Tip:
Highlight text to annotate it
X
>> Umit Topaloglu, Ph.D: Thank you Johnny.
Hello everyone and thanks for the opportunity.
Today I'll be talking about the Comprehensive Research Informatics Suite
or the CRIS that we have been developing since pretty much 2007.
It started with the Cancer Institution Dr. Laura Hutchins,
Tom [inaudible] and Sheryl Lanes,
region of having clinical research informatics platform at our Institution
and then I have involved and start initiative pretty much end of 2007
and early 2008 and since then we are actively working
on developing this infrastructure and today I'll be talking about some frames
that I feel important I'd like to share and please feel free
to at the end ask any questions you might have.
So what's our vision?
We want an Informatics support
for the entire biomedical research lifecycle in our institution.
As you very well know it starts with hypothesis development,
securing some funding, conducting your study, analyzing your result,
maybe tweaking your hypothesis and this is a complete life cycle.
We want this to be supported by the informatics tools and services as much
as we can and as you'll note we are focusing on genetic and biological factors
and now we are looking at behavioral and social factors,
trying to understand the disease,
the environmental factors also playing an important role in this paradigm
and as we think this is biomedical lifecycle a network the patient is
in the center of this network with a lot of attributes with other notes.
This comes with a lot of multidimensional data, a lot of variation in data
and it's very challenging as pretty much everybody is working towards addressing
these issues, pretty much [inaudible] and this cannot be addressed
from our point of view having a map in place
and this is the workflow we have developed starting 2007-8 timeframe
and we have been following.
Everything starts from recruitment competence
and the dark blue competence are the research tools that we currently use
in production mode so everything started recruiting participants registering
on the right side and then in the middle data management section that's
where the studies are being conducted and on the left side genomics
and specimen management and the left corner is the research billing side
for clinical studies, that's another big issue that I will be talking today.
And at the end we want our data to be accessible to researchers.
Towards this effort CRIS currently supports National Children's Study,
I will briefly mention about what we did on that.
Our cancer clinical trials, community-based participatory research
of cultural public health.
We are conducing many studies in different counties
across the state using CRIS platform.
Our Translational Research Institute,
our CTSA but is currently using and implemented the CRIS.
We are working closer with genetics department with Family Health Portrait
and other components of CRIS.
Gene-expression profiling transcript on genome sequencing
with birth defects and Cancer Institute researchers are on their way
and finally tissue procurement facility results also using CRIS platform
to collect and annotate their specimens.
So with this we have 258 projects internally currently
in our CRIS environment and 273 users.
I want to note that that 258 is prospective studies not [inaudible]
and other type of similar research is not included in that number
and here are the logos of the institutes that CRISS uses.
We have implemented CRISS for 5 out of 6 institute in UAMS,
those are Cancer Institute, Translational Research Institute,
Psychiatric Research Institute, Myeloma Institute of Aging and three colleges:
Medicine, nursing and public health,
as well as Arkansas Children's Hospital Research Institute
and our own National Children's study side
of Benton County are CRIS' current users.
In addition to our internal views,
we are in process of migrating the following Universities our CRIS platform
within the National Children's Study Program those are we are currently live
with the University of Florida, we are in process of moving case faster
and as I speak training is underway and UC Davis also we are in process
of moving the data and Saint.
Louis, UT Southwester will be the next to follow
and as of last week we are told we will be given we will be keeping Cast Western
and Saint Louis University County research, National Children's Studies research
and we will be given 8 more counties starting from South Dakota to [inaudible]
in the central region and this graph shows briefly our start of the CRISS
and in active projects in our institution.
So one thing I like to notice here is we have gained a lot of knowledge
as we work with different groups and one thing I can proud
of saying is positive feedback.
As we learn and implement things in one area
that becomes available or useful for other area.
For instance consent [inaudible] options created for cancer clinical trials
and that was very useful for National Children's Study and some other studies.
Familial relationships among participants was implemented
for National Children's Study and others are also using.
Grouping activities collapse expand for the PSC, other use of time,
not showing the time on each activity
in the patient study calendar is also useful
for National Children's Study as well.
So we have also implemented one feature you can move epoch back
and it turn out in a couple of studies in Cancer Institute that was a useful
and required feature so and currently we are in process of implementing
of finding a collection for National Children's Study community based
participatory research, translational research institute
and cancer control researchers who are traveling different counties and state
and trying to collect data from participant in rural areas as well
where in turn is not always readily available.
So if you go into a little bit more detail in the CRISS I want
to briefly mention the basic function
and the next slide will have the complete list of components of CRISS.
So for recruitment we have our trial search and we use caBIG for registration
and participant management.
Patient study calendars are Study Calendar Management system.
For Electronic Data Capture OpenClinica
and LimeSurvey are the tools we currently use.
Tracking Regulatory Events is being managed
by our Internal Developed Event Tracker.
CA tissues are by specimen and caArray is our data management solutions we use
from caBIG again surveys are being implemented by using LimeSurvey.
Adverse Event Reporting and Toxicity Grading we have three tools we are working
in our environment.
caAERS/CALAEGS from [inaudible] and UAMS AERS have been developed internally
and for free text extraction ca times
from Pittsburg is the tool we currently use and this is the list
of the tools we use currently in the CRISS platform.
The red ones are from caBIG and in addition
to the caBIG there are other open source tools that we use as you can see
on the right side like MIRTH Connect and other tool we heavily rely on,
Family Health Portrait, I2B2 as well.
Again if there are questions at the end of any
of these tools I would be happy to discuss those.
In addition to having those separate modules and providing services
to different groups, we needed to implement several integration scenarios,
those are basically broadcasting the participant registration
so once you reach your participant you get time consuming
and less better possibility we want to broadcast that to other tools
in the CRISS Platform and for
that purpose we have implemented a MIRTHConnect enterprise service bus
that will register same participant in caTissue if a consent
in C3PR indicates a specimen collection is allowed.
Patient Study Calendar, OpenClinica,
LimeSurvey and caARES are other tools are MIRTHConnect enterprise service bus
broadcast The way it works if study is created
in this platforms, subject is registered.
For instance, for sponsored studies,
OpenClinica our ECRF solution is not being used
so OpenClinica will not register any further return a message saying subject
hasn't been registered because study's not there and that's quite useful
and our users are very happy with that feature.
In addition to broadcasting we are pulling patient registration and demographics
from our DMR system procedure DPR so we don't want people to write patient names
as long as it is allowed in their consent to use medical record number.
Clinical laboratory results of our participant to PSC mitigate
and toxicity another feature we have recently implemented
and I will be talking briefly about that in following slides.
We use patient study calendars for research orders to address billing problems
and I will be talking about that as well.
Family Health Portrait/caTissue/LimeSurvey integration is another use case we
have implemented and I will briefly talking in following slides as well.
As we are moving into CTCA version 4
for adverse event their collection we have implemented AERS tool
that is being used currently in the clinic and after it has been entered
into that tool we broadcast that adverse event to caAERs,
OpenClinica and our EMR for not
and the last thing I will briefly mention will be the budget tool
as we develop budget for RIB approval we have functional expert
that patient study calendar to continue registering subjects without needing
to create new calendar on approved budget.
So let me briefly show a couple of screen shots about our platform.
So this is a screenshot from the National Children's side group.
As you can see there's go to on the right top that has a lot of groups in there
and based on your login you're allowed to see some or not
and you see a test site from one of our National Children's Study as you look
and you will be able to see the links caTissue, LimeSurvery and a red line here
and there are some port lets in [inaudible] that allows users
to utilize different functions such as for National Children's [inaudible]
if you are going to call somebody
and collect some data you can use this Google map feature
to travel and find where they are.
The next thing we are heavily involved with in training any users
so you see a screen shot where our participants are have access to a lot
of documents and if you are able to you are going to see a lot of videos
about how to implement some functions and on the right
or left tab you will see calendar,
documents and [inaudible] screen you are seeing right now.
Red mine is our issue and feature tracking
and management system that we currently use.
Question and feedback is an environment where our users ask questions
and discuss with everybody of the portal user and one of the features
that you cannot turn to see on the right bottom,
you see who ever is online currently using the system and you have option
to chat with those people as you are on the portal and Wiki is the last feature
that we involve our users and get them trained
or find what they are looking for.
The next feature is the record component
as the trial research this is our internal developed tool,
all of the data is coming from our IRB protocol management system,
we don't want users to change any of these after IRB and one change we are
in process of doing is using [inaudible] to classify those diseases for people
or investigators to easily access the studies they are looking
for or interested.
The next feature is again internal developed event trackers.
This tool has several options like you can track your submission into IRB,
parents or different communities and I have two screen shots about this tool,
one being this delegation log.
We have implemented this to allow PS or staff to go there
and acknowledge they are assigned to different functions for a study
and at the end they can email PI and PI and can come and sign off these
and it also has a feature to remove these people from the project.
This is one of the important features our users wanted us to implement.
Another one is the contracting our translation research institutes clinical
services core charges for day of services and they want to track their time
on each study and this module allows them to go in
and specify what they have done and on this study
and how much time they have spent and based on this we create reports
for each study which I will be showing towards end of my presentation.
Next thing is the patient study calendar I will want
to discuss with the audience.
This is the patient study calendar developed
by Northwestern University [inaudible] a big program
and we have made some changes for instance since the computers
in clinic are not big screens, they wanted to hide timeline feature at the top
so it will give more area for them to see the activity itself,
also expanding or collapsing activities as you can see
like September 7th have several different types of activities
and they are in plans that as you register your participant everything it
becomes planned and then we order and terms scheduled as t progress
and one thing I want to note each activity has CPT codes,
ICD9 codes and R or C at the end in the label field specify
that activity is dedicated for research or should be charged for research
or if it's culmination of you'll be seeing Cs for instance
on the fifth lab activity it's supposed to be charged to culmination
and it has ICD9 code and safety code in it.
This allows us to track the activities in a research study and charge properly,
which I have a couple of slides about that.
One other feature we have done
in patient study calendar is emulating the lab viewer so as you can see
in the yellow kind of right top is the lab result.
As soon as you click on this it will open up a new screen where you'll be able
to see a grid with all the lab results and with that [inaudible] grading.
The way it works then we'll receive an HR7 lab result,
we look if that patient is our participant, if so we display the lab result
to you at the same time we have a lab service call the [inaudible] grading
system and if there is a toxicity it displays the result over here.
Again since this is a test patient
and we don't have a test patient test result coming
from our lab information system I cannot show any
but in a real system you'll be seeing a lot of lab results.
One other feature I like to mention as we are ordering labs
from patient study calendars
and if you see those results back we automatically mark those lab activities
occurred as we know that has been ordered within the study
and we have received the result.
The billing and research order is something I
like to briefly discuss with this group.
That's one of the big issues we are trying to address in our institution
and one of the challenges are the terminologies.
As we create our budgets before IRB approval,
everybody uses CPT codes in our institution
and we develop a study calendar template using CPT codes
but as you may know CPT codes are not very clinic friendly
and Office Visit is a very good example.
There's one Office Visit CPT code which may include 1
or 8 different items you may or may not need to do for specific visit.
SNOMED CT is giving that [inaudible] for us so we are trying
to convert those CPT SNOMED codes for the patient study calendars.
Again it's not an easy process and it's not automated but we are doing as much
as we can manual and automated and the challenge again after it has been ordered
by the clinic when it comes through billing part,
billing system doesn't know SNOMED as well,
they know CPT and they charge with the CPT
so we convert those back to CPT codes.
Again very challenging process.
I don't know if anybody has easier way to do it I'll be happy to listen to that.
This is a budgeting tool that we have implemented internally called Clara
that is an institutional requirement that you need to go in
and answer your budget into this tool for every human subject studies
and this gives us all the culmational research or invoicable items identified.
On the left top there screening
and treatment those are the impact design we have created so it will sync
with the patient study calendar and [inaudible].
Again at the end you will have per subject total for each activity
and when those activities will be happening and where to charge.
Another web service we have developed we pushed this
after this is approved we push that patient study calendars
and you will be seeing calendar and patient study calendar
after you register all of the activities with CPT codes and charge types
as you may remember from previous slide.
I didn't want to have several slides in this one
as in the background I don't know how clear it can be seen
but you can collect activities that you want to order in the clinic
and on the right side there is mark ordered option.
As soon as you hit mark ordered,
if you are not an authorized physician you will see this screen asking you
to pick an authorized physician to place an order.
If you are an authorized physician logged in you won't be seeing this screen.
After selecting authorized physician order will be placed.
This diagram shows a communication flow.
After order is placed in the patient study calendar,
we send a copy of that to our center's EMR system I will have a screen shot
in next slide and then based on the schedule
and the day of lab we send an HR7 order to our lab system
with the proper charge type if it's R or C and what's the ICD-9 code,
what is the account number.
I don't know others but in our institution we have different account numbers
for billing purposes for research and culmination of types and our culmination
of accounts changes monthly for each institute.
It's very complicated and at the end every activity completed then hospital
and professional billing system knows what activities for this study completed
for this subject and what our data charge types at the end.
Here is a screen shot of a research order we sent our EMR system
and as you can see we summarize [inaudible] lab and the account numbers with it
and the in detail you're seeing for instance this is authorized
by Laura Hutchinson and entered by this person.
We catch that parent ID is something needed for billing purposes for those
who knows billing and all the details is listed here.
So I have mentioned these so I will skip that.
Again, our goal is to track all the activities being done for research purposes
and properly charge if it's going culmination or Medicare or it is going
to be charged to research study.
[inaudible] data capturing part for returning CRFs we use OpenClinica v3.1.2
and we have created a lot of documentation about 21 CRF Part 11 for IND studies
and we have put lot of CRFs created using common date elements.
Currently we have one staff person that's trained to be a creator
for common date element she has worked for Diane
and learned about creating common date elements as we create these CRF pools.
I want to continue with adverse event and lab grading.
As I have mentioned earlier we have developed at AERs system
to capture CTCAE version 4 the reason being
as our EMR system doesn't have CTCAE version 4 implemented in it and it wasn't
that easily available so we have integrated our AERS with EMR system
that can jump from EMR system to our AERS to capture CTCAE version 4
and after that it sends message back to our EMR AERS and OpenClinica.
We have used LimeSurvey to implement I believe 7
or so number of CTCAE terms reason being
when the version 5 comes it should be not too complicated
to implement the new version as well.
For the lab rating we have implemented CALAEGS
from COH currently it's working in production environment.
Again, as I mentioned previously
when we receive a lab result HR7 you get the result and sent web service message
to CALEGS and it returns if there is any adverse event or if it can grade.
Here's a screen shot when you log in when you're trying
to enter new adverse event you will pick who is the provider
on said date and affected system.
We have supplied symptoms and diagnosis for easy access and if you can see
if you bring your mouse over it explains what's the adverse event.
I don't know how it is readable but oral pain, stomach pain,
those are the two disabled, the reason being is this patient has already ongoing
adverse event on those two items that has not been resolved
so it won't let you add new oral pain,
you should go and *** ongoing adverse event first and you cannot add new one
until you act on the previous one.
One small thing I like to mention being very helpful
as the CTCA forward link on the right top.
This is the complete PDF guide
and as our clinic computers cannot access internet
and nurses were carrying those CTCA booklets all around with them,
this is easily they want to know how they can grade they can just click
on that link and it will open up the whole guide
and I think they find it very useful as well.
Again this is the screen shot on the ongoing,
as soon as you log into this patient
if this patient has an ongoing adverse event,
this is the first screen you will be seeing.
Wherever you see those pencil shapes those are the ones can be updated.
Again we keep grade and ongoing as you add ongoing with grade change it will add
to this table in the middle and [inaudible] locks this record
and it sends message to those tools.
And here's a message that we send
to our adverse event message our clinical note one feature working
with [inaudible] if you provide a tech number
and we send with that tech number it can insert that section of adverse event
into the clinical note that provider is working on and as an example we send
to our EMR system and here's the web service call and the same adverse event is
in the care system and from now on you get
to continue reporting and report accordingly.
For other data collection purposes we use LimeSurvey tool, it's very easy to use
and we have integrated our sign on and CRIS platform.
I don't want to talk too much about other
than we have implemented offline data collection.
Again internet is not available all areas in the state
and that's why we are working on developing this tool to allow researchers
to collect data across the state.
Family health history is a area where our department of genetics is needed
so this is the screen landing page if you will.
When you look and you will be picking your study
and after you pick you'll be seeing everybody from CPR here,
the green checkmark states this person has family history has collected
and if you click on that it will open up this show the result.
The plus on FHH column shows that person doesn't have family health history
and as soon as you click that it will let you continue heading the family has
history of that person and for this study LimeSurvey external has those
checkmarks, those indicate this person has completed environmental questionnaire
and the last column is the CA tissue
that checkmark states this is a saliva collection study
so that person has a saliva collected in CA tissue and if you click
that checkmark it will take the CA tissue screen it will be showing
the specimen.
I think our genetic department found this very useful as they are working
on different similar studies and here is a screen shot of pedigree
from surgeon general's family health portrait tool that we have implemented
and we have made some changes to store into data base and so on and so forth.
The next area quickly to go over is the data quality.
Again as we collect tons of data there's always some problem
with the data being collected so with working with information quality program
in our system they have developed this open sYSTem Entity Resolution tool
which works on the different set of data.
For instance in one study we have run this tool and among the 150
or so participants it was able to find several issues, for instance row 9
and 10 is the same person but they are assigned different participant ID,
maybe just because their address was wrong.
Anyways it's very useful tool to identify issues with our study and data.
Quickly going over to next topic is the reporting in data access we use SSRS
as our reporting tool for instance our [inaudible] report is being put
automatically you don't need to enter much information
and here is how you access the report.
Because of time I'm going fast at the end.
If you have further questions we can discuss and here is a screen shot I
like to show if you may remember on event tracker page you are able
to enter your time adverse event or different submission time for a study
and in this report our translation research institute services core can see
those and create a charge for this study and in this next screen
if there is any submission or if there's any subject activities it will mark
and at the end it will give you a total.
This total giving our core to be able to track all of the activities they do
as a nursing or regulatory services and charge the study accordingly.
One thing another thing I want to mention is the specimen collection,
I'm not going to go into too much detail other than we use extensively
into different biobanks across our institution and Arkansas Children's Hospital
and we are in process of migrating a lot of specimens in that.
What I like to mention is I2B2 to for those
who don't know I2B2 is a cohort discovery tool allowing you
to find queries on limited data sets.
We have worked on and mapped our tumor registry, ca Times and caTissue in here
and allowing our data warehouse to complete the data for researchers
to search on the data they want.
One thing we are currently working as we have mapped ca tissue
into our I2B2 environment but as I mentioned we have several biobanks how we can
map those several bio banks into I2B2 environment.
For that purposes we have worked on creating an ontology for biobank
and using those for annotations and CA tissue so it will be mapped I2B2 properly
and readily available to researchers, that's one ongoing project.
Another project we don't have publication yet but implementing another ontolgoy
for LimeSurvey as different researchers developing their forms they can use
those ontologies and after they complete
that data collection it can be easily mapped to I2B2 environment.
One other thing I want to quickly mention is the paper or project we worked
on we have worked on [inaudible] to find direct link to adverse event
and we have a publication about that.
The reason I have given those two examples are eventually for eligibility
and finding eligible patients.
The area we are working currently,
the studies we have if we can cluster those properly,
we can use the existing ontolgies in our current data
and maybe run a best process to identify patients.
Another one, again based on the same cluster if you can go to social media
such as [inaudible] or others if you can identify potential participants
for our studies that will I think increase the participation
in studies in our institution.
Again, we are working on different compliments to implement
and put into our CRIS environment.
So what's next?
We want we are talking several other institutions that want to implement CRIS
and we want to work more on the 'omics' side of things
and participant record and registry.
Due to time I'm skipping this.
One thing we have worked on this project for several years
and we have gained a lot of experience on clinical research from different range
and one thing we have seen the functionality created like therapeutic studies
as becoming very useful for observation studies or vice versa.
With that I'd like to thank the whole team.
We are 20 plus team working on CRIS team members working on CRIS.
We have Cancer Institute, TTSA
and the National Children's Study contributing on the project.