Transformation in Trials

Exploring the Impact of Open Source Software on the Pharmaceutical Industry with Katja Glass

September 20, 2023 Sam Parnell & Ivanna Rosendal Season 4 Episode 4
Transformation in Trials
Exploring the Impact of Open Source Software on the Pharmaceutical Industry with Katja Glass
Show Notes Transcript Chapter Markers

Prepare to uncover the transformative potential of open source software in the pharmaceutical industry! Our distinguished guest, Katja Glass, an influential open source ambassador, joins us on this enlightening journey. We'll explore the rising popularity of open source solutions in the pharmaceutical industry, its myriad benefits, and the profound impact of sharing tools, processes, and software.

Turn up the volume as we unravel the advantageous features of the Open Study Builder. This unique tool has not only helped pharma companies in embracing automation but also presents an innovative approach towards modern metadata repositories. We take a closer look at how its  properties facilitate creating protocols, bringing a refreshing twist to the traditional processes.

As we conclude, we ponder over the future prospects of the Open Study Builder, the implications and the exciting opportunities for tech vendors to engage with the Open Study Builder ecosystem. Katja shares her perspectives on the possible transformations. Tune in as we draw parallels between the pharmaceutical industry and other sectors and explore what lessons can be learned from their uptake of open source. So, buckle up and get ready for an intriguing conversation revolving around the changing dynamics of trials in the pharma industry!

Guest:
Katja Glass


________
Reach out to Sam Parnell and Ivanna Rosendal

Join the conversation on our LinkedIn page

Speaker 1:

You're listening to Transformation in Trials. Welcome to Transformation in Trials. This is a podcast exploring all things transformational in clinical trials. Everything is off limits on the show and we will have guests from the whole spectrum of the clinical trials community, and we're your hosts, ivana and Sam.

Speaker 2:

Hello everybody and welcome to another episode of the Transformation in Trials podcast, and today we're delighted to be joined by Katja Glass, who is an open source ambassador within the pharmaceutical industry. Katja, really great to have you on the show.

Speaker 3:

Thanks, sam and Ivana, to be here.

Speaker 2:

Katja, as you well understand, as we discussed, we try and base each episode of the Transformation in Trials podcast around a specific theme or topic and, given your interests and where your passion lies, we thought that today the topic and episode theme would be the promise of open source software for the pharma industry and beyond. So with that, we're going to learn a lot about you and your background. But let's get going over the first question. So we have a broad church of listeners to this podcast, so, to set the scene, perhaps you could kick us off by describing what is open source and how does this differ from non-open source or other licensing models?

Speaker 3:

Yes. So open source is the way that you can provide software for free, but not only for free, but mainly that you can get access to the source code. So the main purpose of open source is that you can access the code, you can use it as you want, you can change it as you want and very important you can also share as you want. So this is very different of classical licensing models, so where you buy a software, you can use a software and that's it. So with open source, you are really able to update the software, to learn from it how it's coded and also to share it.

Speaker 2:

Sounds pretty cool, but for the non-technical listeners? What does that mean to them in reality? What are the benefits here? I'm sure all of the technical listeners already know. But for those of us who aren't as technically minded, why should we care about open source?

Speaker 3:

Yes. So open source really enables a very broad independent interoperability software development. So instead of having a company which is developing software for a specific purpose and you really can have much more use cases implemented, for example, you do have a broad community and if a software is used by different organizations or people, it's really typically meant to be developed on very generic purposes. So it's much more open. It really allows also on a high level of collaboration, so on the user side. So if you're non-technical, you might see that you can download software and use software for free, even though you might not necessarily have a maintenance model, like if you're using open-word or you're using Microsoft Word.

Speaker 1:

And that's a good example. I'm curious for applications in the clinical space, because what I've encountered so far is that, as you probably also have, especially in statistical programming, there are some licensed software that is dominating the market. But then what I've experienced, especially in later years, is as we enroll younger generations to our pharmaceutical companies, who are used to working with open source, we suddenly also get these open source offshoots of software that can do maybe similar things. Have you also experienced?

Speaker 3:

that trend, definitely so. This is also what we see throughout all the conferences. So whereas traditionally we had a lot of presentations about SAS, now and more and more is upcoming from the R universe. So we do have our studio, which became positive, and there's really in the R environment and in the R world there's a lot of progress, specifically also in our pharma industry, in going into this direction. For example, lately the pharma world has been founded. Also very popular is by now the Admiral package, which gains a lot of attraction because we do have two very big pharma companies really creating open source.

Speaker 3:

But R is not the only application. So in R you can share scripts and it's comparable to SAS macros and also huge frameworks. But really outside this scripting environment there's much more we can do. For example, there are two persons who provided the visual defined XML editor. So if you need to edit the defined XML you can really do it in a graphical user interface. So that's pretty cool and there are also very many nice new small tools. Also the big organizations like CDISC are coming up with open source solutions, for example the core project. And very cool is the Novo Nordisk open study builder project, which is a complete metadata repository or study definition repository.

Speaker 1:

That does sound very cool. That's a big company like that would venture into trying to create a repository like that.

Speaker 2:

What's been the capitalist catcher for open source really taking off? And it might just be because I've been exposed to open source recently at conferences that I feel like it's really kind of taking off. But you tell me, is this a new phenomenon or is it something that's been ongoing for some time and is just getting some momentum of late?

Speaker 3:

It's as usual in our pharma industry it's gaining momentum quite late.

Speaker 3:

So what we also saw, for example, when we see there had been this graph, topics and linked data, which it really started small and slowly and then really, when you just have people taking it up, it will grow, and then really immediately, which is great, because I mean CEDIS has taken this up with the Cosa initiative. So the CEDIS open source Alliance, few say had a working group about open source also since quite some year, but the importance is really growing and for this you can see, for example, that the Fuse organization even created a director for the open source space. So it's really an increasing and growing thing which really needs a lot of mind shifting. So it's really do we really share what we can do? There's a lot of. I mean, we only share our processes and tools. Specifically, on the conferences, we had been talking a lot and showed a lot and we also created working groups, and now sharing, next to just the ideas, also the software itself is the next level, but as usual in pharma, it takes quite some time.

Speaker 2:

Does validation come into play here? Because often that cited as the reason why things are slow or tricky to adopt when it comes to new technology. Is that a consideration?

Speaker 3:

Validation is typically a consideration, but I mean what you experience if you have bought software. I mean you can buy software and it can be buggy as well, and then you have to the disadvantage that you really have to wait to the vendor corrects it. But on the positive side, you do have a vendor you can contact for For open source. It's really depending on who is developing and maintaining this tool. So if you figure out a bug, but at least you always have the chance to either do it by yourself or find someone who's familiar with this approach.

Speaker 3:

And as what we see, for example and there are even now starting up new companies which are really relying on open source models so when they release open source tools they live on the fact that there will be change requests coming in or education requirements or release stuff or validation support. So this is how they can make money with. But yes, validation is a topic. But really see how, how sure you can be that the vendor has done appropriate validation. Of course we do have our SOPs, but small issues and special situations we experience always.

Speaker 2:

Got it Right. If I capture what you said correctly, you've got like this whole support and services industry that is wrapping around this open source movement as well, Right?

Speaker 3:

That's currently not too strong. That's currently in the R space. There are two or three bigger ones. We do have quite some smaller ones, but I see it growing. But if you're looking outside of R, the market is still very slow or very small, which is also because the interest I mean if no pharma company would pay a vendor to create open source solutions, then they wouldn't do open source solutions.

Speaker 2:

Understood and I want to move the conversation on to really the reason why I asked you to come on the show. I was at a conference, the CEDIS conference, and there were a few exhibitors there, but one booth stood out in terms of the amount of people that were bustling around this booth trying to understand what was going on. And that was the booth that you were effectively at, where you were talking about the open source study builder project, and I know you touched upon this during a previous question, but to me it sounds just from a booth attendance standpoint. It looked like this is a seriously hot topic in our industry, so maybe you can talk us through open study builder project from the vantage point of somebody that and we have many listeners here somebody that hasn't been exposed to this before.

Speaker 3:

Yeah, so the booth was from the C-Disk open source Allianz and we do have presented their multiple open source projects and the open study builder is really a very super huge, super big project which has a lot of functionality and capability. Many colleagues followed the C-Disk 360 project, so this is also the base for the open study builder In that project. It's also a collaborative project with many participants. We really would like to have end-to-end automation, starting from study protocol to data collection, to data evaluation. So creating SCTM standards, creating ADEM standards, creating the final TLS and also all the authority documentation, the publications. So if we could just streamline more information and have it supported by really this end-to-end streamline. That's the ultimate goal and an outcome of this C-Disk 360 project was also the Transcelerate Digital Dataflow project, which is aiming for the same, and also the open study builder.

Speaker 3:

And Nuvo Nordus created the open study builder because they need to get rid of their old LADIC system, so they do have to switch to a new metadata repository. But they couldn't find a solution fitting really good to what they want to have and for this they decided okay, we want to go with C-Disk 360. We don't want to go with the Transcelerate DDF, but we don't want to wait another two years, and for this they started to develop it. But their intention is also not to create another custom solution. What they really would like to have, they would like to have a system which is growing over time and which is also modernizing over time, and not just by them but by anybody. And for this they decided to go open source which really allows continuously development, and when really I mean by now it's only Nuvo Nordus who's doing all the development and also the promotion of this tool, of this open source project. But when later on it comes on, different companies will have different use cases and if you can embed it all into one tool, you would have a lot of opportunities. And actually it's meant also I mean, the metadata repository is just storage in a graph model in this case which contains the metadata, and you can do anything with metadata and you can enhance the metadata in any way. And you can also connect very different tools, like you can connect your EDC system, you can connect your clinical evaluation environment, so the connected with the clinical data, that you can do automatic evaluations if you are having metadata driven programming, for example. So there's a lot of opportunity and I have to admit the open study builder is very complex because it has already so many functionality in there, but still we're still in the front processes.

Speaker 3:

So the open study builder it's a metadata repository and study definition repository, starting with protocol data. So whatever you need for information in the protocol, you can manage all the structure information into the open study builder and then all things which can be derived out of this. So for example, there's also like standard for Fs. You can manage in there Also all the trial domains you need for your clinical studies. You I mean the trial domains are really an outcome of the protocol. So if you have the protocol in a structured way, you will get the trial domains just out of it quite straight. So that's a pretty cool thing.

Speaker 2:

Sounds pretty cool to me and what like it may be obvious to a lot of people on the call, but what are the real sort of outcomes that pharma companies can expect to achieve if they embrace greater automation, with what this open source study builder initiative should be able to offer?

Speaker 3:

So they could benefit to really also automate processes or to harmonize tools. I mean, one of the main goals of the DDF project, for example, is also to have standard interfaces and the open study builder will provide also the same interfaces. So if you will have created tools like EDC systems or dashboards or protocol tools which are using these standard interfaces, it doesn't matter whether you use this metadata repository called open study builder or from an alternative vendor. So it's really the idea is, the more we invest in using these standardized interfaces and when we participate in providing also additional tools, it will be a hub where you can just it's like, it's like an app store. So you do have your heart, which is a metadata repository, and then you plug in your EDC tool, you plug in your if you have automated metadata driven evaluations this is something you could plug in. You can create protocol automations or vice versa. If you already have protocols, you can get the information out there and just feed in the open study builder to get, for example, your trial domains out.

Speaker 1:

I am curious about this protocol builder piece of the whole workflow. Sometimes ago I was involved in a company that was trying to rethink their protocol and also enable a higher degree of reuse, but also designing the protocol so it's ready to collect the data that we need afterwards, and there was a higher degree of coherence and I had no idea that someone was already working on this as a collaboration between our industry. But could you tell me more, maybe, about that piece? What does it entail? What is the aspiration?

Speaker 3:

So really the inspiration is I mean, traditionally we do have Word documents with a protocol and we do have all core information in there, and then whatever following up processes we have, they need to read and interpret the protocol, which is really difficult. And if you have different protocols, even for similar studies, you are using different objective endpoints, for example, so you don't have standards there. So and this is really the idea Okay, we're not thinking about having a Word document and just derive the information from this, out of this, which is what we do now, more or less. It's really it's also a training issue. So it's meant that the medical writers, in this case, are entering the protocol information into this tool. This tool will create, for example, an automatic protocol flowchart for you and it will also create the activities and this you can just import in your protocol. So Novo Nordus decided to still go with Word as major interface for medical writing to finally create the report, but the core information from the report are coming through an interface from this metadata repository.

Speaker 1:

And that's also a big chunk of the way that, even if we use text as an interface, that we can transform it into something that's more usable.

Speaker 3:

My dream was always having everything a C disk annotators from the very beginning, even in the protocol, but we're inching our ways towards that and also one way we have in the Open Study Builder is also a lot about templating, For example. We do have objectives, objective templates, so you can have, like you have, a comparator versus treatment and so you can have a generic text and you can reuse this text in different protocols and just exchange what it is inside which really allows for really having very similar protocols.

Speaker 2:

So, Tatsuya, one of the reasons why in the past or historically, people have probably gone to a vendor for software is because the vendor might have an ongoing program to produce new releases in the software. They take on board all of the maintenance, hosting, etc. How does that work in a world where you've got an open source project? Who takes that on?

Speaker 3:

Currently, I mean Novo Nordic is a pharma company, so it's no IT company, so a lot of implementation is actually done by vendors. So and they I mean with the current technology and the current agile way of working we are having a lot of pipelines. So there is an automatic process that when you do changes, automatically test scripts are running and if the test scripts are running and everything is fine, you can just click the release and then the pipelines are started, that's, a new version is rolled out. So currently these pipelines and processes are, of course, set up internally, so within their hosted environment everything runs smoothly. For the open source release it's that way. Whenever they do an internal release, then there's a parallel pipeline to push the code onto the public environment, which will also run To compile.

Speaker 3:

The tool itself is very, very great because we can use Docker environment. So you just need one command line code and everything will just, yeah, install itself on your local machine, which is great. Of course, it can also be installed on a server and, if you would be interested in automatic like a cloud subscription and so on, we also have a vendor. For example, our database is Neo4j database, and they would also offer services, or they do offer services to host it for anybody. So they also and this is one of the reasons why they joined this open source project because they hope to get also additional customers. Kaja, what kind of technology?

Speaker 1:

is this open source builder created on.

Speaker 3:

This is, I mean internally. It's using also another lot of open source tools. So the open study builder, the database itself is Neo4j graph database, so we do have like biomedical concepts, so really it's a graph model. Then we do have the API, which is a standard interface, how you can get the data. This is created in Python. Then we do have the application itself, which is a web application. It's using the Vue framework and then there will be soon a word plug-in or add-on available which is be written in Visual Basic for applications, and the DDF API adapter is even created in Java. So it has a lot of things. But you really can see specifically the word add-in or also the DDF API adapter. It's like different components. You don't need them, but it's a nice. It's just a growing system. You don't have to use all of it, you just use what you need. But yes, we do have already quite a lot of programming languages included in this project.

Speaker 2:

Can I ask a question? You may. Yeah, this might be a slightly awkward one, but what happens if, like for arguments taking new CEO of NovinOrthis comes in and goes? Why are we doing this? Let's change direction. We don't want to do this anymore. Does Open Study Builder still exist or does the whole project go to waste?

Speaker 3:

So whatever is released as open source will stay as open source, so everybody is able to take it up. There could be a vendor company, there could be a pharma company, there could be a coalition, like really a community. The long-term idea is also that, when we get enough collaborators, that this tool is also not at some point in the future not managed by NovinOrthis but by a community. But until then you do have to have a lot of collaborators and really a management company, organization, whatever.

Speaker 2:

And what about the tech vendors then? As part of the wider ecosystem, is there opportunities to integrate with what is being done with Open Study Builder? Should we be viewing this as a collaboration opportunity rather than a threat?

Speaker 3:

Definitely. I mean, the scope of the Open Study Builder is still metadata repository and study definition repository, so there's still a lot of need for connecting tools. Or also like, for example, we do have a Neo4j database, but then Microsoft or another company would like to exchange it with their own graph database. So this is things they could do. If you're a vendor, like you want to show your SCE, so your computing environment, and would like to show some automatic cases to connect to a metadata repository, you can just go and just use it for demonstration purposes. If you're using the standard interfaces, like which is from the DDF project, you cannot only connect to the Open Study Builder. You will also be able to connect to any other tool which is supporting the DDF standard. I mean, another important aspect is when really other pharma companies coming into this and would like to use it, they definitely would need support. I mean, either they do have an internal IT or, as what we typically do in pharma, we just hire vendors for this.

Speaker 1:

That was actually my next question. How does this change the dynamics for the internal IT organization in a pharmaceutical company?

Speaker 3:

Yes. So either you do have an internal IT which then gets the additional tasks I mean, if you would like to apply it internally, you have to host it somewhere, you have to implement some kind of security, so the access write management there's a lot of things you have to still do or you might also want to do some validation, scripting and so on, because also the SOPs with respect to validation differ from company to company. So there's always these things you have to consider. But it's similar as you have now. By now you will get installation instructions which you follow With an open source project. It really depends on how much information is available and maybe so we are working on the documentation and we would be very grateful if others will also join in. So if you think or if you investigate, you would need these additional steps, for example, to create authentication, so that we can get those guides on.

Speaker 2:

Yeah, sounds like it. Do you have any perspectives on how other industries have embraced open source catchers, and is there anything that our industry could maybe learn from other industries, regulated or otherwise?

Speaker 3:

I mean what we see, for example, in the web development industry. This is a super fast industry, so and they are mainly using only open source projects. So you can not only you can use plain HTML files, but by now you really have huge frameworks which do everything for you, and I mean also like the R Shiny, it's also like a web framework. You can just plug in whatever you are used to and it will deploy a web page for you. So this is much quicker and you will have a wide variety by having this industry.

Speaker 2:

But I think that's a good kind of segue into learning a bit more about you. Catch it, so you've obviously got a big passion on this topic. Tell us how you came to be passionate about this. I'm pretty sure you didn't just wake up one day and think I love open source. There must be some background here, so tell our listeners a bit about yourself.

Speaker 3:

Yeah. So I started quite some years ago already in the pharmaceutical industry and in the technical area and I quite quickly joined the fuse organization, with first going to the conferences and on the second one I already presented and the atmosphere was so great just to talk to people around about the real issues and the real, yeah, problems we have, about the processes, and I mean I also saw these working groups and I joined these working groups and I really enjoy this. But I really always see that okay then. Okay, but let's not only talk about the processes and anything. The next step would really be the open source towards, and I also very love very much coding, so I would.

Speaker 3:

I thought about, okay, let's, let's do some open source projects.

Speaker 3:

But then, when I made myself independent with the scope, I quickly figured out I wanted to create something which was related or similar to the visual defining some editor, and then I found out, oh, there is already an open source tool.

Speaker 3:

So so one of the main reasons I started now with open source portal for clinical study evaluations is that you can have one site with information. Okay, which open source solutions do we even have? Because what I also see is that open source developers, specifically if you're one or two open source open source projects, they don't. They don't have a strong voice specifically at conferences and say it's hard to participate. If you're like like I'm also just doing part time open source, if I go to a big conference with those high costs, it's a little bit strange. And for this I see many open source tools are not promoted enough and for this I'm very happy that, maybe because I whispered some words into the right ears, so the cause I has been found. I'm super excited that in our everything is going smoothly on anyway. So this is, I don't need to put my energy in there and I'm super heavy happy that also no one orders decided to really do this open as a project the open study builder, which is really a super big project as open source.

Speaker 2:

It sounds like you're heavily involved in open study builder. That's going very well. What does the future hold for open source in our industry, and maybe you can give us a bit about what your vision is for the next few years?

Speaker 3:

I think that we definitely will go very much forward in the art universe, because I really see a lot of companies looking into the details, specifically as we are sharing complex scripting and processing, like we have with Admiral, and there's I mean Admiral is from STTM to Adam mapping, but there's also a new initiative which is called oak, which goes from C dash to STTM, so it really will cover the complete flow and and the interest is really high and the acceptance is high.

Speaker 3:

So we will definitely see an increasing our usage of open source tools and also our open source tools, and I hope that we also will gain more contributors for the open study builder, because then we really would also see more and more use cases. For example, there's a preclinical group dealing with protocol automation and for them it would be valuable as well to use just the principles and the processes and the database from the open study builder, and we already created a small proof of concept. So it's it's working. It's just there needs to be also bigger pharma companies investing and looking into this, this opportunity, which is really a very good one. But because it's such a huge project, it's also for the pharma companies difficult whether they want to start with a small use case, or whether they want to stay with their own processes.

Speaker 2:

Well, hopefully our podcast can help get that message out to lots of pharma companies and align with what you're thinking, that as we start to wrap up the podcast culture. We normally ask what we always ask our guests the same question, and that is if you had a magic wand and can change one thing about our industry, what would that be and why?

Speaker 3:

I would love to see, specifically, the big pharma industries to become more and more open minded, for pharma to really support their employees. When they want to publish anything, open scripts, to do so, but also to enable their vendors so the contractors, the other oath that they to allow them to create their scripts and everything to open source. And I really would love to see really investments investments in having people doing open source or investments in hiring vendors to create open source.

Speaker 2:

I think that's a very honorable wish and it's going to go out to the masses, so let's see if it comes true. Perfect, all right. So I'm sure there's going to be lots of people who want to learn more about yourself, more, learn more about open study builder. Where is the best place for them to connect with you or learn more?

Speaker 3:

So of course I do have an homepage. So if you're googling for Katja glass, you will find my homepage, which is glaucombeeu, and there you also find the link to the open source portal. And the open study builder also has open study builder dot com or org. I don't just Google open study builder, I think it's dot com and yeah, there will be all information and I'm available in slack and linked in so people can reach out.

Speaker 2:

Perfect, and I can attest to Katja post some great content on linked in, so go and follow her there as well. Awesome. Hey, katja really enjoyed this discussion. I think our listeners are going to take a lot from it as well. Looking forward to seeing what happens with open study builder. We'll keep a close eye on that, but thank you very much for coming on this podcast and we will be in touch with you, see.

Speaker 3:

And thanks for inviting me and it was a really pleasure and a very great talk, thanks.

Speaker 1:

And get the episodes hot off the editor.

Open Source Software in Pharma Industry
Benefits of Open Source Study Builder
Future of Open Study Builder
Open Study Builder Discussion With Katja