00:00 Welcome: Georg Neuwirther
06:58 PHAROS – Implementation of ISO-IDMP Concepts: Georg Neuwirther
25:41 Integration of EMA SPOR services – OMS data: Georg Neuwirther
35:48 PHAROS – Importing eAF data based on the new FHIR format (Variation of CPs): Noel Diamant
53:32 Integration of EMA SPOR services – RMS data: Noel Diamant
01:09:19 Integration of EMA SPOR services – SMS data: Noel Diamant
01:16:06 Providing Medicinal Product Data in FHIR 5.0 format – Outlook: Noel Diamant
01:21:23 Conclusions & Take Home messages: Georg Neuwirther & Noel Diamant
The slides displayed during the webinar can be consulted here: https://unicom-project.eu/wp-content/uploads/2024/03/2024-02-28-AGES-best-practice.pdf
Thank you good so I’d like to welcome you to this breast practice Workshop from the Austrian agency hello to so many participants from inside Unicom I’ve seen many members and I also see many members who are not currently part of the Unicom project so that’s even much better that we can share this
Information with Unicom Outsiders as well um I also would like to welcome my agus colleagues I saw some names in the list so thanks for joining this meeting and as you can see on the starting slide thanks for all your contribution uh to make Unicom a success at our agency so
Thanks a lot today together with uh my colleague Noel damant and Sonia kixa we tried to set up interesting 90 minutes hopefully this will be enough because we have a very dense dense program uh my name is G NOA I’m head of the Austrian medicines head of it I’m
Not head of the agency I’m just head of of it of the AUST Austrian agency and I would like to ask Noel to introduce himself yes good afternoon from my side I am afraid the camera is not giving a proper picture so I’ll just uh I just talk
And where am I I’ve Been Working for with G in the austan agency for the past 15 years we’ve you look fine don’t worry ah now I can see myself perfect uh we’ve done from it everything there is to do uh starting I think the well nearly all of the projects um I’m
Bioinformatician and I’ve been working together with EMA for the last few years uh with the Unicon work package three to try to improve upon the electronic application forms and the product life cycle management so today I here to talk to you from the other side as a as an
Employee from the austan agency and how we integrate the things that we also um try to create together with EMA thanks Noel so let’s have a look at our agenda ah so first of all please uh consider uh the disclaimer and please be also aware that this session will be
Recorded and published on YouTube or wherever Luke who is our communication officer plans to publish this important information um our agenda is quite dense so if some things are too fast then please let us know via chat and there’s always the opportunity to have follow-up sessions if you are interested in any
Details of what we present today let me introduce our agency so we are part of the arus limited company so the arus medical medicines and medical devices agency is part of a bigger company uh and we act on let’s say if you know this this um uh this objective
One Health so we have different divisions working on patients and veter AR health and on food safety so we are dealing here with licensing authorizing of medicine products but also do um post marketing authorization uh post marketing activities and are also are responsible for medical device vigilance and many other business processes it’s
Very interesting to work in our agency uh we are publicly owned so we are owned by the Republic of Austria and uh the division we are working for is also the business unit to provide information to the federal office for safety and health care who delivers the official letters to the
Applicants in this presentation today we focus on deliverables that we covered uh in the last years with within the Unicom project so we have let’s say input material and input services from EMA that we will cover today we have new opportunities with the new application form that Noel has already mentioned a
Bit before uh another input of course are the European implementation guides where we take information from then we will see things that we have done in our core it system which is called fos that we introduced more than 10 years ago so we had to ref fact to extend our core it
System system meaning new US interface is a new business logic to handle new data elements and of course a much more complex business data model that comes with the ismp concepts apart from technical stuff we had to do a lot of updating Sops and training material for
Our colleagues to deal with the new elements that we introduced with this and this also has impact on our data consumers so you will see some uh details H how we export data to ealth uh consumers ealth data organizations uh how we support crossborder prescription which uh is an
Objective in Europe and we will also get some information about a prototype that we are are now evaluating together with Elga Elga is our National ealth Organization in Austria before going into details this we would also um inform for you that we are contributing in another work package
Uh in work package three that sets up the new PLM portal together with EMR and we are very proud that agus are is contributing with the network product owner in person this is Noel damant I hope the video and the photo are still the same yeah similar artificial intelligence could
Find the match between the photo and the live picture from Noel so we’re contributing and acting as the network product owner together with Christina pasari from EMA and there we are working on the implementation of the new application form creation tool and I’m pretty sure Noel will mention some details a bit
Later first we start with what did we do in fyos as I said before fyos is our core it system this means faros supports I would say 85% of our business process that we handle uh in our agency so ISO itmp had a massive impact on this
System here you see an overview from EMA about let’s say ISO idmp Concepts this is a very old picture but I still like it and in the purple purple is our color uh um circles you can see which elements we touched or even fully implemented so nearly every concept from EMA was
Considered in our Unicom work some of them are a bit delayed to do missing specifications or to missing information or we only could implement the structure but still have no data because there’s no data in the regulatory activity available but uh we covered most of the of the iso idmp ISO idmp
Concepts so so what were the key changes in the core system first of all was uh we had to implement the administrable product which we all know is also named by with a pharmaceutical product please do not be distracted in this presentation we use the term administrable product this has some
Reasons are in the EU implementation guidelines or the standard it’s called pharmaceutical product so it’s a synonym now um it was not that bad to implement this because our database was already uh based on a reference data model called RDM version 3 where some of the iso MP
Concepts were already included so it was not a total new database model in this aspect we had to detail the packages you know now we have a more detailed structure with ISO MP this was our a big challenge in in in the Unicon project in indications is a new information that we
Didn’t have before in our system manufacturer details were enriched with new data elements we also considered uh to have part names in the future and we implemented some minor thing but also interesting to mention the additional monitoring attribute are as described in Chapter 2 of the EU
Implementation guide and then I have one slide describing or giving some details how we did the migration of our human medicinal products into this New Concept so let’s start with the administrable product and please let me know if you can hear if if I’m too loud
Or too not too loud or too fast so that I have some feedback from the audience here um the admin product concept was new in our fars world so we didn’t have this before before we had something which was very similar or was similar to the manufactured it
And this similar manufactur item class was let’s say the the data element leader uh and this we had to change uh into the administrable product while doing this we followed the following principles first we wanted to uh implement it in such a way that the migration of Legacy data can easily work
So we we try to find synergies between business rules that we have for migrating all the data from the old world into the iso MP World second we wanted to make sure that this New Concept can be populated with ef data so that we don’t have to manually type data
Into this new class third of course RMS terms are mandatory um pretty sure all of you are having this principle already then we wanted to minimize data entry effort idmp brings a lot of more data elements in a much more granular way this means if you don’t find clever Solutions you
Will have a lot of lot of more manual administrative effort and we try to invent something and we will show this to you to minimize this administrative effort apart from automized import from application forms and we wanted to find a good balance between enriching the user interface with new elements versus
Uh complexity of the concept on the next slide I will give you a screenshot and we’ll explain what you can see are as you can see here in our in our it systems we are German of course but some terms are in English not because we are too lazy to translate but
When we speak to each other we sometimes really use English terms and that’s why you find them in the user interface too so first of all you can see in this red box now uh the overview of the administrative product uh where uh colleagues can get the most important
Overview information of this product which is administered to the patient uh you can also see that we have a national ID for this uh administ product but also placeholders for IDs from imas Spore um in this in the in the screenshot you see an ID for Austria and
For le le stands for lonstein that’s because we are processing marketing authorizations and doing life cycle management for lin as well the next box shows you how we reference B effected items there are different views how we do this but here it’s very Sim simple shown uh this administrative product consists of two
Manufactured items and here you can see the numbers of these manufactured items for referencing uh internally we use the internal National IDs next one um a very important information um talking about ingredients and composition of administ products are denominators and what we did we are we put this data element into the administ
Product and you will see the same concept in the manufacted item as well so you can see here uh the strength denominators here per One V or per one millit in here uh and this is only defined once for all the ingredient strengths that you can see below so this
Is one of our or let’s say tricks or solutions to minimize the data uh effort and also reduce error reduction because you define this strength denominator once per um the composition of this product last but not least the root of administration is now in this administrative product and below this
And here you can only see one line but of course there can be 10 15 four lines of ingredients uh which are part of this administrable product the ingredient line is the line where you have your substances for from SMS and the strength information and many other
Information uh this we didn’t have to change too much so this was already very let’s say um similar to ISO MP Concepts in our in our formal database model second we go to the to the implementation of the manufactored item so the manufact item is what is produced
And may be combined reconstituted in administer products and this screenshot is a special one to explain one of our ideas that we implemented um quite often administ products and manufactured items are the same so for example this is applicable for all tablets or any product that doesn’t need reconstitution
So we followed the goal to minimize effort and doing error reduction so you can see in our in our uh in the screenshot here in our um far system we duplicated the administrative product Concepts because it’s they’re very similar to the to the administrative product then we introduced a flag that
Marks that this manufactured items is the same or is is the administrative product so people don’t need to enter the same information twice and we enable that in this scenario manufactured items also have a root of administration of course this is an exception when you have a power
Solution you won’t find this root of administration in the manufactured item so this is one way and then of course you can see the list of ingredients of this manufactured item very similar to that Min product uh this is quite the same as before so
You can see with this we can reduce the amount of let’s see data handling or tasks to our uh colleagues and uh all the reports that we’re doing or the search um um search filters or act data exports uh are evaluating this flag and are handling accordingly uh if it’s the
Same or if the administrative product consists of one or many other uh affected items then we come to the package the packages I like the packages because this is what you can buy in the pharmacy and the package concept is is much more detailed and granula than we had before
So here you see the concept of the package medicinal product I have a new approvation I didn’t find an official one so I call it P PM P um and this p medicinal product has shelf lives one or more uh that are now able that we are now able to
Store and it consist can consist in this example of two layers so it’s a box first layer and then inside the box we have two vials with two different manufactured items so you can see here the shle is the box and then you have two D flash vies which are inside the
Box and you can describe this with the container and the container material and the container Parts material but also here you have find the place where you link the quantity of the manufactured item um so you link the package layer with the manufactured item and give the manufactured item quantity here so to
Inform how much of this manif item is inside uh this package item layer please be informed that the unit is not displayed or inside this table because this can be reused from the manufactured item so this is another trick to reduce data administrate data Administration or even avoid errors that you have here
Different unit then you have it on the manufactured item here you can see the reference so you can see this National numbers again but also including a texal description so if you have a glance on the package that you have um let’s say summarized information what this manufactured item
Is we also enriched information and data elements for medical for devices that are Copa or part of the package layer and oops I go back sorry and um we tried to introduce some user experience elements so you can see on the left that you can uh collapse or
Expand the package layers so that if it’s a more complex situation that people can easily navigate through their package layer structure yes then let’s move on to I think two more concept that are not that complex that compared to the admos product and the packages um but they are
Giving future opportunities to faros the first one is we implemented data structures uh that can keep indication information so we are now able to import I need to say automatically import uh indication information from the smpc from section 4.1 and Below you can see data elements uh that can keep coded
Indication uh so these elements are prepared and the plan is that we will consume this data from sport PMS once sport PMS is available for data transfer so this this op opens up further use cases in our agency in the future Another one not very tricky but
Interesting uh name parts we are aware of ongoing name parts discussion in across Europe so how much will be mandatory who will do their name parts um splitting for legacy products Etc so we didn’t invest too much time in that but we set up the foundation that means
Uh business colleagues can select or can say which name parts they would like to enter into the system so there we have catalog saying this is the um strength name part or or some other name parts sorry they are in German here it’s too hard to translate to me on the fly but
This is a catalog that can be enriched or limited however and then you have a dynamic table where you can enter the name parts or you can import and this is our plan we will not do this manually we will import this information from uh the
New PLM portal created EFS where Noel is working on together with with EMA last but not least in this here um Manufacturing information becomes more and more important when we think about the European shortage monitoring platform we know that we need information about manufactures Etc so
This is a domain where we invest time and money to enrich information in our system So currently we have already good databases So currently we are importing this information from the application forms but now we added a new information let’s say we prepared this uh we also want to have information about the
Licenses and the um licensing um Authority so you can see here in the red box there are new data elements that are prepared for the future when we get this information from the electronic application form we will import this and can display or report it or summarize this uh for our business
Colleagues as the next slide um I’d like to say some words about how we dealt with the migration from all the Legacy data into the New Concept so I’d like to thank again my colleagues and external Consultants who helped us in thinking how to do this because this could could
Become a disaster if you do it if you don’t do it well here my colleagues or our colleagues did it perfectly I need to say so we planned the stepwise migration of Legacy data from bottom up so we started with enriching the manufactured items so this is something
We had very similar already before this concept and so we started with this then we created this composition of which administrative product consists of which are Mees much of that could most of it could be done could have been done automatically with business rules and with great ideas from experts in our
Agency and then we moved on to the package layer and Link the M with the different package item layers um and so we had this let’s say core con isod concept where uh a lot of Legacy data migration had to be done the other data elements like manufacturing
Etc uh this was not so much of concern to us for legacy data migration um manual preparing Legacy data was needed for an old n for number of old products so some of very old products were not let’s say in in not the quality the quality was good but not
In the shape or not in the structure for Automatic Transition and we had to enrich the unit of presentation information into our manufactured items so this was a lot of manual work we had to do but finally did it uh and the existing package information our system was very helpful
To find out what is the the correct unit of presentation the package layer structure we found a way to do much of this uh automatically so had to have all these layers um but then of course some correction had to be done um and the most effort was for the need to correct
The M manufactured item quantity information so this was where we invested uh the most work when we talk about packages perfect um this was let’s say the the the the presentation about the idmp work that we had to do in fos and now we will go into three presentations the first
One is how we consume OMS data in faros the second and then I will hand over to Noel is how we consume referential data and the third one will be about substance uh cons consumption in in faros and I will start with OMS oh that’s that’s me it’s not Noel sorry Noel
Um good and I think we’re good in time Luke yes I think Noel what’s your opinion yeah it’s perfect um yep good perfect another hour yes I think that should work um I don’t want to spoiler too much but we need time because we have also prepared a live demo we we’ll
See if this works um so what was the idea about integration of OMS so what was the concept in the middle we have our core system which is faros already said before the input for organization data is the application form so we do import organizations from example the
Application form and then we do the update of OMS data via regular OMS API LS yeah so we use uh only applications like application forms like clinical Tri or other applications we are using to create new organizations in our system and for the for the regular update so
When any any data changes in OMS we use a new OMS API uh concept to receive this information and to process process it Accord ly I will have a workflow in a second about this um yes it’s not on the slide but maybe it’s a good information we do not keep
All OMS entries in faros we only have those um records from OMS that we need for our activities so for example for regulatory activities then I would like to inform that for official letters we are still using faros data so we’re not relying on OMS data due to some reasons that maybe
You are also aware um um so for the official letters we are still relying on faros data um but all the fos data in organization aspect are linked to OMS now so we know even if my street is written in a bit different way we exactly know which of my record is
Linked to which record at OMS for this we have some for this sying concept we have some kind of process which I would like to uh present you not in detail but you can read later but I would like to explain the the the the thinking that is behind this process so
First of all we are calling the OMS data repository via rest API calls and we are receiving a CSV file comma separated file so we are not are having multiple OMS calls for every OMS record we have one call that delivers a CSV file from EMA
This is one option EMA provides and we had an architectural decision uh to choose the CSV file option due to some instability if you call the OMS API too often uh this is done daily we receed this CSV file on the daily basis and then we compare the gaps between our
Database and uh the OMS database daily we are checking if there are some changes in the time stamps if yes then some our actions follow and on a weekly basis we check if the combination of the location ID and the organization ID still exists like it’s in in Emma
Database so we check if our combinations or lock ID is the same as we have it uh in OMS and this check is needed to follow merging and splitting of OMS records but also uh to follow all the active inactive situations that we have in in OMS systems here’s one box um crossed
Deleted in a way but I didn’t want to Blind make it blind because I wanted to say to you um we learned a lot when we implemented the system and it was an hi process so if you start doing this and if you then see the behavior of OMS if
You realize how OMS data stewards are doing their work then you might need to adapt your initial thoughts happened in our agency and I’m pretty sure that this flow will change over the years based on Lessons Learned but when we do the Gap analysis then we process the results
Many um changes can be done automatically and some but some of them also require manual interaction so this means that in our workflow system the system creates a workflow task for the people responsible for address management and they confirm if they want to take uh the change that comes from
OMS with OMS we had a lot of findings and challenges so compared to RMS synchronization OMS is much much more effort for our agency so first of all we are aware that the CSV file option is not the best one so we would like to get rid but at that time
Uh it was more stable than separate OMS calls maybe this has changed uh in the last one and a half year but at the time when we started this was the best option for us and now we are having some problems that the CSV file file contains different data than the user interface
Of omis so there’s there’s one open ticket and there was already one before maybe it’s already solved but when I did the presentation this was the case um we have a lot of now we have high administrative effort in dealing with OMS data quality so this is nothing
Unicor specific it’s we we know this from from stakeholders externals giving us this information I think EMA also aware of this and trying to deal this this topic with a key user group but it’s still a lot of effort for our business colleagues in this department to uh follow tickets to follow rejected
Requests Etc I don’t want to go into much more detail today then of course uml so you see it here in the item are uh always challenging then fundamentally we have a different approach in faros we have semantic lines so meaning Street uh house number Etc in OMS we have
Non-semantic lines we have four address lines and yes maybe there is some kind of rule in what stands in line two or line three but if you have a structured semantic line system like faros and a non-structured semantic line system like OMS then you can imagine that some problems
Arise um the district information might be normal for other member states in Austria it’s not normal to have the district inside the address Fields so it creates effort to delete it when we um update or when we receive this information from OMS um I was informed that there are also
Sometimes irritation because German company names and translated English ones or vice versa um have some kind of differences that raise questions and this is also a situation with the alternative names uh that are able in OMS and this also uh creates questions that we need to handle uh in our business
Processes for this for example I’m informed that in the application forms group um there’s a new backlog item a new user story that in the future you will be able to define the which alternative name you would to have in the application form so maybe this problem of alternative names would be
Solved when this will come possible in the application form but finally we now have OMS data in faros and we are so much proud that we also help our our inspectors because we of course also synchronize as you can see here the GPS coordinates so this is helpful for yeah let’s say finding
Addresses at Brad but also for doing travel planning when you visit more than one um location in one Journey so I think this is a big success we love it but we have still some troubles with the challenges that I presented before perfect I think yes I think that we can now move
On to Noel who will give a presentation about how we deal with application forms data uh what we did inside and before unicor with this and Noel I will do the time uh checking uh thank you G and thank you for your presentation it was so nice to relax a
Little bit and listen to you but we are here to work and I’m going to tell you three things today one is the importing of the EF so for once I can I can actually feel myself from the other side usually we’re creating EFS as part of EMA today we are
Receiving them and trying to put the data automatically into our system the second part as G mentioned is going to be on RMS how do we keep those synchronized with our list and SMS of course um the last bit that we’re going to touch on is
A bit of an Outlook and we want to provide the data that we have also to our ealth organizations and we are going to use similar Concept in order to facilitate that so I’m going to take over the screen share if that’s fine with you and hopefully you can see the exact
Same screen that you have just seen yes confirmed so starting with thef I mean it is quite everybody knows data is the reason why we’re all here um it is what keeps Us Alive it is what we receive what we validate and what we then provide so having the same data is
Kind of the quality standard that we want to get to having the same data in our system in applicant systems ideally in EMA systems is of big importance it increases um the quality of that data and at the moment the vessel that gives us the VAR well the product data is the
MAA form that starts with the product and the variation form that changes that product so for the MAA at least um we can we could take a lot of data for the variation we do import some as well not so much with different product data but at least enough to facilitate our
Systems and we are now prepared to receive more and more as soon as we have more data now just a bit of a concept of uh of where we are and where we’ve been and where we’re going to so the applicants at the moment are still most of them are just drafting EFS
The PDFs the Legacy PDFs that we all know and that we’ve known for many years but there is also the PLM portal product life cycle management portal uh at EMA where applicants can now at least draft um Central applications and hopefully soon they will also be submitting MRP DCP and National
Variations so this is going to use the entirety of sport so not just the organizations the reference the referentials but also the substance es and the products uh that should in fact already increase data quality by a lot and because it means well depends on the data quality
You have in your National system but at least then we starting to talk about the same data set now in order to submit this data we are still using the PDF forms but the difference is that in the PDF forms today we have something called Des which
Is our data exchange standard behind the PDF which is an XML that is proprietary to Adobe so the exchange format uh since we’ve been looking at it which is kind of 2015 uh is since we are importing this data um has always been the one from Adobe now there could be changes in
Theory that were unintended there were some difficulties with um weird things around Mt elements and so on so it does make sense that we are now going to use a a standard that is making its way through the ealth community and now through the regulatory um regulatory organizations like well ncas or
EMA um so at the moment we have these two options we still have the desk format for the Legacy PDFs but we also so are able to receive in theory um the PDFs that contain a fire message so fire is the other standard uh that is going to replace the
Desk now we just take the eect from well look at the module one extract the EF and import this information into our regulatory system to be well from a project point of view where we did this this kind of work package free has been creating the structure work package 4
Has been reading the structure uh in AUST we have this glad case where um these people are kind of the same uh so it is a bit simpler for us now to be a little bit more technical just a little bit um how this works and
This is a very old slide you can see the very old robot here um the applicant uploads a well to uh a PDF or an entire ecd we get that ecd we receive it via sesp and then we have a little robot that looks into a folder and checks oh
There is a new application new ecd let’s look inside module one look for the PDF if he finds the PDF see if the PDF um has an XML and if it has an XML take that uh so this is pretty much autom automized and the next big thing would
Be to see all right and what is it now so we need to Det protect the type and the version of this so is it an MAA is it a variation is it a legacy variation is it a new variation with fire um this is now a little bit
Improved because the new Fire versions will start with 2.0 so it will there will not be an overlap because the old ones always had 1. 26 well 24 25 26 and so on and the new ones will have a 2 point something after that um we do need a little bit of
A human interaction that confirms the workflow scenario that should be started through this system but then again we use mainly XML and transformations of that XML to extract the data that we want and we convert this data into Java because Java is the language we’ve been implementing at the agency now for a
Long time and we save this to our database we start workflows we create procedure folders and then we start to involve people actually looking at this from a more validation perspective now we did this in fire version 4.6 and this is still what we have now in production because fire
Version 5 was released well the XML version two which uses fire five was released by EMA uh just a few weeks ago but it is the new stable version now so 4.6 was the version for the last two years and this is the version we’ve been looking at nationally we’ve been testing
And we’ve been implementing uh we’ve been waiting for 5.0 to go further with our implementations because it was always a little bit unstable now we finally do have this version that we do hope we keep for the next few years years which means that it doesn’t matter
If you are on version 4.6 now and if you’re kind of in the same boat as we are or if you haven’t started importing at all um I would say we can recommend to start now and to find this path on how to get from 4.6 or how to get from
Nothing to a 5.0 version there is now some very good documentation so there is a link behind this presentation that you can click once you have that presentation and I’m sure L will share this and let’s just click this um and it will guide you to a page
That will always be the same page this is the EF Byer XML release note page it is a knowledge article where if there is a new version it would just get uh attached to it meaning you bookmark this one and you always always have the newest information that you need in
Order to update your system now as this was a very very large update and as we are aware that not many people are importing this right now but many people are trying to and developing this there is a little package right at the end that you should click and what
It gives you is a download of a zip file that contains everything you need to start this upgrade or to start um your your input routines so it has the data requirements it has open this up for a moment it has examples in 4.6 it has examples in 5.0
It has an HTML representation it has the change log and it has examples for every resource that can be inside uh an application form that is either part of the variation or part of a medicine product so this should give you everything you need to
Start and we can recommend to do this in order to um decrease your administrative burden now the other thing that we can provide is if you are still very new to this topic we did give a few trainings that give you an overview of the structure of
The medical product in fire that gives you an overview of the structure of the procedure envelope so the variation and how it looks like in fire if you don’t want to Dive Right In but just touch well top of this take 10 fields and we’ve did a presentation on
How to get these top 10 Fields what that we think are the most important ones for example we used to know which procedure to start and the very last presentation that we gave is the most advanced one um it tells you if you want to track
Changes and how to see this this is not yet relevant as you can’t actually change a medical product well you can change like very basic things that like ATC code and device but not the entire thing nevertheless the EF is ready to do that so the EF that you receive has an
XML attached this XML is ready to receive any kind of change from a medicinal product and as soon as the user inter faces from EMA can do this as well thef will include those changes and the presentation last one was on how to understand the track changes so these
Presentations they’re all on YouTube they are time coded so you can just jump to the interesting part and Skip all the boring introduction now of course uh look at the introduction and uh that should also give you kind of a good start now we are here to import those
EFS and just so that you believe me that we actually do this and we’re not just talking about yeah you should do it um even though we have a robot that does this today I’m going to be this robot and I’m going into our power system this is how it looks
Like when you start um this is what I have to do but uh we have here uh a little section to create a new procedure now I could just type in which medicinal product I want to do a procedure on or I choose a an
EF and in this case I will just choose an EF and I take um this one now choosing this EF tells the system a couple of things for example it already knows now which products I actually want to change and it will suggest these same
As I would have gotten them if I would have searched them it says all right I have identified these two products in DF and I am assuming that this is a central variation for human please confirm so I could change this if um if it did a
Mistake and then I can say all right this is it now create my procedure now this takes a minute or so which means we’re going to look in the meantime at the fields that we are extracting right now so this is what we have in our system and we have developed
This is not our final uh our final tool but we had to start somewhere and as we don’t actually yet receive National applications this is just for centrally now um we’ve chosen this limited number of fields now the first two are the most important ones the ma number and the ma
State it’s not something that we need it’s something that we need to identify see which product is changing this is exactly what I’ve shown you before the other things are informations that we that we want to show in the procedure folder that we open up so in
The meantime this has opened up and we have ribbons here for every part of the medicinal product and we have one for the uh well we have kind of an an overview as well um for the medal product itself um for the ingredients vet or for application data
There is one then for variation which is probably the most interesting right now because there we want to see the Scopes when I see the procedure type implementation date I mean domain is pretty straightforward uh it’s going to be human for now um type of authorization for now is just cops but
We will receive different ones we take the variation procedure numbers we take the the type of application if it does include a line extension if it’s a work sharing or an IG then the procedure type the ma holder but only if it is for Austria because
There can be more than one ma holder but we’re only interested in the one for um well for our country and for that one we take the email the company um and for kind of more Precision information we take the precise scope for change not yet the the other changes not yet the
Specific changes just the precise scope for change so we can extend this in the future but for now it’s it’s kind of a minimum set then we were interested in free checkboxes which is a summary of the product characteristics if it’s attached the labeling and the package leaflet and we’re showing this
Automatically in the procedure folder where people can not only look at it um for example here you have the SC it’s a it’s a 1B so I can change this now if I think it’s not correct but I could just look at the PDF see right this is well I
Actually don’t have to look at the PDF just look at the procedure folder here I’ll look at the ecd to see yes this is okay and in the end I just say okay it gives my name my date and I say all right I’m done and
Next and this is what we do this is how we extract EF if you’re interested in more specific and technical things I do believe we had um talks and knowledge sessions around this um so this is kind of a summary of it if you have questions ask them in the
Chat G is waiting eagerly to answer them in the meantime I’m going to go to the next topic which is RMS you have one question in the chat Noel yes G feel free to yeah I will I will answer Luke can you enable me to directly
Answer one so that you can see the tree in the chat so that Noel can continue presentation I think you you’re allowed to do so no you don’t no not yet because strangely everybody can can can write in the chat ah no but I uh directly
Answering to one so that I don’t need to reference know what I mean yeah yeah okay good RMS RMS is a well pretty large system by now but even though it has quite a lot of lists at EMA we have even more in Austria list are
What man manage well list kind of manage our our far system before there was far as there were lists and we didn’t want to hard code too much so everything is kind of adaptable you need a new procedure type you create an entry in a
List you need a new uh type of medicine product you create an entry in a list so lists were all very important for us and our data structure is not easy nevertheless it is now necessary to get the IDS that EMA or RMS is using so
We want to match the list that we have with the lists that EMA has and this is not just we can’t just take the list that em em has or RS has um because we need to map them to our terms and many of the terms that are coming now from
RMS we have actually already created a while ago so in order to map this we first try to directly call RMS uh this went quite okay for a while until some unnamed um very friendly person did a few too many calls to RMS and shot the system now they have implemented a fra
Link while they are trying to figure out how to enable RMS for a larger user base this throttling unfortunately prevents us from doing a proper synchronization um which is why we’re using a private well we bought a tool called sporify which is for us kind of a
Pass through API it caches RMS and it allows us to do as many calls as we want uh we not limited to 50 calls per minute the other nice thing about it it has a 99 something per up time and if we look at Spore at the moment it is quite
A bit below that I don’t know if it’s 80% but uh we do experience quite a lot of downtimes as we don’t want them we went through kind of this proxy API now the second bit is we have a lot of national extensions so even though we
Want to use the RMS lists we need our own database and our own extended list um which we have our own system Above This which is in Oracle Apex developed that gives us access uh and the ability to change lists nevertheless once an ID is is on the one Z something term we
Know it’s coming from EMA if we’re using the 900 something for our national ones um the principle that we use to synchronize is uh the modified after uh that if you do synchronize you probably notice you can put a date to RMS and it gives you the changes that happened
Since that date this is quite useful um we do still need to post-process this and there is an algorithm there which is on the next page which I’m going to show you in a minute the last thing is that we have now decided if a new term comes along or if
An update comes along many of them we actually want to look at some of them we don’t so we have this synchronization tool that gives us a list of terms that a person needs to check before they update our own National system if we identify a term where we
Think ours is better uh but we can’t actually submit it as a change request we have the concept of a blacklist meaning an update comes along we don’t want that and the next time it comes along we also don’t want that so we Blacklist the term and it will never again
Synchronize how does this work so we have decided that we’re not doing ad hoc we tried ad hoc synchronizations once um it was okay but it was not necessary it was a bit too much of a strain so we’re not doing this once a week uh it seems like it’s enough and every
Synchronization results into either well an existing term needs an update or there is a new term ID so we’re matching by term ID if you don’t match it should be new but it might not so there is the other possibility that if there is a new
Term uh it could be that the name is exactly the same so if we if the system does match the name all it does is it will up update our local ID with the ID that comes along from RMS and the next time we have the link
Established if we go down the path the left path here where we say all right we do have this ID already so this is a change to our current to our current term there is for one two big options one is a text update and the other one is that it has been
Invalidated the term that we want that we have is no longer valid it is now non-current could be nullified on noncurrent current is uh most likely especially for well I’m going to get there for substances um then we also set our term non-current uh we call them valid two we
Give them a date and we still show them in our systems but you can no longer select them actively so next time you touch it you have to change it to something else or it comes by an EF and then it would not take the term the other option is the text update
For text update we have actually three options uh it can be the name the description or the state state being a different update but for name of descriptions we look at the English and the German ones um saying German ones because we actually have two Germans so let’s spend
A minute to think about the difference between Austria and Germany uh it’s mainly the language we do speak German but we kind of well have our own little words that we really like so there is not just German there is the German which is called
DDE and German which is De a t and our principle is we are fine with using the German Germany term but if there exists a German Austrian term we’re going to take that one because we like the suckle that of well whatever Germany uses and if we don’t have either of them
We’re going to use the English one because well we still understand English now if there is a change in the German one so the German German one we want to check that want to see if this is fine so it comes into a list if there
Is a difference in the English one we’re fine with it we don’t need to check that uh we assume that the English update that comes along has been checked by other people we are not too keen on this we’re using this as backup anyway so we’re just going to update
This and if there is a difference in German Austrian one it’s probably us anyway so we’re going to update this and you can follow this algorithm but you would need to understand German I’m sorry I have not translated this now to look at this I have prepared
The second demo for today and the second demo will put us into a system that is currently being developed right while we we’re actually while we are talking and presenting here um our team is well our team of developers and analysts are actually doing the work behind the
Scenes and changing this constantly so this is not yet finished but it is good enough for me to show you uh it’s starts with a list of not all the lists yet because we’re still working on it but it already has a lot of them um and for each of these lists
We have the English name we have the name that it is called in our own system and we have a flag that we can change if we want to synchronize it or if we don’t want to synchronize it as it already does this on a weekly basis you will see
When the last synchronization has happened how many new terms um there currently are how many updated terms and how many terms there are there would be updated but we don’t want them because they’re on The Blacklist now if we go if we say all right there are nine for the combined
Term there are nine new new terms and seven updates just looking at the time GE how long do I have uh you’re good in time 24 24 minutes yes we good in time so we have time until half past 3 so so we’re going to look at the next
Ribbon yeah so but if you continue that pace I think it should work perfect so now we can see the combined term let’s make this just a bit small too small okay combined term is the color loock it’s called in German it’s called the AR form and uh most of the
Time and there are a lot of RMS term names which are new and we do not have an Austrian term name for it we do get a German one right now and if I want to do some changes to this I can so below that if I select this will
Come um what is coming along from EMA and we should have here now what do we currently have in the a this is not yet implemented right now I can overwrite it but I don’t yet the um what we have here so it should say
This uh so this is how it would work like if this is a new term what I would do is I would say yes uh this is what I want and then I click this little button here and it will um create the term in my
System if I’m unhappy with this term or with this translation I would add it to the Blacklist and I would say just because it is a Wednesday and it will add this specific term ID to my Blacklist so the next time I do a synchron synchronization it will
Not bother me again if this is a new term um there will be another thing that we have already implemented but it is not yet shown in our quality system which is a similarity match so it what it does is it takes usually a Leen Stein algorithm and which calculates the
Distance in letters inside a word and it looks for similar terms because quite often a term comes along and we actually already have that term we just have a nine something number for it so what I can do then it will give me a similarity search I can click on the similar term
And it then um assigns a new term ID to that term and that’s pretty much it I can filter this on new or on change terms so I could say uh just a new terms place um I could do a filter on on a date these are the new
Terms uh so some materials that we currently don’t have for example um there are also some status changes here that I have well this is actually a new term that is coming along that is all that is non-current so I don’t know if we’re going to take this but it is good to
Know now Blacklist is currently this is a quality system it has the exact the entry that I just made um I am no 25 because I was 25 when I started at the Austrian agency Look at me now uh history and log is not yet fil uh implemented but what
We want is a run history where we have the logging if the run is successful or it failed and a log if somebody did create a term it will lock it here what we did though is a test ribbon and this is quite a fun bit so what we can do
Here is ask our testers or developers to give us some pretty examples um because well we have to test this quite extensively and this if you do want to implement a system like this testing it is not easy because you need updates from RMS and you can’t just ask RMS to send you
Updates so what we’re doing here is we’re simulating updates so we have this own little test system for it where you can either rever a catalog because you’ve messed it up as a tester uh or you can say well I just want to simulate I have this um this catalog here with
This list ID and I’m going to add a Jaguar a test jaguar and um it would I would say start simulation and what it does it will give me a new Jaguar add a jaguar and then I can go to my term checks and see if I can find the Jaguar
Probably no I cannot so let’s see if I can find myself course I cannot this is what happens if you do a live demo from time to time um let’s search by term ID here’s my Jaguar it’s test jaguar in English um so you do need users who know how to
Use the system and then it will find them uh so this test Jaguar would be simulating a new species and I could accept this new species or I could say no no just speak I know I know you love RMS synchronization no is no worries but I
Think now we need to go to something that you even love more the substance substance integration of course thank you for looking at the time I do run into these um these nice pretty demos for too long substances uh if you have looked at thect list or SMS list that is currently
There you have you will find more than 75 I think it’s 76,500 something substances which is way too much it is very messy there are a lot of duplicates so the positive thing is it is a complete list which is the reason why we want it we want the IDS we want
The complete list we are not happy about the duplicates that we have there our national list in faros has about 20,000 substances so we know we’re missing a few and we want them but we also don’t want the other 50, duplicate that probably exist in
SMS so how do we tackle this um for one not every substance that we have in our system is going to be used in our life cycle management because what we do is we create substance Master folders so just because it is a term doesn’t mean
We are actually using this term in the ingredients so the ingredients have to be selected from a substance master phold folder and this master folder has more information than just a uh a substance in a substance list it has well synonyms they are there as well but it has information on doping narcotics
Psychotropics and drug interactions pharmacovigilance issu had in the past different IDs from different systems uh the type the base substance calculations and the documentations like as smfs now we synchronize SMS with a few principles first principle we handle SMS like any other RMS list just because it
Has its own API doesn’t mean it’s something special I mean of course it is but you don’t have to treat it that way second we only synchronize substances on demand so we don’t want to have the 7,000 substances in our 20,000 list we are going to take them whenever we need
Them if we receive an EF with with a substance we don’t have we’re going to go through our RMS sync tool going back to now and the substance will be here most likely and we just select it and we say yes this is the one we
Want um we use the field commment from SMS so SMS has a very neat field uh called comment where they put put if they put a a substance non-current they’re going to write in there why they did it and what is the substance that um well most likely because of it’s a
Duplicate what is the other substance and the substance ID will be in the common field uh we checked this last week and it actually is there uh we’re going to take a few new Fields because they’re neat um not just the the substance name Source because
Well we need that already and we have that already but we also want to have the type we have that but slightly different so we’re going to try to synchronize this we’re going to get the weight and the formula and I’m going to show you right in a minute why we’re taking the
Formula uh last but not least the synchronization will help us a little bit because it will also tell us if this uh substance is being used already in our system because that means we need to update our Master folders um so most of these checks they are
There because we are expecting a lot of mergers and substances in the next few years from EMA and we need to handle those non-current terms that come along now yeah so I’ve mentioned before we have this um this formula that you get here now the formulas uh look like
This here in the title WM o u Etc this is how you encode a substance structure now what you can do with this is you could query pcam pcam is a public register with all chemicals and it has all substances so you take this uh this
Little value here and you use pop cam by looking at the compound and then you put this string value and then you choose do you want to have a 3D structure or a 2d structure if you choose 3D um it will show you the image of this
Structure which is very neat um if you don’t want to burden your system you can do the same thing with 2D um and the other nice side effect what it does is it takes this string as an input and it tells you also the name of this so if for whatever reason EMA
Made a mistake SMS is not correct they give you this huge list of substances but um so this is the SMS list um probably know that you can take I can pick any any string here at random this is will always be here if it is a
Substance and you put this inside this link and you can show this inside your own system and it will show you the image but it will also show you the name and if you translate this this one is the Polish one but if you look at the
English one it is the exact same name as in SMS so this is a neat little add-on um if you want you can you can just take this little piece of code that I’ve attached to the PowerPoint and show it as an iframe an iframe is something
You can show your web page in your own procedure system or if you’re pretty more modern you can use a rest service to get this same image no maybe to add to this um one’s EU SRS so they European substance registration system would provide such a
Function then of course we would link to this e SS system but currently this is a good good idea in the meantime yeah so I’ve been using pcam since I was a little pop uh we learned that as UN in University of bionformatics this is the
System that you go to because it’s well it’s the most complete existing system in the world uh of course um it has everything every chemical so we are only focused on on specific ones and there usrs is of course more focused and would be probably a better
System so last topic for today a bit of an Outlook a bit of taking whatever we learned right now importing fire and trying to do it ourselves so we are providing to our from well to our ealth communities to pharmacies um in a in an indirect way but we people kind of rely
On the data that we give them and this is medicine product data so we thought about well let’s create our own fire version 5 export and this is how you would start doing this so at hl7 orire you will find the current 5.0 version what we do is we’ve start with the data
Requirements from the electronic application form because it has the links to all well it has all the fields that are available uh you then choose the fields that you want and it has the fire paths in 5 so there is not a lot of mapping that you need to do there is no
Ambigu ambiguity um you can take these uh they will also match the next version of the IG that will be published so you can take the U IG as well but you will find a little bit more information in here Point number four would be to add your National requirements now we our
Eof community wants for example the PN number for every package this is a national number that identifies every package in Austria and we have to include this now this of course you will not find in any EU lists so you have to create an extension in fire extensions
Are how you get information in that you kind of come up with yourself or your country comes up with them and I have two examples here one is the PSN I’ve mentioned uh where you do have to kind of find the name for this extension and
For that one we took the one that exists so we get into touch with our communities in Austria there is the ho7 fire Austra community and you see all right is there something already that is called like this and if not you come up with your own extensions and there you use
Your own URL you say extension and you put the name of the attribute you want and then you create this using well we use Java so the only option that we have is Happy fire um which is the library to use uh to create a Json or an XML fire
Message uh currently this is how it looks like uh which is clam uh which is a system from Germany developed um it is a very old system it is okay but um we thought well let’s use the uh well it has one downside you can see if you have five substances it’s
It’s encoded you have substance one two 3 four Etc you can’t it’s not object oriented you can’t just say okay I have n substances there is a limit uh so this is just a list so why did we go there we’ve EMA decided to use fire as a messaging
Format which is kind of the main driver for this uh eal is using it um apple is using it I don’t know if this is a uh an show of hands that it is something that is going to stay there for a while in this domain we are doing now this prototype
Uh in farus with uh the R5 format and we’re giving this to the ealth organization that is responsible for the electronic um health records in Austria and they’re going to look at it and see if this is usable for them now fire is object oriented which is well for one
The good thing you can do a lot more complex things the downside may be ealth doesn’t need those complex things so we we had a talk with them and we gave them this this is the structure that we have now look at all the beautiful things we
Can do with a medicine product and then we went through it and we said well actually we don’t need that part and this part is not of interest for us and there we don’t need this so we crossed off everything that eal was not interested in and we remained with a
Very short list so this is the Prototype that we’ll do now um the nice thing is it is object oriented we will give them the concepts of manufactured items and administ products which some ealth people don’t even know about because they don’t need it they just want to
Know the active substance and that’s it but we tell them well if you manufacture something can be an active substance in one and the other manufactured item can be different in the adal product so maybe this is information that you should be interested in and they said
Yes give this to us and we said all right let’s do this and that’s it thanks no so much well done I think thank you um N I I propose you you keep sharing your screen because we have only one more slide and I think I can do this
Via your screen sharing um so let’s come to the conclusions and this is the last slide and I’m proud of all of you because we are still 106 people so I hope you enjoyed what we presented uh please next slide no yes so first of all I really need to
Say thanks to Sonia Kik SOA she didn’t she didn’t speak in this presentation but she helped us to let’s say find find information that you might be interested into and gave so much business input about this so thanks a lot Sonia uh thanks Noel for spending so
Much time as well and colleagues I really hope that we increase your motivation to work on IDM Concepts we we we really hope that you have seen that this can be done and it helps to bring us the regulatory Network into this data driven future because we think data is
The key for data driven decisions and for many many use cases that we have uh around and so this is really helpful and Unicom and IDM sorry idmp did we just lose G I’m afraid so you might take it over take on message well I can take it over
Um so I think what G wanted to say is it took a while to understand all of these Concepts and it is not something you do from today to tomorrow it’s a marathon there are all these little Parts they’re not just guideline improvements and fire versions there is the Improvement of your own
System um if you look just at the standard you need to understand it you do ux improvements based on this like the thing we did with admin product manufactured items where I think this is a very huge help for the people that you date don’t have too much
Info um yes I’m back I’m sorry for this hi K I tried to take over one last thing I wanted to say I think the best thing that we did is like is the the manufactured item unit and having this the same for all ingredients it just makes sense there is
A a ux Improvement that is not just ux it’s data Improvement that we did there just because idmp tells well this is the option doesn’t mean you should do this um so if you look at the recording go back to that section uh this is one of
The best things I think we did thank you sorry for my connection problem this happens when you’re not at home office so we have finally three take home messages so the first one is or when you do this we all know in unicor it’s not a Sprint it’s a marathon
And we also are now aware that we will have future or we expect additional effort in the future there will be updated guidelines from EMA there will be updated fire versions uh ux improvements will be raised by by our users data enrichment will be uh will
Have to be implemented once the EF will will be prepared for this the sync with PMS so maybe you mentioned we we uh talked about RMS SMS and OMS but not about pm so we’re planning to download PMS data especially for centralized products improvements on Ms will come
And we need to finalize the mapping to all uh to EUR European terms and get rid of national ones second day call message maybe you have the same situation as we are um maybe the structure is more enhanced than the data enrichment so we are expecting more data when from the application forms
And as no it’s for you again from the application forms and we are evaluating the PMS API to download cups I’m saying evaluate because um PMS is still on a very old fire version we’ll see if we want to use this and also getting an account from PMS will take a while
Um they’re going live with the read API in may only um so right now the access that we have is very limited um so this is the concept that we need to look at y are you back yes yes I’m back so and F yes and finally I
Think uh the idmp language is perfectly for the communication inside the network inside the regulatory Network maybe it’s too complex for some our data consumers outside and that’s why we have have this prototype and maybe this is something you also can consider doing your present your implementations colleagues it’s
Half past free we are finished on time thanks a lot for joining us that long as I said before this will be recorded So other interested people can watch it on YouTube I I never watch because I can’t see me in video thanks Luke for preparing this recording
And goodbye from Austria and please if you have questions if question contact contact us and it was a pleasure to be able to present our outcomes in this meeting thanks a lot back to you Luke thank you Gil and I would like to give just the floor to CHR uh to finish to
Announce the next uh knowledge transfer webinar CHR are you with us yes thank you Luke I’m here thank you gell for a very extensive presentation it’s very good to have it on YouTube actually there a lot of information you still have some maybe unanswered question in the chat
Maybe you can do that afterwards uh but so thank you thanks a lot from the austrialian colleagues as well Sonia and others and I just want to announce that the the best practice session of the knowledge sharing the last one actually uh during the Unicom project is the
Second of April and it’s Portugal and I think the time is 10:00 but Luke will surely send an invitation to you all to to participate on the last one because every NCA has done at least two uh best practice session and knowledge sharing session so Portugal is the last in line
For this project at least so thank you thanks a lot all of you thanks for your active questions in the chat as well and see you soon again yes for all the questions that I didn’t answer please drop an email to me I will uh reply in a written manner
Thank you I will help you in that also thanks L make sure to make sure you don’t we don’t miss anything uh just a last communication so uh the project is slowly but surely getting to an a first end hopefully something will continue after it uh but uh we we have we will
Have a final event in Brussels on the 26th of uh April 25th and 26 but 26th will be open to everybody so you will also receive uh very soon now uh the the information about that that event that we want to be also to be a high level
Event so meaning trying to try to engage all the people who have a say and a decision making to to to make things happen quicker and better in the future still better and still quicker uh so you are most welcome uh you will receive um an information on this soon uh with if
You can of course the the the link to the registration uh as it’s a free of course event uh and um we will be most happy to see you in Brussels if you can that’s it for me and bye bye Luke bye bye Christa everybody have a very nice day hope you
Will join tomorrow at 10 o’clock here at the meeting 10 tomorrow yes I will be there I will be in free meeting