging can
be considered a form of a PACS but it's a more universal type of a structure. Now
DICOM as I said before is Digital Imaging and Communication in Medicine and we'll go
into a little bit more depth of that here in just a little bit. HL7 is the text communication
protocol called Health Level 7 and it's actually how we get text back and forth from
VistA because imaging needs to know what in the world is going on inside of VistA. A
protocol, and again anyone who's not familiar with that term in the computer world is just
- a protocol is anything that adheres to a standard; whenever two devices try to talk to
each other you have to use a protocol so that both devices understand what's going on.
HL7 is a protocol, it's a standard; DICOM is a standard of communication. Just like
speaking a language, we have to each understand the same language in order to
communicate. A modality, we'll throw that around; and I remember when I became an
imaging coordinator and signed on and started walking around this murky world of
imaging, modality was the first term that kind of got me between the eyes and it took me
- I was too embarrassed to say what's a modality? Because I was afraid people would
look at me like I was a geek or something, which I still am a geek but I don't want people
to know it. But modality is simply something that is, again, conforming to a standard; in
imaging it's simply an image capture processing device that conforms to the DICOM
standard. So anytime that you hear the term modality referred to, it's simply a medical
device that can allow processing, capturing of medical images in electronic way that can
be exported and imported; that's all they're talking about. CPRS is the Computerized
Patient Records System which we showed you earlier, it's the GUI front end or graphical
user interface for VistA. The clinical workstation is what we refer to just as the
workstation that clinicians use to use the software in order to get access to those images.
A diagnostic workstation, this is the first thing that usually kind of confronts IT people;
VeHU 182.doc 4
Transcript for 2008 VeHU Session #182
this is your first contact probably with a FDA class five medical device. It uses very
specialized monitors that allow radiologists to do soft copy interpretation and there's a lot
of regulation that's involved in that and you can assure that the radiologist is seeing the
images in a very clear and accurate way and so that they can render a proper
interpretation. We'll go into that just a little bit later too, the details on that. And also
we'll be talking a little bit about hard copy versus soft copy interpretation; soft copy is
when a radiologist interprets images from a computer electronically; hard copy is when a
radiologist use film and light boxes, just like they have for years and years and years.
Now to kind of form up the overall view, I like to take everything from a bird's eye view
and kind of go down into the details so that we kind of have our bearings a little bit.
When I got to my hotel, I wasn't sure exactly how far the convention center was so I have
to bring up Google Earth and rotate it around so that my laptop is looking out my window
so I can see the Carnival Cruise ship over here, okay that's the waterway and look oh, the
convention center's right across the street, well that's no big deal. So I kind of overshot
that one a little bit, but I like to get my bearings; I like to have a top down view of things.
The several levels of implementation with VistA Imaging, the basic configuration is
actually the required configuration, it's fundamentally - VistA Imaging is required by the
VA to be used as the fundamental archive for all patient images and I'll go into that a
little bit more in detail, but it's basically because all of these medical devices and all these
modalities store, if you buy them from a vendor, they're stored in a proprietary format a
lot of times and the VA wants those images to be portable, to be seen across the VA.
And so the best way to assure that is to make all those images land in VistA Imaging that
way it's assured that everybody can see those images across the country. DICOM
configuration takes the basic plus DICOM gateways to handle DICOM modality and I
actually misspoke just a little bit; the required configuration actually includes No. 2 there.
If you have DICOM devices, they need to be secondarily stored also to VistA Imaging.
VistARad takes the DICOM configuration and adds VistARad, which is the VA's custom
in-house written version of a diagnostic interpretation viewer for radiologists; we'll talk
about that just a little bit later. Image routing takes VistARad and uses technology to
allow performance over slow wide area network links. And then we add the DICOM
VeHU 182.doc 5
Transcript for 2008 VeHU Session #182
configuration with a commercial PACS which is actually the alternative to VistARad;
they can be mutually exclusive, they can be blended or hybrid; we'll talk about that in just
a little bit also.
Now let's first look at the basic configuration. As I said, per VA requirement, each
facility has to have VistA Imaging installed, which everybody does, that's been that way
for a number of years. The minimum configuration is that you have RAID storage,
which is an array of magnetic disks to store your images; they're very fast, the capacity of
RAID is increasing and the price is getting cheaper. So I remember when we started we
only had about 300 gigs of storage which amounted to about 3 months worth of images.
So we didn't have a whole lot of images online and quickly available at the time. Now
we have tons and tons and I think it goes back years and years and years. For the
overflow back in those days, and actually still, we use something called an optical
jukebox, which is a big jukebox full of magnetic platters, magneto optical platters they
call them or UDO platters, you'll see that term used. And it's the long term storage that
holds them 75 years per requirements. I don't know what a computer's gonna look like in
75 years, I don't know that I want to know what a computer looks like. You'll probably
have these little 3-D goggles and can go like this and grope in mid space and manipulate
your images. The background processor is a component that actually processes the
images to and from the jukebox as well as some other activities. The clinical display, as I
mentioned earlier, is the software component that runs on a clinical workstation that
presents you with the images and allows you to work with those images. And Clinical
Capture is actually a component in that same software that allows the user, if they have a
scanner or some other device that can produce images, you can actually pick if you have
the right authorization within the system, you can capture an image and attach them to
notes, attach them to various structures.
What you have is a - you start out with the clinician who has a patient record that they
want - they have the text of the patient record but you know what? They want to see that
consult; they'd really like to see that signature on that surgical record that says they're
authorizing the consult to happen. What happens is you add this component over here,
VeHU 182.doc 6
Transcript for 2008 VeHU Session #182
the clinical workstation with capture and the person who receives the signed consent will
scan that in and that will interact with VistA, it will tell VistA, hey I'm scanning this
image, and the image portion will land on the magnetic storage component, it will
simultaneously get processed by the background processor, land on the optical archive
for 75 or 100,000 years and then the Clinical workstation will say hey, here's the address
here and the address here of where those images live; VistA this is where they're at in
case you need to get at them, and this happy clinical user up here can pull both of those
pieces of information up at the same time. Now we're gonna begin to delve a little bit into
the nuts and bolts of what makes imaging work the way the pipes lay out with each other.
Configuration two takes that basic configuration and adds DICOM processing. It allows
text information or work list, we're gonna talk about that in a minute, to transfer to an
image storage from DICOM modalities. The applications include computed radiography
or CR or radiological images like chest films and head shots and things like that, dental
image capture, ophthalmic image capture, transferring DICOM images to and from other
modalities such as DICOM printers, and also encapsulated PDFs; we see those
increasingly coming out. Medical devices of all shapes and sizes, I mean patient vital
systems that will produce really nice printed formatted reports in a PDF format they can
be exported to imaging and retrieved on demand.
Now before we talk about the DICOM configuration I want to get your head into DICOM
just a little bit. Now this is a weird world, I mean I know several people who are really
heavy into DICOM and they're weird people too. No I'm just kidding, I'm just kidding.
But they take things - the DICOM terminology takes things that are probably pretty easily
understood but they attach a weird term to it. So it makes it just even more difficult to
kind of get your head in to it. But for example, service class user, is a term or SCU, is
simply a device taking a client role device that sends print jobs for example. Or a more
down to earth example is you walk up to a ticket counter you want to go see a movie and
you hand your money through the ticket window, you're a user of the service; the movie
theater's providing a service, you're using that service, you are a SCU, you're handing
your money through. Well, a service class provider is someone that provides a service.
VeHU 182.doc 7
Transcript for 2008 VeHU Session #182
For example, someone that takes a host role like a DICOM print server or the girl that's
on the other end of the glass that gives you your movie ticket to buy; they're a service
class provider. That's an easy way to remember SCU and SCP. And the only reason that
you'll kind of want to know those is you'll hear those terms thrown around quite a bit,
especially when you're talking to support personnel or other people who are more
involved in DICOM, you'll hear those terms thrown around, it's good for you to know
that. Now DICOM images are in a composite format, which means the best - the way
that they've chosen to communicate this information from DICOM devices or modalities
is that files get transmitted over the network in the form of one file that has all the
information about that file along with the picture. The composite format means that
imaged data for the actual picture is kind of at the bottom of the file, literally taking up
the bulk of the size of the file and a header area, which is at the beginning of the file and
it is used for holding the text associated with that image; such as the patient's name, the
exam information, attributes about the file, how many pixels high and wide and what the
bit depth is. So that the device that receives the DICOM image has all that information
and it can properly reproduce the image. If it didn't have that information, the computer
software would have to sort of make certain assumptions about how the image was
originally captured and that doesn't bode well for accurately reproducing a medical
image; accuracy is everything in medicine period. It's even more so important in medical
imaging.
Again you have your user here in the center of the screen and at the bottom you still have
the VistA system, you have your magnetic storage; a little bit about magnetic storage, it's
very fast, typically not as high capacity so you use it for very quick access to relatively
short term storage items. Optical archive over here, which holds lots and lots of data but
it takes longer to get at those. When we're talking about currently - I know in Huntington
we've been in medical imaging now for going on eight years and so far all of our images
are still on the magnetic just because it's so easy to keep that much storage available
you're like why not? You know increasingly providers will need information going back
five years maybe; once you get back to the seven year mark you get to a point of
diminishing returns. Even the film library will purge films after about six or seven years,
VeHU 182.doc 8
Transcript for 2008 VeHU Session #182
depending on the site, depending on regulatory requirements; they get sent off to off-site
storage. The only time you'd need anything longer than that if a tort claim came,
attorneys show up with briefcases and they want a film because they're pursing a case.
The optical archive VistA Imaging will hold them for literally decades. So we have a
means to store that information long term. The background processor that manages the
storage to and from and now we add two components there, the DICOM image gateway
and the DICOM text gateway. What happens is this DICOM text gateway on the left
hand side will use the HL7 protocol to allow VistA to say hey here's a new exam entry;
someone has just ordered a chest film up in radiology. That information will get
communicated real time over to the text gateway and the text gateway will store it in
what's called a global or a storage unit on the machine itself and it'll just sit there waiting
for someone to request it. Well up here in radiology you have a DICOM modality, you
have the work list component - in our facility it's actually a little device that's mounted on
the wall where the clinician will walk in with the patient and they go up and they walk up
to the list and oh the patient's not on the list, they hit the refresh button and that work list
component will call down to the text gateway and say hey, give me the most current list
of patients, and it shoots up there and there's Mr. Smith or Patient 1, Patient 2 as we have
it up here and they say oh there's the patient, they touch the screen and they associate a
stack of plates or if they're doing it manually or whatever the modality, how it processes
it, with that patient. They walk into the room, they capture the X-ray in this stage right
here, then the tech goes over here and QA's the image; someone's usually at a QA station
of some type and they look to make sure that the image looks correct, it's not blurred or
off center or whatever; they put the correct markers on it as prescribed by the radiology
department. Then they say okay I'm gonna store this down to VistA Imaging. What
happens is that image gets sent down to the DICOM image gateway as a DICOM object.
The DICOM image gateway does several things; first of all, it contacts VistA to say hey
I've got images for Patient 1, here's the Social, 123-45-6 whatever and here's the
accession number, which is the case number for radiology, do all three of those things
match up with an active case in VistA? If they do match up, it says okay that's good to
go; I'm gonna process the image. If it doesn't match up, it gets stuck there and it stays
there until you release it and that's kind of one of the little catch all kind of things. As an
VeHU 182.doc 9
Transcript for 2008 VeHU Session #182
Image coordinator, people call in from radiology or other places that provide DICOM
images and they say hey my images aren't going anywhere, that's the first place you look,
see if they got stuck. Sometimes if the tech doesn't see the patient on the work list over
here, they might try to cheat and manually type in the stuff and if they miss it by one
digit, it won't go through, it's a safety feature. Also what will happen is let's say a tech
sends through an image and it successfully stores, this is a second scenario, but they
realize after they stored it that they messed it up; well they try to correct it and try to send
it through again, depending on how you have yours configured, we have ours configured
this way that it will not allow duplicate unique IDs of images. Whenever it creates that
DICOM image, there's a unique ID associated with that and when that image comes
down and it sees a duplicate, it ignores the new image; it just discards it. So if you have
someone that says oh my image is wrong and I can't get it to update, that's usually kind of
where you're going. When the DICOM image gateway sees that all that information
matches up, it first of all stores the image portion of the data on the magnetic, and we'll
talk about a couple scenarios that we have; there are other variations of this. But
essentially it stores the image data. It stores the text portion, like from the DICOM
header in a text file on this RAID storage in a distinct format. The background processor
takes a copy of that image and stores it over here onto the optical archive for long term
storage and the DICOM image gateway then takes all those pieces of information of
where it put those things and stores it on VistA so that this clinician can pull the images
back up again. That is the bird's eye view of how VistA Imaging works in the most
fundamental way involving DICOM.
Now, configuration three takes that same DICOM configuration but we add VistARad.
Now VistARad diagnostic workstations allow radiologists to interpret radiology images
on the computer, or soft copy interpretation. VistARad is simply imaging software with
specialized tools helpful to a radiologist and we have a number of VistARad sessions; if
you're just curious about what VistARad looks like, feel free to step in to one of those
sessions and I'm sure John and the team would be happy to show you the product and
show it to you in action. It runs on selected types of workstations, equipped with
diagnostic monitors, I'll talk about that in a minute. It communicates with VistA to
VeHU 182.doc 10
Transcript for 2008 VeHU Session #182
update exam status when interpreted, I'll talk about that. And we are - Patch 65
introduced many commercial features to VistARad; we're currently on Patch 76 with a
couple more patches coming up to bring some more cool features with it. One bonus
about the software, it's basically free. It's written by the VA and it's available to your
facility if you choose to use it or try it out.
This introduces the scenario of the VistARad workstation and we have up here a
VistARad unread list; that's kind of - there are a number of lists of patients that the
radiologist can use. Pivotal to all of this is VistARad uses something in the radiology
package called the read status or the exam status and it involves several stages. And it
basically follows the image from the patient being registered, the patient having the
image taken, it being interpreted by the radiologist, it being confirmed by the radiologist
and finalized finally to be an official part of the patient's record. The first status is
waiting for exam, when the patient walks up to the counter and checks in and says hey
I'm here to get a chest film, they check him in and their status for that particular exam is
I'm waiting for my exam; I'm waiting for my image to be captured. The next status is the
patient has - the radiology tech has captured the image, stored it on VistA through that
process I just showed you a little bit ago and when it's in the examined status, that's the
trigger that VistARad uses to say hey, they're on the unread list. It's an image that has
been captured but it's not been looked at by a radiologist. When the radiologist eases up
to the workstation in their dark room, they take a look at their image, they make an
interpretation, they dictate it on a Dictaphone or a voice recognition software; dictating
the report, there's evidence of this; there's evidence of that. By the way, I'm always
interested in - one of the things I thought was cool was how radiologists dictate the
images; they never say there's such and such there, they always say there's evidence of, or
I'm suspect of, they never knock theirselves into a corner, it's always just sort of I think I
see something, but you'll have to check it out and make sure. Once the radiologist has
cleared all that and dictated the report, they mark it as interpreted and the status gets
changed to the interpreted status. When the transcriptionist or when the voice recognition
software, one of the two, converts that into text, the report, or the exam goes in to the
transcribed status. Well then someone in radiology has to basically read that report,
VeHU 182.doc 11
Transcript for 2008 VeHU Session #182
ensure that it's correct and poignant to the exam, then it goes to the complete status; at
that point the exam is complete.
Now, if you have VistARad workstations over a wide area network link, in our facility
we have two radiologists that actually have VistARad workstations in their homes; we
have T1 lines that are extended out to their homes, which compared to the facility, are
very slow. They're like probably, I don't know, like probably five percent or less the
capacity of a full blown network connection in the facility. So images don't fly across
those lines as quickly as they would in the medical center. To kind of compensate for
that, VistARad uses something called image routing which allows the images to be
prepushed if you want to, over to the VistARad workstation and stored until the
radiologist is ready for them and then it just pulls them up off of the local hard drive.
Much quicker than having to drag them through a straw when it's used to a fire hose.
The text data stream still comes over the T1 line because text is very slender and it still
comes through the lines very easily. The image data stream gets processed the way I just
stated, where it gets preprocessed. Now the images can be routed either on demand or
they can be auto routed; routing on demand means that the radiologist goes ahead and pre
chooses a whole bunch of studies and says route me these images, goes off and gets a cup
of coffee, comes back and then the first study is there, they go ahead and start interpreting
and while they're interpreting and deciding on the case, the remaining images are coming
through automatically. If they can preselect a group or a list of images to read for
themselves, they can basically read with very little interruption, if any interruption
because of bandwidth or performance.
Now DICOM with a commercial PACS is if you have a commercial PACS environment,
which is a vendor produced soft copy interpretation system and capture architecture. It is
a self contained system that includes its own patient management system separate from
VistA; some of the PACS out there are trying to integrate more tightly with VistA so
there's a little more uniformity, but most of them as far as I can tell are still pretty distinct
from VistA. Includes a separate patient management work list, the image acquisition and
QA components are separated, storage and diagnostic interpretation are separate. Now
VeHU 182.doc 12
Transcript for 2008 VeHU Session #182
VistA Imaging sees PACS pretty much as a black box as like a modality. You still have
the work list component, the processing, all those components. It basically provides
work list to it and it gets images from it and that's all VistA Imaging really sees at that
point. It is required at this point if you do have a commercial PACS to store those
images, as I mentioned, in VistA Imaging for portability.
A similar situation in a commercial PACS, this entire yellow box is basically the PACS
environment. The acronyms are very similar, you have image storage, you have the
equivalent of VistA which is a RIS or Radiology Information System, you have the
PACS Diagnostic workstation which is similar to VistARad and you have the work list
capture and QA components and it's all encapsulated. But VistA Imaging living outside
of this yellow box sees it as one big modality.
There are some pros and cons to VistA Imaging versus VistARad - or excuse me, PACS
versus VistARad and it's almost like a feverish kind of thing among certain people; it's
almost like Chevys and Fords. I mean people embrace or love one or the other and they
have shouting matches across the ball field, you know if you're not, you know. But there
are pros and cons to both; VistARad is custom written and historically has been kind of a
clunky package and four or five years ago it was very basic, it was completely
outstripped by commercial PACS environments out there as far as the capabilities and the
bells and whistles and things. But with Patch 65 and subsequent patches, it's rapidly
catching up. So I would encourage you that if you've been a commercial PACS site, to at
least take a look at it because the cost factor is pretty compelling. You do have to depend
on in-house staff as opposed to commercial support staff so much, so if you have a good
strong IT department biomed component, I would encourage you to at least take a good
strong look at it. Now the pros are that it's self contained; it's not dependent on VistA or
outside systems to function so it will function in a stand alone mode, especially if its on
its own switch and all that kind of stuff; if VistA's down, the hospital network's down, it
will continue to function. And they typically offer a very robust feature set, at a price
point of course. But with their own components comes all this unique communication
information that goes to the PACS and it can provide some unique functionality. Now
VeHU 182.doc 13
Transcript for 2008 VeHU Session #182
the cons, just like pros it's self contained; because it's self contained, it's another copy of
the data that has to keep synchronized. So your quality assurance or QA really has to be
strong; you don't want to assume that because it's in the PACS correctly that - you don't
want to blindly assume that it's correct in VistA Imaging. Seattle might be pulling your
images and seeing incorrect or bogus image and make some kind of an assumption of the
patient based on that image and that would just be a bad thing. In general, I will take a
second to mention that QA is everything in VistA Imaging or actually any discipline in
the medical center, but we've always encouraged, and I have to lift up our Radiology
Department, if you're out there listening to my voice, kudos guys because their QA is
second to none; they literally, even in the most pressured of situations, the processes, they
capture, they send, they pull it up in VistA Imaging and look at it. They will tell us
before there's a problem and so kudos to those guys; that's just been ingrained in their
culture and they're shocked to hear that anybody does anything any different. It's a very
good group of folks in that department.
The cons are again in the commercial PACS, they are very expensive, come with support
contracts and such as that and more times than not the data is stored in a proprietary
format. So ask lots of questions, if you're considering a PACS environment, actually for
that matter when you're considering any vendor system, please call your local Geek
Squad; get them at the table and give them full permission to drill the daylights out of the
sales and technical reps that come, please. Let them go at them because it's better to find
out ahead of time all the little gotchas than after you've bought your system; once they've
got your money they're not nearly as inclined to make you happy. And Michelle will be
talking a little bit more about Directive 6500 and the procurement process that's been
instituted recently.
Now DICOM text gateway up close runs inner systems Cache on Windows 2000 or
2003, strongly trying to get everybody to 2003 operating system. It is a work list service
class provider modality; remember that term, since it responds to query requests from
clients. And it queries HL7 messages from VistA, we went over that earlier, and converts
the data into DICOM messages for work list.
VeHU 182.doc 14
Transcript for 2008 VeHU Session #182
Now the DICOM image gateway, and before I go too much into the image gateway,
DICOM text gateway and DICOM image gateway are not unique necessarily unique
systems. It is one piece of software that you basically turn the functionality on and off
per system. You can have one box that provides work list and image capture processing
at the same time but if you're doing it at any capacity, you really want to keep those
distinct and separated so that you don't bottleneck your resources. And also these
systems can run, besides running on server, they can run actually on a workstation. So if
you get in a crunch you can actually install this thing on pretty much any PC in your
facility and it should run fine, as long as you've got proper capacity. The image gateway
is a storage service class provider modality, among lots of other functionality and as we
discussed earlier, it does the matching with the name, Social Security Number, and
accession number. And if there's a match, it queries VistA for an internal entry number,
gives it a big old long number to uniquely associate to that image set, splits the DICOM
file up appropriately, if appropriate and stores and copies the file to the RAID and the
jukebox.
Now I need to kind of, hope this doesn't get too confusing if you're not familiar with, and
it might sound a little murky, so overlook this slide if you don't have a little bit of a visual
with regard to what I'm gonna talk to. But historically when these DICOM files would
come from these modalities, as of 10-15 years ago when they started the project, DICOM
was not quite the standard that it is today; they weren't quite sure that when they were
getting these DICOM files that if they stored them that way that there would be a good
way of pulling them back off again. So they made the decision to store the files, back in
those days, in what's called a targa format, which was very open and is still a very open
format. The targa format is pretty simply if you've got a bit in the image, you got a bit in
the file; it's not compressed at all. It's a very one to one relationship, as a result the
images are pretty inefficient as far as storage and also the process of taking the DICOM
file and breaking it up into components is problematic at best. When you have to try to
put these files back together again as DICOM files later, say to send them to another
DICOM service class provider such as a DICOM printer, lots of issues have ensued
VeHU 182.doc 15
Transcript for 2008 VeHU Session #182
trying to assemble these things back together again and they pop out the printer and
they're backwards, color and monochrome is one way or the other way. So there's been
an effort in the last 5-6 years, and its still ongoing, to try to capture the images in their
native DICOM format and store the DICOM file because it is a standard now and it can
be stood on. But you will still see some modalities that they don't have the process in
place yet. What happens for those images is that it takes - it splits that DICOM file up
into several pieces; first off you end up with - you can end up with what's called a "BIG
file" and that's just simply, remember I mentioned the DICOM file has a header piece
which the text information and then the image data at the bottom? It's basically the
image data taken out and stuck in a file. If it's coming from a modality that's very large
in geometry, such as a dedicated chest room or something of that effect, you'll end up
with what's called a BIG file. If it's not coming from a large format modality, such as an
ultrasound that have the relatively small like 512x512 images, you don't have a big file;
they go to what's called the TGA file. Now you almost always have a TGA file, even
when they have the BIG file they have the TGA file as like a miniaturized rendition of if
so that if you pull up - most of the time the clinicians just want to see a smattering of the
image, they don't want to see the whole chapter they just want a couple of verses. So
they can pull up the TGA file for speed, especially across slow WAN lines, that's very
efficient. You end up with a text file which is the text, or DICOM header component and
then just a plain old text file and if your people who are - your imaging people or you
become an imaging person with direct access to the storage, you can carefully access that
text file to see the contents if you need to troubleshoot something. However, please be
careful 'cuz when you're in the place where that stuff gets stored, you can accidentally
save it, drag it to the wrong folder and it's a very open and raw environment in there and
so consequently, there are procedures in place that very few people have direct access to
that piece of the system.
So you also end up with, regardless of format, whether it's DICOM or not, you end up
with an ABS file or abstract and that's basically just a little teeny thumbnail. So that
when you pull up these thumbnails on the abstract screen, it doesn't have to drag the
entire image set over for performance.
VeHU 182.doc 16
Transcript for 2008 VeHU Session #182
Image quality basics. Now this is especially poignant for the geek squad out there, the
fellow geeks, all you IT people; we historically look at a monitor or an image producing
device and say hey if I can see an image and if I can read it, heck it's good enough for me,
but that's not the case in medical imaging. What you don't see is what hurts people,
literally. What is necessary in medical imaging is good equipment, equipment is working
properly and appropriate equipment; all three of those together.
Now appropriate resolution, in other words the image matrix, you're talking about what
you see on that screen is actually an image matrix; you have bits going across of various
colors, bits coming down of various colors. That might be up there like a 1280 resolution
type of display; you have 1280 pixels going across. If you have a monitor - and this is
primarily talking about in the radiologist's room where they're in the dark room and
they're reading this for critical interpretation, the ideal thing is, at least my consideration
is, that you have what's called a one to one pixel ratio; that you can initially present the
image up there without having to squeeze the image to fit on the screen. As a result, in
our facility we spend a little extra money and get the very higher capacity, the higher
resolution reading monitors, they call them the five mega pixel, talk about that here in
just a second, so that these larger chest films will fit on there without having to be
reduced. Will the radiologist miss it if it drops a few pixels? Eh probably not, but we
like the idea of showing every pixel on the screen and then the radiologist can choose to
reduce or expand at their discretion. The image type requirements; depending on what
type of image you're trying to display, ultrasounds are smaller matrix, you can get by
with pretty much anything to display an ultrasound and it will display properly as far as
image matrix goes because they're so small. The larger displays are referred to in terms
of mega pixels; the five mega pixel that I mentioned is a 2500x2000 display. They have
three mega pixel monitors, they used to have four mega pixels but I don't think they
really have those anymore. IBM has a nine mega pixel color monitor which is really
nice; people have been going to those for mammo's which tend to be very high resolution
images. Also you want to look, when you're choosing equipment, you want to look at the
intended use; is it for clinical review or is it for diagnostic review? The key is if a
radiologist is gonna use it to do a diagnostic impression, it's got to be a diagnostic
monitor and you can't just go to the CDW and buy those things, you have to buy them
VeHU 182.doc 17
Transcript for 2008 VeHU Session #182
actually specifically I think because of OI&T we're kind of restricted where we can buy
things, there's just certain places we are supposed to buy those. But you're not gonna put
you a Sanyo monitor in a diagnostic reading room without being in violation; it's not
worth the money savings. Now besides resolution, resolution is something as an IT
person I didn't have too much getting a grip on, yeah you need lots of resolution. The
part I kind of had to put myself into a little bit was grayscale; now grayscale is where you
get tripped up. Zero to 100 in 1/60th of a second, that's my - basically my way of saying
that grayscale is measured in terms of percentages; zero percent grayscale is absolute
black, there is no luminance on that particular pixel whatsoever. One hundred percent
grayscale is complete white and the key - what affects perceived grayscale, and let's face
it, socially they say perception is everything; it's everything with regard to medical
images. It's what the radiologist's brain can register and see and perceive and recognize
is everything. What affects that is hardware limitations. Again, is the monitor
appropriate for medical interpretation? Also the set up, even the ergonomics; is it
properly calibrated so that you're seeing the entire range of grayscale, and I'll discuss that
in more detail in the next slide. Also the environment and biomed is very strong at our
facility about this, they even go so far as to say the rooms need to be painted with a
certain color of paint and the monitors need to be oriented so that there's no direct light
sources behind them; you can't stand any glare on those things, they need to read these in
the dark or in a very diminished light environment. If you have any ambient light
whatsoever, it has to be heavily controlled because, I mean let's face it, pretend that your
grandfather who's in the hospital has some breathing issues, there's a nodule in the lung
that's very faint and it's only just a slight shadow, you want that radiologist to see that
shadow and having a glint of light in the room can mean the difference between seeing
that or not seeing that. In fact, that's one of my mantras I guess, as I approach medical
imaging or actually my work in the medical center, I have to make myself think is
everything that I do one of my relatives, or somebody's relative, is probably on the other
end of my action; I really have to be conscientious about what I'm doing and keep
everything above board.
VeHU 182.doc 18
Transcript for 2008 VeHU Session #182
Now image quality and grayscale calibration, producing a good curve; we use an old item
just for roughing things out called a SMPTE pattern.
Now this is a SMPTE pattern and I didn't check this out before - this is actually a pretty
good illustration. If this projector had the brightness and contrast adjusted correctly, you
see on this little piece right here; you see that little faint gray box? That's 95 percent laid
over top of 100 percent; you have to be able to distinguish that on the high end of the
grayscale. But can anybody see the box over here? No, that's because the brightness and
the contrast is not properly adjusted on this projector; that's a very simplistic way of
saying that this display is not properly calibrated. What if that gray nodule that I was
talking about in your grandfather's lung was that 95 or 96 percent level of gray, it'd be
invisible right? Completely missed it and the radiologist wouldn't even have a way of
even seeing it.
What we do at our facility, we have a little software package we bought called Image
Smiths I think, Image Smith, and again that's not an endorsement of the product, that's
just saying what we did, but we actually installed that; from the company you buy the
puck that's the calibration but the software's free. So we installed the software on all the
workstations, we have the same workstation, same monitor that goes out all over the
place; when we do the master build we calibrate that monitor, we reset it to factory
defaults, we calibrate the monitor, it puts the grayscale curve in the video card and then
we clone that all over the place. So we know that a clinician, all they have to do is reset
the monitor to defaults and they've probably got adjusted because they got their window
blinds wide open, they've got florescent lights on in the ceiling, which means it's all out
of whack as far as that goes. But if they're looking for something that they see in a
radiologist report and they don't see it in the image, it kind of messes us up a little bit -
messes them up, the first thing we say is okay reset to factory defaults your monitor and it
puts the curve back in place, and then they can see both ends of the grayscale spectrum.
That's been kind of a handy thing. The term SMPTE actually stands for Society of
Motion Picture and Television Engineers and it was actually used back in the old days to
calibrate black and white televisions and monitors. But we use it in the medical
community just to get those two pieces of information.
VeHU 182.doc 19
Transcript for 2008 VeHU Session #182
At a minimum you should be able to discern 5 percent and 95 percent grayscale at the
same time. And on some monitors you'll find, and video systems, you'll find this not
possible; it's just not possible and that's when software like the Image Smiths thing, the
calibration software comes into play.
There's a close up of that 95 and 5 percent.
Now if the video card supports look up tables then the card can be calibrated for the
attached monitor and this is called gamma correction, its part of DICOM part 14, the
DICOM standard. They must be recalibrated periodically, or if the monitor's changed. If
this is in the clinical environment it's not that big a deal unless you have the clinician
that's very intense on seeing an accurate image, they might want to demand; you might
want to educate your users that if they demand that level of accuracy, you can maybe
provide them with a little bit more appropriate pieces of hardware. Calibration is a really
good idea for clinical workstations; it's an absolute requirement for diagnostic
workstations.
Michele: So it's helped to ensure that you get successful interfacing. So there's a DICOM
conformance statement that your IT department will need that comes from the vendor,
they have to work with Silver Springs to make sure they are meeting all the DICOM
standards to talk with our VistA Imaging. It has to be done before you buy the device,
you will need to make sure it's on the approved modality list before you go forward.
Stick to your guns because your medical center staff people, through all the different
services will try to get things passed because OI&T does certain things that are IT
budgeting, the medical centers can buy medical devices. Everything has to go into the IT
tracking software now so we do get a chance to catch a lot of stuff, but I can tell you from
my own facility, they bought three pieces of equipment two years ago and the company
told them oh they're on the approved modality list, after they were purchased and on site
and wanted me to interface, I said no it's not on the approved modality list. So still one of
the devices is still not interfaced yet because it just got approval a few months ago. So be
VeHU 182.doc 20
Transcript for 2008 VeHU Session #182
careful with that. There is now a new directive that helps with that, the 6550, do I have
that - yeah, required components, it's gonna ask for network specific information.
They're gonna look to see what kind of rights the users need to operate this system. In
the OI&T world, you know giving people administrative rights to devices is becoming an
issue, a big time issue. They're gonna ask for physical characteristics, so this form is
what, five pages long?
Michele: That has all kinds of data in it asking for all these things; these are all supposed
to be a pre-assessment, that 6550 is called "Pre-Procurement Assessment". Your biomed,
and I know our biomed is really strong right now with that, we're making sure we're
working together to get these documents filled and actually this year when they started
asking for medical equipment, having this has been a great deal of help. Partly because
people who would buy servers, they have to be in our computer room, that's another
issue, you know do you have room for these things, let alone interfacing them to imaging.
Contact Silver Springs they'll tell you if something, if you don't see it on the webpage that
it's an approved modality, Silver Springs is great for telling you where it's at in the
process. And yes it must be approved, stick to your guns; be forceful because they'll try
and get you to do stuff. If the device is not approved, tell them to get back and as I said,
don't back down. At the bottom we gave you the web link to the PDF with the approved
modalities; it's on the VistA Imaging website, just go there and pick it up and look
through the list. It's fairly easy to get through the PDF, I used the find feature on the
PDF, pop in my device name and it tried to find it on the list.
In addition to that, that's while you're getting it improved, after it comes in there's a
technical imaging acquisition technical data sheet, there's two different ones depending
on what you're doing. If you're doing a DICOM modality, there's one that's for them, it
goes to the vendor and the service line who's asking for it has a couple details that you'll
need to be able to interface the device. And then there's a clinical capture; these are for
scanners, anything importing so you gotta make sure that you cover yourself. You know
sometimes people say oh it's just a scanner, make sure it's approved and you have your
technical data sheets for those. And for each modality you're gonna submit it to Silver
VeHU 182.doc 21
Transcript for 2008 VeHU Session #182
Springs so that they have a listing of what you're using on your imaging system and as
part of the 510K compliance. Then once you've got this interface, you're gonna need
your subject matter expert in the service line who's gonna be looking at these images,
doing any interpretation from images to review the images and fill out an Image Quality
Certification form. Don't have them say, well you fill it out, no you need one of the
clinicians that's gonna interpret the images. Periodically I've had to check back,
particularly with our Plan Mecca device, periodically check back with the end user
because we've had some degradation that happened on the device so he actually filled out
a second one and he said it wasn't working and we ended up getting biomed and we had
to get it recalibrated with the vendor. So you might want to check back; the forms are
always up on the website, get the latest ones up there. I don't even download it anymore;
I just pick the latest one in case there's been any changes so that Silver Springs gets the
most important ones up.
The other thing I can give you a recommendation for, as I stepped in and I was brand
new, the person who previously had left almost a year prior to me, is I asked Silver
Springs to give me a list of what they had on hand for our facility, went through the list
and then I had to back track and get a couple items sent back to them, so you might want
to do that and see what they have. Preventive maintenance, this is not an official or
comprehensive list, these are just things from a pair of people who have been there, done
that. Each morning review the basic operations of your gateways; at my facility we use
Dameware, you could also use PC Anywhere. I do a quick check from my desktop in to
all my gateways, check make sure all the modalities are working. I check my
background processor and make sure things aren't piling up. Check that my backups
have run and there weren't any hardware problems. If you have a jukebox, check that
that's running properly. So you will have to go down to your devices, but you can do
some spot checks before you get started; sometimes they give you a big head's up that
things are going on. Review your system logs and events; see if there's anything unusual
that happened. You know sometimes things look like they're running smoothly but
there's something going on in the background. Look at error trap, the VistA error trap,
especially if you put a brand new patch in, monitor that. That can sometimes bottleneck
VeHU 182.doc 22
Transcript for 2008 VeHU Session #182
things up. Look at your backups to make sure that they ran properly; last week we had a
problem with the tape drive, it just wasn't recognizing the tape that we had just put in. So
lo and behold the next morning we didn't have a backup done. If you have the jukebox,
remove any completed jukebox copy media and move it to a secure location, save it for
that 75 years. Walk through your area; so as I said, I do a lot from my desktop first thing
in the morning but then I run down to the computer room later on, and check to make
sure I'm not having any problems with the RAID array, any problems with the jukebox.
Make sure things are there, and we have a little log book that we keep on top of the server
rack that we monitor and we put down any problems that we've had; if we had to take
down the system, reboot something for a reason so we can keep track, look for issues that
might be repeating itself or if we see trends; it sometimes can help when you're calling in
to the help desk, have you seen this before? Has this happened? Kind of those tracky
things.
Monthly check your backup operations, remove the media to off sites, it is recommended
and we probably can tell you from when Nashville had their incident, you want those
backups somewhere; it's been an issue. Purge your DICOM gateways, you know clean
them up, make sure things look good. And check your RAID free space; we are a facility
that two years ago we had three terabytes and we were quickly purging things and
keeping about two to three months on our local RAID. We now have 19 terabytes but I
still check and make sure, yeah I probably can keep a lot on there, but I want to make
sure that I'm not having issues again 'cuz sometimes you can use that data to get more
equipment too if you're seeing that too, you can get new hardware.
Plan for disasters; as I mentioned, Nashville had an issue where they had a flood in their
computer room and without the backups it can be, you know getting things back up to
where you need it with your incremental tapes, can take time. So make a point to have
your disaster guide easy to follow, that its in a place where everybody can get to it, in a
place because you may be on vacation, that somebody else can follow it and if you're
getting national help that they can help follow it. Label your key information on your
critical equipment. Label your local RAID, your arrays, learned that the hard way. But
be discrete, if it's in a public area you don't want to say here's our imaging server, crash it,
VeHU 182.doc 23
Transcript for 2008 VeHU Session #182
but label things. We have certain racks within our computer room that are specific to
imaging, I have them labeled. That came in handy when the engineers wanted to come in
and replace a circuit and they said, well we'd like to split this circuit, and it happened to
be the one that all the imaging servers were on and I said no you'll go split another circuit
so if you have things labeled people can get in touch with you and figure out what's going
on. I'm not the only one who has stuff in the computer room so it makes it hard.
Document your formula, again it's not official, but document everything and sometimes
you may have to consult with your ISO. Keep an accurate table of contents; with all the
systems with the ISOs, they want you to have your system documentation that includes
your security plan, your backup plans, all your plans. So keep an accurate table of
contents, make sure your emergency contacts are up to date. As I said, in March I
changed roles, I'm now supervising the imaging person so now we had to change our
documentation to now have the backup person who became primary's name as the
emergency contact first line and then I became second line. We're adding a second
imaging person starting next week so I'll have to update the software again to include
both names before it gets to me. Make sure you have your system overview; what's on
your racks? We're replacing three of our G1s with G5s this week and we're adding a
fourth gateway in addition to replacing the text gateway, two previous DICOM gateways
and we already had a third gateway, we're getting more modalities in and we're now
putting in a fourth gateway. So keep track of all that, update your diagrams as you add
more equipment. And that's key is when you call for HP help if you have a hardware
failure, what are you running? Are you running the G1, are you running a G5? They
don't like when I call and say I still have a Compaq G1 sitting on my rack, so they're glad
I'm updating this week. Keep everything there, keep your key information on each piece
of equipment; the imaging stand and overview section, a component section, what
functions, effects of what an outage can do if you have pieces, what your start up and
shut down procedure. Lot of times - we now have a regional data processing center for
our VistA system so we're not doing the monthly down times as we were before, but
before the VistA system manager had to take down the imaging system when he did the
monthly VistA down time; so I had to have procedures specifically for that person on
how to bring everything - take it down correctly and bring it back up. So we have those
VeHU 182.doc 24
Transcript for 2008 VeHU Session #182
documents and we actually institute a internal wiki at our facility, we're gonna post those
so even if the imaging person's not there and there's an emergency that happens,
somebody who's on call can take care of it. Look at your contingency plan, make sure
everybody knows where it's at and keep your notes.
Implementation. As you want to put in a new device, or even starting up VistA Imaging
as a whole, anything, is make sure your group of people that are taking care of each
device, it's a comprehensive stakeholder group. Make sure it's key individuals - in
Directive 6550 now it actually asks who all your stakeholders are so that you have them
all involved. There's nothing like being brought in to the last minute to a project and just
say well we're already this far through it, you're the stop gap now, so there's nothing like
being that. Look at all your services; right now we're looking at a cardiac PACS system
but they want to integrate it with the radiology PACS system too, so they want things to
go together so look at that. Make sure your imaging coordinator, you, are part of the
process. IRM, if you're not part of IRM, 'cuz in some places the imaging coordinator is
not the imaging system manager, so it could be two different people. Make sure biomed's
involved, that's a big key, these are all medical devices and they will be a really good ally
in helping getting things done. We have a tiger team that works with our radiology and it
includes a hardware person from our peripheral section, the imaging system manager, a
biomed person, the radiology ADPAC as we've been introducing the new commercial
PACS along side VistARad, is keeping everybody working and making sure things work
together. It's key for implementing these large projects that you have all the people
involved; it helps run very smooth and as I use Brent's thing, it keeps the devil out of the
details if you plan ahead. Every medical device, hopefully now, will be integrated with
VistA. It's key for an electronic record to have this, so if you can, find devices that do
have the imaging piece. We actually use that as a criteria when they were looking at an
anesthesia record keeping software, there was only one vendor at the time that had some
type of integration into imaging and we were able to convince them to choose that
because of that, so now we have the flow sheets from the anesthesia going right in to
VistA Imaging through a PDF function and nobody has to scan, nobody has to monitor.
VeHU 182.doc 25
Transcript for 2008 VeHU Session #182
There is a little bit of monitoring in the background to make sure things are flowing right,
but it's now there as one document.
Getting support. Everybody knew this is the best key, I can tell you. VistA Imaging,
being from VistA background, being a clinical application coordinator, VistA Imaging
support is probably the best one out there; you put a ticket in, somebody's on the phone
within ten minutes to you to figure it out. I can't speak more highly of it. They'll help
identify who you need to contact. You put the remedy ticket in, they'll figure out from
what you've rate, even calling you whether they need to get HP on the phone for your
hardware, whether they need to get Dameware in, actually they go in through PC
Anywhere. They'll hop in the machine, they'll walk you through stuff they'll say watch
and teach you along the way, they are really great with that. I've had no problem with
them even paging somebody on a weekend when something really crazy happens.
Remedy for all the VA packages, VistA Imaging's one of them, it tracks the statuses,
they'll update you. The HP Expertise Center number is here, they'll help with server
hardware issues. I actually put the remedy ticket in first and let the imaging support staff
help me get - 'cuz sometimes I know it's HP that's gonna end up with a hardware issue,
but it keeps everybody on track because I've sometime had to use both the national
support staff and HP to get things working again.
Get yourself on the listserv, which is our last item, but there's weekly imaging conference
calls every Thursday at 12:00 Eastern time. Forum, if you don't have Forum access, get
Forum access; there's a lot of information that goes back and forth on Forum. There's -
sometimes you can get, instead of sending it out on Outlook, send it out on Forum, it's a
nice string of messaging back and forth if you're having problems and we put the call in
number for the weekly call here. Monthly there's an HP imaging hardware conference
call on the third Tuesday of the month at 2:00 Eastern time and that's the forum for
addressing hardware issues; they'll bring up anything that's going on with hardware, HP's
really good with that. And as I said, get on the listserv for any critical updates and
requests. The monthly minutes from the monthly call and from the weekly call go out to
the listserv; any updates, any tips and tricks that come through from imaging go out that
VeHU 182.doc 26
Transcript for 2008 VeHU Session #182
way. Even stuff about this conference went out through the listserv, so it is a really great
messaging. I find this to be one of the best supportive products out there.
And how many of you have ever been a test site for imaging? Was it a great thing to do?
Probably a lot of work, but it's been a great thing. It helps move the project along. It
helps your site too, you know help it figure out problems that you're dealing with. And I
also found out as a new coordinator, start out with a little test patch, it kind of helps you
really understand the system. A lot of times they'll have weekly calls, so it starts out with
an E3OR or a new service request going out there to say what do you want to do? What's
going on? How this would help, it goes through a concept development internal testing at
Silver Springs because it is a medical device and has some strict requirements, goes out
for alpha testing so that the first couple sites will alpha test it, look at anything, design
flaws in a real system to try and go through any bugs and get it to a point where they add
a few more sites and beta testing we see as we expand. And they try to get a large site, a
small site and a multidivisional - I'm getting my SQA person back there giving me a nod.
And then it goes out for release, and sometimes even with release things happen and
remedy tickets are great when you put them in. If you do become a test site, you'll be
given a script with some scenarios to go through. There are feedback forms that are a
must with this that you'll send back and forth. A lot of times there's weekly calls with the
software to discuss what's going on and seeing if there's any issues. So there will be
things that you'll have to report back but it really is beneficial and has brought this
software so far along.
And then this is our soap box, Brent and I as customer support, we are kind of in a role
where we have to deal with the front end users as well as the IT part of our job so it's how
us techie people tend to approach support. You know usually it's your problem and I just
have to fix it. Don't get in the way of my nice day that's coming along and sometimes
I've talked geeky too and they go can you change that wording around? That's 'cuz
sometimes I want them to know, oh yeah it's doing this, this and this. So we tend to talk
in that techno speak and want to just get them off our back today. How clinicians
perceive our approach is that they hesitate to call us unless there's smoke coming out of
VeHU 182.doc 27
Transcript for 2008 VeHU Session #182
the machine and I've seen that, you know why did you wait three days to tell me that you
haven't seen the images for a week? Because they go, because you don't tell me right.
The correct approach is their problem is your problem and if we work together we can try
and fix it. Your problems are part of why I'm here, it's job security, that's why I'm here.
And speak in terms that they will understand; how they perceive when you do this
approach is that they're more willing to communicate, they're more confident that the
system will work for them at all times, and they're much more forgiving when there
really is an issue.
And I have to say from one of our areas with some of the support is as we were
introducing the commercial PACS along side the VistA Imaging, and all they wanted to
see was this commercial PACS, commercial PACS, commercial PACS, we had a tiger
team that kind of went out and kind of made things work both ways because VistARad is
our backup if the commercial PACS is down, so we had to make them feel good that both
systems would operate pretty much 99.99 percent of the time. So we actually said okay
let's brainstorm together, and we now have instituted a kind of monthly radiology what's
your problem kind of meeting; get them on the table because they were actually sending
messages up to the director's office at one point and it was being more of my headache
than trying it, so we ended up saying okay we're gonna address this head on.
Brent: And I can say that I'm pretty proud of my imaging folks at my facility, they've
carried a great legacy of keeping in touch with the users, they've gotten to know them by
name, they take walks up there periodically and say hey how you guys doing? If they
have some free time that they can spare, they see it as an investment of their time to walk
up there through radiology and actually go looking for problems, and that's not in our
instinct as IT people or technical people, we kinda like want to let a sleeping dog lie. But
we want the users to be comfortable to communicate with us and tell us because too
frequently they get frustrated with a process that's not working and they just shut up and
you assume that because you're not hearing anything, oh everything must be working
better, and then you find out when you're in a director's conference that it's not working
and we've seen that that approach, that pro active support approach, our users have
VeHU 182.doc 28
Transcript for 2008 VeHU Session #182
defended us in front of high profile meetings and that just - you can't put a price tag on
that. Plus it's just good business, it's just a good way to do business.
VeHU 182.doc 29