 Hwych gweithio. Mae gynnyddau a gwahanol y ffyrdd pan oedd hynny yw'r Eisteddf fel hynny'r llans y byddai, ac yn oed yn dda i'r enw i rhai gwybodaeth gyda'r ddatblygu Gwlinnol, i gyhoeddoddau trwy gweithio i gyd-lwg, sefydig i, bengylch, cysylltu, youron yng Nghymru. Fy hoeddi'r cyfan! Yn mynd i'r adwyng iddyn nhw, yn roi'r cyfniadau i'r gytigol, I have a quote here. Social and emotional well-being is the foundation of physical and mental health for Indigenous Australians. It's a holistic concept that encompasses the importance of connection to land, culture, spirituality and ancestry and how these affect the well-being of the individual and of the community. But as a group, Indigenous Australians continue to experience poorer health outcomes than non-Indigenous Australians in some areas. Throughout our work with Hysander and the People Research Data Commons, we've heard consistent feedback that this gap is an issue. It's an issue that Australian researchers see as a priority to address and that enabling culturally appropriate and equitable research is foundational. To the design of research infrastructure and that empowering communities with their own data allows them to take charge of addressing priority issues of health and well-being. That's one thing to think about as we go into this afternoon's discussion. So I'm Rosie Hicks. I'm the CEO of the Australian Research Data Commons, ARDC, which is enabled by the federal government's national collaborative research infrastructure strategy, or ENCRIS. Our mission to accelerate research and innovation by driving excellence in the creation, analysis and retention of high quality data assets. And I'm delighted to welcome you all to the second of the 2023 ARDC leadership fora. And as you know, I hope you're in the right place to discuss considerations for trusted research environments for health and medical data. Excuse me, I'm just going to do this. I've been given my speaking notes. There we go. I'm not pretending. Virtual research environments that allow collaboration between researchers across geographical, institutional and potentially jurisdictional boundaries. Secure research environments that provide additional technological controls to help ensure the platforms that researchers use working on their data are safe and protected from unintended use. But trusted research environments go beyond this. And they help to ensure that the expectations of all research stakeholders are met. This means that not only can the service provider have confidence in the usability and security of the technology for research, but the data custodians and the people whose data is being accessed have confidence there is robust governance, an oversight of how their data is used, and that this kind of research infrastructure meets their legal and ethical and in fact cultural expectations. Be safe data spaces, address the needs of the public sector, the private sector, the research sector and civil society. They allow researchers to work on priority health and well-being issues for the health system and for health consumers. ARDC's People Research Data Commons provides national scale data infrastructure for health and biomedical research and research translation. You will know the health and health research system is a complex multi-stakeholder environment with multiple government jurisdictions, a myriad of health services, a thriving private sector, several national research facilities and a wide range of research institutions. I believe that would go under the description of a wicked problem. So today's forum is really looking at how we can simplify, transform the way the researchers contribute to this health system. The data commons builds national level capability, coherence and coordination in data strategy and discovery, trusted research environments, data integration and advanced analytics. We look forward to catalyzing the public debate, envisaging for this system a more coherent framework for trusted research environments across those many players in the health system. Today's discussions will inform our own national infrastructure programmes and partnerships. I would like to recognise there are a number of organisations that have developed solutions in this space and are working further to meet the evolving needs of researchers, including importantly our fellow ENCRIS capability PHRN, which supports a number of data linkage units around the country and environments in this space, including bioplatforms Australia, research organisations, research infrastructure organisations and commonwealth and state-based health agencies. I suspect that all of you will belong to one of those on the list. Let me know if I've missed you out. We're committed to supporting Australia's researchers using trusted data environments to address the grand challenges. So now I am going to hand over to our facilitator for today, Joanna Batston. Joanna is the inaugural director of the Monash Future Data... I'm going to get this bit right. The inaugural director of the Monash Data Futures Institute, responsible for bringing together data science and AI capabilities from across Monash University. An exceptional thought leader in the development and application of AI and data analytics, a passionate person for the benefits of AI in driving lasting and transformative change for social good. But particularly I'm very grateful for Joanna's contributions to the ARDC research advisory committee, research and technology advisory committee. It's a committee that has truly partnered with us as we go through evolving the thematic research data commons and Joanna's wisdom has been of great benefit to us. I know it's only a tiny part of the many many hats you wear Joanna, but we're very grateful and I'm delighted to welcome you here to guide us through this afternoon's conversation. Thank you. Thank you Rosie for those kind words of introduction and it's great to see everybody here in the room and I know we have a large audience online as well, so welcome. Public trust in science is essential in a range of fields of enquiries and not least those where we work with personal data. In the fields of medical and health sciences, the development of innovation and new knowledge generally requires us to assess accurate, reliable, and at times very sensitive information. The ARDC is seeking to address these challenges of data privacy and data security by exploring trusted research environments for health and medical data. Trusted research environments often called TREs, another TLA, three-letter acronym for us to remember. These TREs are highly secure and controlled computing environments that allow researchers to gain access to data in a safe way. They're also often called safe data havens or secure data environments. TREs are not entirely new though. In the United Kingdom, TREs have been used to share sensitive data since at least 2013 and more recently we've seen governments around the world use TREs as a secure long-term solution for research and clinical needs and clinical use of sensitive health data. And within Australia's national collaborative research infrastructure, the Population Health Research Network has developed secure access facilities with both state and federal governments to provide linkage of population administrative data sets. At the Commonwealth level, the new data access scheme provided by the Office of the National Data Commissioner includes notions of trusted data environments provided by accredited data service providers. These trusted research environments are also implicitly a baseline capability of research institutions and researchers to be accredited users, including probably many of you in the room as well. Several solutions are now emerging across Australia and internationally to meet these needs. However, the needs of researchers are also changing and evolving to incorporate new technologies and new ways of conducting research. In my role as the director of the Monash Data Futures Institute, we're particularly interested in the application of new technologies, particularly data science and artificial intelligence, as it relates to the way that we think about access and analytics associated with data. And more broadly at Monash, we're very focused on looking at the opportunities with the School of Public Health and Preventative Medicine and other groups across Monash, including our e-research teams, our Monash Helix technology platform teams, to think about how we protect and secure sensitive research data. Ensuring health data can be analyzed in a secure, trusted research environment is critical to enable us to maintain public trust in the research process while enabling researchers to bring together a range of data types from a variety of different data sources. I'm delighted to be able to facilitate a fascinating panel of speakers today. We have speakers from a wide range of perspectives, including from industry, from a platform provider, from researchers who use research platforms. And I'm sure you'll enjoy a very lively discussion on trusted research environments this afternoon to understand what's already in place and what are the emerging trends and directions in this field. So before I introduce our panelists and ask them to come to the stage, I'd encourage all of you, as you listen to this afternoon's discussion, to think through how would you address answers to these types of questions. What do you see as the opportunities and challenges presented for trusted research environments? What do you think are the best practices today in trusted research environments? And interestingly, how can we facilitate patient involvement and input to develop trusted research environments that address their research priorities and interests as well? So we're in for a very lively discussion. In terms of the format for the South Tynun's event, each panelist has been asked to speak for about five minutes, introducing their own perspectives and experience and the work that they do today in their respective organisations to look at trusted research environments. And so I'd like to invite our four panellists to come up and join me on the stage now. Dr Charmaine Tan, Professor Dougie Boyle, Professor Jim Buttery and Associate Professor Bernie Pope, if you would all come on up and take a seat on the stage here and we'll get into the panel discussion. So I'm going to introduce our first panellist and then we'll go one by one and I'll introduce each one of them in more detail as they start. So our first panellist will be Dr Charmaine Tan and Charmaine represents the industry voice in today's panel discussion so we're delighted to have her join us today. Charmaine is a senior consultant in the ecosystem business unit at Telstra Health and within that role, she oversees an aggregated data and insights research programme to help address population health and research needs. Prior to joining Telstra Health, Charmaine led a clinical analytics team that worked with millions of routinely collected electronic medical records within a metropolitan health service and she also spent 10 years working in health and medical research and so she's very passionate about safe and judicious use of routinely collected medical records to drive health outcomes. So Charmaine, looking forward to hearing your insights. Thank you for joining us. And thank you for the introduction, Joanna. Perhaps before I start, just perhaps to give you a bit of a context of Telstra Health as an organisation and the types of work that we do. So Telstra Health is Australia's largest digital health company. Over the past 10 years or so, we've acquired and also developed a range of patient management software, spanning hospitals, pharmacies, residential aged care facilities, as well as primary care. We also run a number of core health services on behalf of the Australian federal government so we also run the National Cancer Screening Registry, 1800RESPECT, and also an electronic prescription exchange. So that's kind of a flavour of some of the software that we provide across the patient care continuum. Just to give you an idea of the scale of the interactions that occur through our software in primary care alone. So each year we process about 80 million consultations through primary care. Three out of four of those interactions will result in a prescription, one in four will result in a pathology or an imaging request, and one in five will result in a specialist referral. So that's the kind of magnitude of data that we're dealing with just in primary care alone. The ecosystem business year that I'm part of, we work with partners across the whole health ecosystem that are beyond our customer base. And so we partner with researchers, farmer, government, whoever is interested in leveraging the reach of the information within our platforms, and we partner with them to drive their own health, their own missions and their own health outcomes. So we do that in a number of different areas. We do that in the patient engagement space. So whether you're interested in looking at the appointment book or payment providers, we do that in the clinical communication space. So looking at secure message delivery between from point to point, between clinicians. We do it also for clinical trials and health programs. So we identify cohorts of patients that are eligible for different trials or health programs and serve them electronic clinical decision support in the software in order to reach those patients. And we also have this de-identified aggregated data program that we look after as well. So that's just to give you an idea of the types of capability that we do. And I guess you might be wondering why is Tulsa Health interested in participating in a panel on secure research data environments. And I think traditionally there's been these silos between academia, health services and industry. And having worked in these different areas myself, I can see the opportunities and the challenges in each of these areas. And I think the more opportunities we have to come together and bring together these different perspectives, the more we're able to keep on continuing to provide a safe quality healthcare system for Australians. So I'm really happy to share this perspective on what we do in the data space for Tulsa Health. Thank you. I'm not sure if this microphone's on. Great. Thank you. Thank you very much. So our next panelist is Professor Dougie Boyle. And Dougie's a professor and academic specialist at the University of Melbourne with a specialty in health data science. He has a leadership position as well within chancellery around digital research academic engagement. And he's the director of the Department of General Practice and Primary Care Health and Biomedical Research Information Technology Unit. Dougie's role on this panel today is representing one of those of a platform provider. And so within the large health informatics team, he is the data steward of a repository of de-identified research data, holding data on over 5 million Australians. And so I'm sure we'll hear a lot more about that this afternoon. So thank you, Dougie, for joining us. Thank you, Joanna. Well, look, it's great to be here. As you say, today I'm representing a platform provider. Some of the things I do get involved in are around access to the research data from a stewardship capacity and things like that. And hence, when I sort of think of myself as a platform provider, it's often around that much wider engagement that we're needing right across universities for instance with industry. So it's great to have you here and be on the panel today with you. So, look, I really got involved in providing platforms really out of what I see as an increasing need here in Australia, privacy, consent, governance, cyber security. All these things are more and more becoming important, not just within small niche research communities and industry. This is about people, and that's hugely important. And in fact, the definition of what is sensitive and sensitive data is an interesting one because certainly the data that I often work with is de-identified, but when you're dealing with millions of records, you know, there are huge sensitivities still on all of that. And I want to see how we can maximise the use of data for research. You know, we actually need to undertake research and we need to be able to see translation happening on that to improve outcomes for all. We have great challenges there with the number of breaches that are happening in different organisations, which is happening not because organisations are necessarily lax, but there's an awful lot of more resources being on the dark web and things. It's actually very hard to secure data these days. So it's really important that we can try and take that burden off of researchers who don't have skills in that space. They're there to answer research questions, not to keep an environment safe. So we've got a bit of a transition to do there and we've got things like the ONDC as was mentioned earlier on, looking at access to government data in particular, which is changing university roles potentially in this space. And we'll come back to that and I'm sure I'll discuss it some more. So anyway, that's me. I'm here to try and see how we can provide the platforms and I've worked around the provision of Erika, a platform that we've developed ourselves in our department which is based on Nectar infrastructure. And the university has got a separate secure research environment that has been launched in the last few months that some of my colleagues are actually here in the audience for that today and are doing a great job. So looking forward to the rest of the conversation. Thank you. Thank you very much, Daggy. So our next panelist is on the panel to represent an academic user of research data. So Professor Jim Buttery. Thank you for joining us, Jim. Jim is the chair of Child Health Informatics at the University of Melbourne and also head of the Informatics Research Group at the Murdoch Children's Research Institute. Jim's research interests include vaccine safety, infectious diseases epidemiology and clinical child health informatics. He's also the chief clinical research information officer at the Royal Children's Hospital. He's a co-director of the Global Vaccine Data Network and serves as a member of the advisory committee on vaccines for the therapeutic goods administration. So thanks for joining us today, Jim. Thanks very much, Joanna. I work at the Center for Health Analytics, which is a Melbourne Children's Campus initiative. Melbourne Children's Campus is a combination of the University of Melbourne, the Royal Children's Hospital, Murdoch Children's Research Institute and also the Royal Children's Hospital Foundation. And our mission is quite simple, it's to improve child health through the innovative use of data, which gives us a fairly broad remit. My, our work, as you've heard, given that we formed during COVID was very much public health related rather than very specifically hospital related and we're also responsible for vaccine safety surveillance in the state of Victoria, which is all ages. I might take a step back before I talk about trusted research environments. I'm an infectious disease doctor and I became interested in health data when I was sitting on a government committee, the Pharmaceutical Benefits Advisory Committee. And PBAC makes recommendations to government about whether to fund on policy, whether or not a medicine or a vaccine is a good investment for the Australian taxpayer to support. And submission after submission in good faith showed absolutely terrible amount of quality of the Australian burden of the particular illness that this medicine was designed to improve. Not because the people doing the submission were doing a terrible job but because that data was just not available. And I knew that that data was sitting there and if only could be made available. And trusted research environments such as Duggies are the research environments that suddenly mean that normal interventions in health care, we can begin to understand either the burden of an illness or the difference a medicine may take at a real population level rather than a relatively artificial clinical trial level that occurs in a limited number of people with very set exclusion and inclusion criteria. So it's really offered a way for us to really understand the impacts on our health system. So we use trusted research environments in a variety of ways. One, we use them in collaboration with the state government where we're able to link major statewide data sets to understand the interaction between vaccines and health outcomes. We also use it in collaboration with external partners either private industry or other organisations which enables us to go in and access large-scale de-identified data and ask questions of it in a secure way which gives those organisations the ability to feel confident that their data is safe by giving us access to near population level data. So for us that represents one, a privilege that we're able to trust it but two, excitement that we can now get access to the level of data that enables us to inform real decisions. Thanks very much. Great. Thank you very much, Jim. Our next panelist will give us some opening commentary is Associate Professor Bernie Pope. Bernie, thanks for joining us. Bernie's from the University of Melbourne and his background is in Computer Science and Bioinformatics and his research focuses on applying computational techniques to biological questions, especially related to human genomics and cancer. He's the lead bioinformatician at Melbourne Bioinformatics and Associate Director for the Human Genome Informatics at the Australian Biocommons. And so I'm sure we'll hear some interesting insights from a genomic data consideration perspective from Bernie. Thank you very much for joining us today. Thanks, Joanna. It's a pleasure to be here and thanks to ARDC for inviting me. So I kind of have two roles. On the one hand, I'm a researcher at the University and I'm interested in cancer genomics, particularly using genomics to identify the drivers for aggressive disease. And on the other hand, I work for the Australian Biocommons where we try to develop digital research infrastructure at a national scale connected to international efforts. And in our team in the Human Genome Informatics team, we're focused on genomics and more generally, omics technologies related to human data. And I think the first thing to point out with omics data, of course, is that or for one thing, it's incredibly large per sample. An individual whole genome sequence is approximately per individual about 100 gigabytes. And now there are projects with hundreds of thousands of those. So if you do the maths, you work out it's petabytes or exabytes of data just for a project. So it's big and expensive. And it's also very complicated. So that's one thing. Another thing about omics data is that it's very identifying, particularly if you're looking at DNA for an individual that's unique. Their sequence of their genome is unique to that individual. It's the one thing. It also connects to their relatives. So if you learn things about an individual, you also happen to know some things about their kin as well. And also omics data brings with it all sorts of fantastic opportunities for healthcare and improving translational medicine. There are also some risks and things we need to be concerned with, including the potential for discrimination. And that could come out in many different ways. It could be around access to insurance. It could be about access to employment. Or it could be discrimination on other levels. It could also relate to kinship as I mentioned already. So if you identify information about an individual, you happen then to potentially know things about their relatives. As well. And there's all sorts of issues around that that we need to be cautious with. And so in our work at the Australian Biocommons when we're looking into digital research infrastructure, I have to say that trusted research environments are not something that we are advanced in Australia. But it's on our radar and it's something we're very interested in looking into. So I wanted to highlight probably the most advanced use of trusted research environments in the world, which is through Genomics England. And if you look up Genomics England, you'll see they have about 100,000 human genomes in their data set. That consists of about 30,000 cancer-related genomes and about 90,000 rare disease genomes. And on about 90,000 of those individuals they have detailed clinical and demographic data. With their trusted research environment, no personal identifying information can leave the system. So there's an airlock mechanism whereby researchers can log in, do their analysis, and then what they take out of the system is downstream analysis results, which is carefully curated. Thinking about the advantages of such a trusted research environment, that there are many advantages. One is that it provides a single point of control for the security arrangements for technology so we don't have a very wide surface area to look after and essentially a single point of access into the system. Another major advantage I think particularly with demonstrated by Genomics England is that it actually enables research on very rich genomic and linked demographic and clinical data, probably more so than you would normally expect in the research environment. Because of those controls around how that linked clinical and demographic data is used, it enables actually detailed linkage across the genomic data, the clinical, phenotypic demographic data for the individual participants and also public health records, which is quite an interesting feature and makes it very appealing from the point of view of a researcher. I think it also, as Rosie and Joanna said, at the beginning, it encourages public trust in the use of this data and it also gives control for individuals about their data that's used in the system and also enables them to understand how it's being used. And I think if we looked at the Genomics England platform as an exemplar, I think it's a really excellent example of how we might be able to do this in the human genomic space. It also points to an interesting fact about it, which is it's a very large dataset, which means that it's worth investing and it's quite presumably quite an expensive platform to develop and so on, but it's very well worth investing in for such a large dataset. Maybe later on in the discussions we'll talk about some challenges around that as well and so how do you connect, how does a Genomics England and other things happening around the world, how would they, how would you do an analysis across these systems or mention that now, but that's something that's on our mind as well. So thank you very much. Thanks very much, Bernie. So you can tell we've got some really interesting perspectives and backgrounds and different representation on the panel here. So I'm going to kick off with some questions to get us started in the next phase of Q&A, but I'd encourage you to think about questions you may have for the panel and I know we've got an online audience as well, so please think about questions if you're online as well and the team will feed them up to us. So I'd like to kick off, each one of you has talked about both challenges and opportunities and often when we think about trusted research environments, we might leap immediately to what technology do we need, but I'm actually interested in stepping back and sort of like to get your perspectives certainly for those of you who are platform providers as well as platform users, what do you think are the key challenges in delivering these trusted research environments? Is it the technology or is it the people or is it something else? What are some of the challenges that you have to address on a daily basis? Maybe I'll start with you, Dougie. All right, I'll kick that one off. Look, I think of course there are human challenges out there, but in terms of actually delivering a platform for safe research, there's a couple of interesting observations here. There isn't really a fixed definition of what a secure research environment or a trusted research environment actually is. So I think there's often for as a consideration of what technology we're going to have to put in place, how far do we take it and usually what it really means is you want researchers to be able to work within an environment where they cannot get the data out except through a very controlled manner. In fact, I can't get the data in either and there might be audit trails in there. But if you think about research on where the responsibilities lie, I'd like to just focus on that for a minute. Because I get involved in human research a lot and we have to put in grant applications according to the NHMRC and follow the guidelines and the principle putting the ethics application is it has certain responsibilities and even legal responsibilities and the universities have responsibilities. So the interesting thing is if you're a platform provider what's your responsibility in that? So that's a kind of an interesting one because if something goes wrong and the data does get out how do you escalate that for instance? So some of our challenges are of course trying to put the firewalls and the protections around the data and the standard operating procedures as well as the cyber security and the penetration testing but for me one of the other challenges there is then making sure that we have sufficiently robust processes in place and open processes to deal with things when things go wrong. So that's the interesting thing is that we think we're building secure environments and we do still a human angle in there and we need to appreciate that this isn't always perfect and it might be as simple as someone giving access to the wrong person to data in a research trial sort of thing that's easily done important things of the transparency around the process after that. It's also something where you need a certain scale of organisation to try and undertake maintaining these you need to be able to make sure you've not got a single point of a single point dependency on delivering aspects of this technology if it's all going to fail where do you go with that? So our projects our enclave we've had approximately 30-35 projects go through so we we've got an awful lot to think about in terms of cyber security and our university think about the university's reputational risk so that then flows down on to the providers of the environment they're having to deal with the university wanting to ensure the protections are there but funding envelopes are limited as well. So I'll leave it there there's lots of other perspectives I would think. Thanks Duggy My other panel members wanted to chime in maybe Jim you're a user of a platform. Do you share those perspectives on some of the key challenges or do you see other things at play here? I think everything Duggy says is correct as users and especially early on in periods where we needed to access linked data for public health research they're almost uniformly as users of data rather than platform providers the longest time is actually building the relationship with the providers of the data so that they trust you enough to actually provide access. The first data linkage project I was involved with we got an ARC grant for three years and the federal government approved the use of the linked data two years after the grant finished my personal record is six years from initiating contact with an independent industry partner I went through three managing directors three chief medical officers and after I gave up the notification the data was now available arrived in my inbox and it just takes time and patience and providing all the information and you have to be aware that most organisations are not there to do research and you just need to carefully build that relationship make sure you're answering all their queries in a way that is helpful for them to justify to potentially sceptical board members everything else is certainly a challenge but that's probably the biggest real challenge is building that relationship over time and maintaining that trust Thank you you talked about trust in relationships and you made reference there to industry players Charmaine, you're our industry representative on this panel from a perspective of Telstra health what are some of the considerations that you need to think about when you're trying to build those trust relationships with researchers that you're trying to serve I guess those points that Jim raised were exactly it even in industry we need to build those trusted relationships and they do take time and we have security protocols that we go through but it is really the time to reach comfort between two different parties about for using this information in particular I guess just to make the point with the health and with the data that is captured within our software there are many stakeholders that are involved the patient, the clinician the practice, the pathology provider or the imaging provider and there's us as software vendors so as software vendors we're responsible for securely storing that information on behalf of our customers but we ourselves don't have right abuse of that information for secondary use cases the trust that our customers have in storing that software and not using that information is incredibly high we don't want to jeopardise that however we recognise that there are needs in the broader health ecosystem to derive insight from that data and so some of the use cases that Jim described when he was on the committee asking the prevalence of certain conditions in the population we have developed a consent and governance model to answer those types of population health requirements some of the ways that we address those types of questions is we have a data governance model some of the principles of that are we capture patient and clinician consent so in our patient management systems we store consent for my health record alongside other similar programmes and in that we capture the patient and clinician consent for use of the data for secondary purposes and that's to provide transparency for our customers we also collect dynamic consent every year so we serve a clinical decision support message to clinicians to actively consent into a de-identified aggregated data programme and we keep that up to date every year and also we provide the insights from that research back to the clinician so that they see what that information is being used for we also only provide de-identified aggregated data and I know that limits some of the use cases that we've talked about here but I guess what we've found is we do that for safety we don't want to provide we don't want that risk of disclosing information personal, de-identified or not information and so our comfort level is aggregated de-identified data as well we also have a clinical governance unit who review every single use case that we do to make sure that those use cases are aligned to our mission and vision as an organisation which is to enable connected healthcare for digital healthcare for all Australians so having said that we've got this very robust model and we want to partner with people who do require insights and who want to be able to action those insights within the software so there's some of the considerations that we have as an industry provider I guess not everyone is interested in storing that sensitive information or they have the analytics resources to analyse it so we provide that service for clients who are interested in those questions and the answers to those questions and then also when you want to feedback that insight but we have the software where you can action those types of those findings that you're finding in research or otherwise Thanks very much and so I think what we're all hearing is different approaches different solutions different technologies that can be used as we think about building these trusted research environments and so what I'm hearing is it's probably not a one size fits all and each one of you has different experiences and the types of technology solutions that you're involved in but you're all focused on how do you address data governance security, privacy trust Bernie you made reference a couple of times to the United Kingdom and I wondered if you could comment on what you see are perhaps best practices based on what you've seen in the United Kingdom and what you're involved in from a research perspective here in Australia as you think about this particularly from the genomics perspective is there anything you can say about best practices or what you think has worked really well in the United Kingdom and that we should aspire to emulate here in Australia? Yeah it's a good question so I was thinking again we see trusted research environments as part of an ecosystem of technologies and I think one of the great successes that Genomics England have had is their connection to the broader community internationally there's an organisation called the Global Alliance for Genomics and Health or GA4GH which has something in the order of 900 participating organisations in individuals it's very large it's the community of human genomics and omics researchers getting together and saying we need to standardise the things that we're doing we need to work together we don't want to build things in isolation and that includes interfaces how these systems talk to each other and so I think Genomics England has excelled in that regard another thing I was going to say somewhat related to the previous question but also I think that something that they would have had to solve there is in the genomics space it's very apparent to us that it's an extraordinary number of areas of expertise that need to come together in order to build these kind of systems if you think about it we have the participants who donate their biosamples and their information most importantly we have the clinicians who are involved with some cases where patients are involved we have people who are experts in the legal, ethical and social aspects we have people who are experts in digital infrastructure computing, high performance, computing, algorithms, data analytics statisticians, mathematics we have all of the work that goes into project management all of the work that goes into funding and organising these things it's extraordinary in my mind all the different disciplines and expertise have to come together for these things and so I'm an outsider to genomics England I'm only in some sense a user of their system but from what I can see I think that thinking about how we might do that in Australia one of the big challenges is thinking about how do we bring all of these areas of expertise together there isn't anybody that covers all of those things in the level of expertise across all of them in the same way of course that's never going to occur so we need to have those groups of people coming together to work together and I think that's probably the big challenge for us Thanks very much We have time for some questions from the audience I wondered if anyone has a question or a challenge from their own perspective in their own organisation a roving microphone that we'll be wandering around so if you'd like to put up your hand and ask a question Thank you, thanks Joanna this is a question from our online audience and I've got two questions here in fact but I'll take one of them first and the question is how to ensure a platform is developed or delivered where the data is hosted and has a big portfolio of tools to analyse the data now that might be something for the platform providers or researchers Dargi, do you want to tackle this one first? Yeah, a thorny one I think it's something that providers of secure environments always struggle over are the sets of tools within the environment because one of the key things here is you by definition within these environments you don't have access to the internet and one of the challenges is even some of the stats packages out there that are in common use don't have mechanisms for licence sign up that don't involve connecting to the internet so there's some really fundamental problems so certainly we see it within our environments a lot that we continually have to think about how can we take the software builds try and get them running within it there has been some advocacy I guess from platform providers and through projects like ARDCs ventures around Erika to try and gain more of a group that can approach things like software providers to ask them to change the way the licensing models work as an example but I think this just requires us to have more of a mature sector here in this space to help drive that forward but you can imagine if more and more research starts happening within secure environments then vendors of software are going to think more about secure environments because that's where the tools to be used so just now it's just an ongoing problem certainly once you've built an image with lots of different standard pieces of software on it you can often just spawn these images for new instances of a secure environment the question is just where the licensing models would allow that to happen when you spawn more of them, depends what you've licensed and then the question is who pays then for the software that someone actually wants to use on there so you'll find some platforms will say these are the limits in the software where by default would offer and there could be challenges if researchers want to use something with a sort of standard portfolio Thanks Anyone else want to weigh in on the question I guess I can speak as a user of the platforms I think very similar to Dougie the limitation is largely around the software that's available within that and our experience largely has been that the platform holders and I quite frankly was not aware of the licensing internet issue is that they're generally happy to install the software you want so long as you can pay for the license arrangement and the reality is as a researcher if you've got a member of your team who's comfortable with one particular statistics or software it's actually cheaper to install that software than it is to train them to use an alternate version just because the opportunity cost of all of that time in learning etc and so one we try and get all of our team comfortable with using one particular software usually are but similarly if we need to bring in software most of the platform providers as long as you're happy to fund it are able to provide that within the environment and that includes business intelligence software and other visualisation software Can I just add to that Jim sometimes researchers are wanting the data itself in certain formats like SQL or Comma delimited so sometimes actually just the format in which the data is requested can cause some challenges within the environments as well I totally agree we have one data partner who we're reliant on relatively near real time data to do the work that we do and that data is available in near real time through a particular BI software if we wanted a SQL update which is easier for our team to work with it's a lot of work and expense for the data provider to do that and so we tend to compromise by accepting the limitations of viewing the near real time data through their BI software which is harder to utilise our statistics program too Thank you, Bernie I can comment from the point of view of a bioinformatics researcher as well in cancer and human genomics typically the analyses we do sort of on the raw data especially when talking about hundreds or thousands of individuals we could be hundreds of thousands of CPU hours maybe a million or more CPU hours of analysis on that data which creates its own kind of challenges around orchestrating that because over at large scale things break all the time and that's just a fact of statistics and so we need to it's not just running a tool or a program it's running a whole collection of tools a big flow chart if you like over a long period of time in addition to that if we're doing cross cohort analysis one of the challenges with these large data sets is not introducing batch effects by accident by doing things in slightly different ways if you analyse it slightly differently over here and slightly over there is the difference that you see an actual biological difference or is it actually just an artifact of doing things in a slightly different way and so there's a lot of pressure to actually use a very controlled exact versions of software run in exactly the same way in addition to the software and all the versions every tool has dozens of parameters usually those parameters are highly tuned as well so there's a lot of expertise and knowledge comes into bringing those parameters onto the system and so on so really what I'm trying to say is that one of the challenges faced potentially with trusted research environments is actually facilitating that kind of workflow in a highly secure environment so getting both what the researchers want and need to do in order to get their results and have reliable results but also not break the confines as Dougie is also mentioning of the actual research platform as well and the security You've got another comment Dougie I can see It's just kind of interesting certainly in that genomics space Bernie just given the scale a lot of Australian genomics is at the national level because it's difficult for an individual organisation to try and undertake the research so I guess there's slightly different challenges there and if there are secure environments around genomics maybe that needs done more at that sort of national collaborative level whereas secure environments for other medical research forms a researcher often at the institutional level whether it's government departments or the universities Absolutely that's one of our motivations at Australian Biocommons is to have a national programme for this work and I will mention importantly at the same time Australian genomics is running a programme called the National Approach for Genome Information Management and they've submitted through community consultation a large set of recommendations to doing exactly this kind of work and that's gone to the Department of Health for consideration now so it is worth saying that there's a lot going on at the moment trying to assemble this because we recognise it as you said Dougie a national and actually an international thing that we need to solve Great we have another question here we've got time for one question before the tea break so if the roving might can come down here thank you Thank you Dougie Bernie just picking up on what you discussed right here a few seconds ago quite a lot of use cases around the world now picked up by the government by Ministry of Health on secondary use of data what's happening in Asia, Singapore just picked up by one of the third party vendor what's happening in Finland and other countries Do you see the same roadmap a path might take place here where considering that it's required to bring a lot of different resources people with a lot of different skills considering the undertake and considering the benefits which we'll be taking from the program like that under the umbrella of secondary use of data or utilisation of big volume of data the government will pick up ownership of developing such an environment because look what's currently happening there are a number of spaces where needs to be recognised and a template made and some useful platforms are under consideration being developed but the scale is not there yet so what's your thought will it happen will be interested ministry level to pick that challenge Anyone want to like to offer an opinion on I can start on that I think it's Natasha is it one of our challenges I guess is our Australian governance our government's model but something exciting I guess is the latest ONDC data place a mechanism for trying to look at access to government data that's under the data and transparency act that's come through what's interesting about that is that normally government departments look after their data and don't allow it to go out anywhere else but they can provide some environments to access data so here in Victoria we've got the centre of Victorian data linkages and vaults so what's just changed the things just a little bit is that with this new act there's a role for universities in being a safe pair of hands around data we still need to see exactly how that plays out in early days so what I would say is at least we don't have a static environment I think things are changing but it doesn't seem to me that governments looking to centralise the capability but it's looking to certainly regulate and have mechanisms to provide a capability so I'm thinking that this is something that could be important as we do things like industry engagement because when Charmaine and Telstra there's a problem there about who do you actually trust so what does that trust mean so under data place they've got concept of users and that might mean a university as a safe user of government data as an example but there's also the concept of accredited data service providers, ADSP and that would be for instance a university providing additional services like region bringing the data together but under an umbrella of the fact that they've gone through an accreditation process to reach that level so I'm sort of interested what that potentially could mean in the future to allow us to have mechanisms to allow that wider access to data and the knowledge that there is some sort of accreditation and well government has put it in place so hopefully we can make use of that in the years to come Thanks Dougie we're going to take a tea break but before we break I'm going to encourage all of you again to think about your questions for when we come back when we come back I'd like to pick up on this discussion about our universities providing a safe pair of hands for data what about industry do we trust industry with our data and what does that partnership need to look like with our data and how can we get access to data so as we come back after the tea break I'd like to pick up on this discussion around data how do we get it, how do we manage it how do we store it how do we protect it and what is that interplay between academia, government and industry as we think about trusted research environments so thank you so far for your participation and engagement we're going to give you a short break now and reconvene at 320 so I mentioned before we broke for tea that I'd like to come back to this question around trust and trusted data and Dougie right before the break you made reference to universities and the way that we potentially can think about universities as a trusted safe pair of hands to share and manage and protect data industry though is chopping at the bit to be able to enable and provide access to data to store data so I want to continue this conversation around the perspective of industry of a university data provider around data and some of the challenge as well as then providing them within the secure environments that are slightly different models so I think I'll leave it there a service then perhaps you could trust it but maybe another challenge is that sometimes the environments are coming from departments or faculties rather than central university so there's certainly consideration there thanks so Chamin I'd like to turn back to you to get a bit of a perspective from industry both from your experience at Telstra but also what you hear as you interact with other industry providers and I know one of the questions we've talked a lot about is around what sort of reassurance to corporations need with respect to data privacy and is there a risk of competition? Yeah so trust is paramount at Telstra Health and that is trust from our customers that is the top priority trust where their trusted partner to store the sensitive patient medical records so that's paramount to us and in terms of being able to access that information for secondary use I mentioned that we have this data governance policy where we actively request clinician consent and patient consent for taking part in de-identified aggregated research and also that they've got the control in that consent process transparency with being able to have that consent requested every year of them so we actively know that they want to participate and then our use case is going through the clinical governance unit only providing de-identified aggregated data having said that when we do that and when we do this transparent consent and governance model when we ask the clinicians we have a 10% opt-in rate for participating in that kind of research so I think that kind of tells you that's kind of the appetite of what our customers are responding back to us when a trusted partner is asking would you like to participate in de-identified aggregated research for these types of secondary purposes so I guess that's kind of what we're working with and we continue I think that's actually sounds low but it's actually a fair reflection of where we're at in terms of society and trust in using that information for other purposes so just to clarify you said the opt-in rate is 10% that's for clinician researchers yep or clinicians who are using the platforms I know we've got some questions online coming in so Keith if I can hand over to you for the next question so one question here is do large cloud providers have a role and how does that impact trust from the public in their involvement in the research i.e. them donating their data to such health data repositories interesting question cloud providers how do you all feel about cloud providers what role do they play I know we had some discussion earlier around nectar research cloud what about commercial cloud providers Bernie go ahead comment on that we absolutely think that commercial cloud providers can play an important role in the ecosystem that we've mentioned before so we think of trusted research environments as part of an ecosystem that would include commercial cloud would also include institutional high performance computing resources as well I should mention relating back to the previous question about industry partnerships and I've been talking about genomics England partly because it's the stand out example for trusted research environments in human genomics their cloud platform is provided by a commercial partner and I won't mention names you can look it up if you're interested to see what they do and in one sense you can see the ability to build those systems coming from industry as well they have a lot of expertise and capability to do that how we kind of make sure that everything's going as we expect I think is really related to something that's been mentioned a few times by many of the panel members which is the governance structure that's built on top of this and that's really critical and to get it right it's difficult to do and it does relate to also the previous question about the role of universities and other public institutions for one I think an important thing that universities provide in similar institutions is they have longevity they're here for the infinite time hopefully and that is great because you can build long running institutional governance structures on top of those and so that's important they're also open to scrutiny and they should be seeking scrutiny from the public as well and I think that's a useful thing too but having said that given that we can establish those governance structures we in Australian Biocommons and more generally people in the community and the proposals that have gone to the Department of Health through Najem have advocated for the use of commercial cloud as one of the options in the ecosystem that we're trying to build Yeah, Jim, you want to comment on that too? Thanks, you know I completely agree with Bernie in that it's inevitable that many of the large cloud providers have the sophistication that enables the existence of some platforms but there are some critical limitations and it comes down to the governance for example the largest electronic health record provider Epic which is based in Wisconsin has a massive shared de-identified data facility called Cosmos which carries the records of about 176 million individuals overwhelmingly North American no Australian organisation will take part in Cosmos because the data warehouse is not located within Australia and so the upside is that we're certainly protecting our patient's data the downside is that we're not capable of asking queries of this massive data set and obviously the solution is to create a solution that enables respect of Australia's privacy laws to then potentially be able to take part in it but the governance is critical but the opportunities are also concrete but the other thing I just wanted to say that we've rightly pointed out in the last two questions what do people think and what is the impression and so the involvement of consumers is critical in all of this process and the reason I mentioned is that on our consumer panels the perception of our consumer advocates is that the consumers are several steps ahead of where we as researchers or health professionals think they are and it's critical that we involve them actively in this process so that we're appropriately having their views represented in not only in how we consider these things but also in how we design things. A question over here from the room thank you. Thank you just Emma Ball from Illumina wearing the industry lens and also thinking a little bit offshore there are a number of tech bio companies that are now or have been for some time developing their own proprietary data sets very large ones so companies like Recursion Tempus I think they've got 600,000 genomes 23 and me I think they've got over 13 million genotypes matched clinical phenotypes so what can we learn from some of these industry players is their best practice are their caution retails how do we feel about that sort of data being in private hands It's done the silence up here on the panel so I think it's an interesting question and I'm very interested in the industry perspective here it's not just universities that are aggregating data hospitals have huge amounts of data I know we've got some of the Monash folks in the room here who work with a lot of health data through Monash partners we've got industry players that you've referenced that are building value added data sets for me the question becomes again back to trust do we trust a private company how do we trust a private company as an individual how do companies trust each other and then the question around data operability and linkage between data sets becomes a big big question we've just heard why perhaps Australians aren't using the epic data set from Wisconsin I think we'll get some of those same questions around how do we trust integration with a private company's integrated data set but this is the way of the future we will see curated valuable data sets being built and I'm interested in the panel's view around how do you imagine a future where you are looking to work with and integrate some of these data sets that are not within the tight control of your own institutions trusted research environment so we'd like to chime in first Bernie and then Dougie here it's a fantastic question and it's complicated I think it's very appropriate that we're in the building that houses the law school at University of Melbourne because one thing that came to my mind immediately was how different jurisdictions in the world are dealing with this probably the most prominent example we know of is the GDPR in Europe and if you look into the GDPR and I'm by the way no expert on it but one of the key principles is informed consent for use of personal data that is written in plain language and also the ability to for participants to control the use of that data going forward so that they can remove the data from studies and so on so I think that those principles if you look in the literature typically people will say there's sort of three mechanisms at least that we can use here there's sort of the legal mechanism there's policy and then there's technology and we've been talking a lot about technology today through the trusted research environments but there's also policy and policy actually comes into play significantly when we're talking about the commercial use and these direct to consumer related genomics technologies and again it changes depending on where you are in the world in North America the laws that govern direct to consumer are different than those that for research and different for those that are clinical as well and that's again different in Europe in Australia again not being a legal expert myself but many people are probably aware that the privacy act in Australia is one of the key pieces of legislation that covers some of this in Australia and actually there was just a review about that and many individuals here and throughout these institutions have contributed to that feedback and if you read the legislation it's difficult for the legal system to keep up with all of this because it changes so rapidly and a lot of the terminology is from genetics where we're studying a gene or a location or a locus and now we're talking about the entire genome and not only the entire genome we're talking about the genome of millions hundreds of millions of people potentially the population and so again maybe getting back to a point before is that thinking about all of the different expertise that's needed to bring to this it's just so critical that we have all of that discussion together and we don't have a technology organisation or a commercial partner all independently working on things so I'm not sure I'm answering the question but to say that actually it is useful to look around the world and see what's happening there maybe see what's happening in Europe is that a sign for what we might be heading towards in Australia in the future yeah it's thanks for raising that that you've got these different levels what is the law around this, what's policy what's trust because the reality is that if you look at commercial organisations you're having to make sure that you have that trust with your customers and that's really absolute if you look at legislation around things like record linked research and how you can conduct it there aren't there are ways to do things safely and there aren't actual legislative absolute barriers stopping this from happening so that trust and the right platforms is really important because it can happen and if you think about secondary uses of data it's very difficult to get some form of opt-in consent if I want to look at population based stuff on patients from the mid 1990s some of them are dead for goodness sake but we need to understand what happens to people so I'm a big advocate for improving mechanisms of consent there's no question about that in fact even in this building dynamic consent people from this law faculty were amongst the first to define dynamic consent in conjunction with their colleagues at Oxford so we know there are some mechanisms to advance this and I think we should certainly pursue those because that's about building that trust but I think the secure research environments have got certainly a place in that so I've heard governance mention a few times so there's governance around the secure environments but bear in mind there's a separation there the secure environments as a platform where data goes and then there's governance around the data that might actually go into those which is a separate thing so I think for Telstra, you have to deal with both in effect so that's what's happening there and with our team we've got our general practice data set of about 5 million patients and we've got some from our independent data governance committee here in the audience which is consumer led and I can see a high sign over there so all of these different components are really important I think trust is the most important thing what we can do in that space Shaman, do you want to comment on this industry question too? Absolutely, so I guess from an industry perspective yes it's trust but it's also the legal compliance as well and so the reason Telstra Health, I mentioned earlier we've got all these different products that span the patient care continuum so we can follow patients from primary care, hospital care into the aged care facility pharmacy so we've got access to all these individual software even within the same business however we don't have that consent across all these platforms to use it for secondary purposes so we ourselves have to develop that right of use across these platforms in order to link them together if there was that need so it is a one by one consent process for each of these platforms it's all there but it is developing that right framework consent governance framework in order to realise that opportunity yes and we've seen that as well so we've had to have contractual agreements with the practices on how that works on behalf of their patients but I think the key things here is I'm trying to get over that there isn't an absolute legit to live barrier to it doesn't mean to say there's not a lot of work though Jim did you want to add something here? Yes thanks it was a great question I guess at a macro level that tension is increasingly being recognised in the last 12 months the National Institute of Health has set up the NIH foundation largely with a view to how do they set up and to a certain extent govern public private partnerships not just around the use of data but primarily it was set up as I understand with the thoughts around data collaborations in my area of vaccine safety clearly there's both the money and data rich pharmaceutical companies and then there's other groups and clearly for a group like mine who may need to come out and say vaccine X is either safe or not safe and the financial relationship with pharmaceutical companies is unacceptable for multiple obvious reasons and so to have a framework where you're able to perform research together and have a respected body govern that relationship is going to be increasingly critical as time goes by Yeah thanks Jim another question that's come in from our online audience thank you Keith Yeah this is a question which was asked earlier on and I've been holding it back a bit but I think we're now at that point useful to quickly drop it in because I think there's quite a few people that are certainly in the online audience are interested in this is there are quite a few trusted research environments in Australia and examples are Erechur Shure Serp Keypoint things like that and the question is do they differ from each other and do we need many of these Yes I'm intrigued to hear the thoughts from maybe the panel or the audience on that Yeah let's start with the panel but I know we've got people in the audience that may well have a perspective on this as well I'm looking at some of my mollish one-ish colleagues here but let's start with the panel How many of these do we need? Is one enough? Do we need 10? How do we enable integration operability? How do you feel about multiple secure research platforms? Let's start I guess Well wouldn't it be nice if there was just somewhere you went and it was all safe and done but I think as I was mentioning it there I can't really see it happening but I guess from a commercial perspective companies have got their own commercial imperatives but certainly universities and hospitals and so on have their own imperatives too around looking after their data so I think that's probably the biggest driver for multiple but it doesn't mean to say that everyone needs to run their own so we've got platforms like Erika potentially that could be running more centrally and there are others so I guess the challenge as I mentioned is that there isn't actually a tight definition of what they are I suppose I suspect they've all got very similar capabilities but there's not actually a definition maybe one for the RDC that's getting accreditation on what constitutes a secure research environment but don't put the burden on us poor providers to actually get certified I'm only joking but you can imagine the challenge there I know we've got certainly at Monash we're using Monash security research platform but we have the Helix technology platform work as well and clinical registries and trials so I wanted to invite some commentary I know we've got at least three from our Monash community here in the room and wanted to hand the microphone over perhaps to David to sort of talk about a bit of compare and contrast but certainly what you're experiencing in delivering a platform at Monash Thank you, David Powell from Monash Research and Helix so I am now responsible for the SERP instance at Monash I don't have an answer for that that ran out many we need to be nice if there was one solution there are differences as far as I can tell and so a healthy ecosystem makes sense but that answer of what should a theory look like I think we all need help with that because I was just told that there is this problem we have researchers which have data in three different theories and they can't move between them which does hinder research how do we solve that is that a common problem I know we've got a number of different institutions represented here in the room but are others of you experiencing the challenge that your researchers are having to access multiple different secure research systems I'm seeing some nodding going on here in the room anyone want to share the challenges with what that means for themselves Jim how about you So I guess at the moment we have gross hypocrisy on our part so we have to access data from at least three different TREs and we're looking to access a fourth but we're also looking to create our own which would round out the fingers on the hand nicely I think inevitably I think some level of standardisation and harmonisation and I think we can move towards at least in many aspects common data models that will facilitate an easier interchange between these platforms we're in the law building I'm not familiar with jurisdictional law but we are a federation and I think at least compartmentalisation at least based on state-based privacy laws etc are almost inevitable whether that can actually be partitioned in a federal resource is a possibility but I understand that but I suspect both for industry-related reasons and federation-based reasons it's inevitable we're going to have multiple and we need to figure out how to harmonise them as best as possible Thank you I know that the ARDC has been spending a lot of time thinking about different research data commons and we're here talking about people and people platforms so I wanted to invite Adrian to give some commentary from an ARDC perspective of things that are going on in the context of people but particularly in this broader context of health and medical research data sets so Adrian, thank you Thanks, Serena Yes, the ARDC is launching a new five-year program really dedicated to the health research community and what are the national scale infrastructures that we can provide and as we've seen today I think Rosie called it wicked and I'm going to keep telling her that that we need wicked resources here, Rosie No that across that there are very nuanced health service providers state, federal governments increased facilities, MRIs universities, private companies ARDC is not going to do all of that that's not our job we are a digital and data national capability and we are looking to join up the activities across that nuanced landscape the 4.5 challenges that we're really putting out there are data strategy, is the data that the researcher needs there in order to address dementia or diabetes do we have a data strategy for that is it discoverable is the data from Western Australia as easily accessible to someone from Victoria otherwise then the next challenge after that is our trusted research environments so we have a whole challenge over the next five years dedicated to this area and then once you've got it then you're trying to bring data sets together in data integration and data analytics so we are looking to operate some like touch national services in those areas and do a lot with across the health system to provide coordination and coherence so to come to the trusted research environment challenge that we we have been doing consultation over the last 18 months and have got that message very clear that there doesn't seem to be any way to just have one secure research platform as convenient as that might seem and we were possibly at that stage five years ago where we were thinking that might be where we would be now but there are just there are lots of drivers to have lots and multiple custodians need to have different control and then multiple researchers needing different access and they're being funded at smaller levels as well so given that we've got two big programs that we're looking to launch over the next months over the five year period one is a national framework for trusted research environment so responding to a lot of things that you were just saying there how could we even if it's being done in different places what are the common features how would we even define one even if I was going to say yes you could use mine what is yours so even to have that interoperability and federation conversation we have to have a common notion of what it is so anyway ARDC will be resourcing that national conversation framework that touches governance, technology integration some of the social aspects as well and then the second phase of the program would be us looking at the research system and saying okay what does a network of these things look like and where would ARDC be resourcing where there are not environments and we need them to be or resourcing some of the interoperability with the private sector for example so the second stage so in the end we do this for the researchers so you've got to put yourself in the shoes the researcher to say it was Nicole who was saying that over T got one research project the data is now in five different places the tools are in a sixth place and then the computing resource is somewhere else so the reason why ARDC is getting involved here is in order to bring some kind of coordination and coherence for the researcher or the policy maker that needs to ask questions about data that's in these environments Great, thanks very much Adrian lots of food for thought there If anyone would like to help us at the ARDC doing that we are hiring at the moment Well that's actually it's a good segue into the closing question and I'm going to ask each one of panellists and I'm going to go in reverse order from the way you presented so I'm going to start with you Bernie and then maybe we'll just work our way down down the panel but my closing question for each one of you and keep your answer relatively brief if you can the closing question and having just heard what Keith has that Adrian's just described with respect to ARDC plans if there is one thing that you could improve in the environment in which you're working in the institution and group that you're in one thing that you could improve what would it be? I want to say more money please but I won't say that Instead what I would like to say one thing to improve is I think due to the nature of Australia the way that it's a federated country it's distributed by large distances and so on bringing this together in a national program is critical mentioned that already and so I think just stimulating that national engagement across the board bringing people together I wouldn't say this is not happening at all it is absolutely happening at the moment and so I just want more of that to happen and if one thing could really go quickly that would be my first preference would be to try and bring the national agenda together as much as possible Great, thanks Bernie Jim, what would you look to improve? I'll keep it short whatever it takes to improve interoperability Yeah Charmaine I'd really love to see a shift in the conversation from discussions about data sets data linkage, mega data sets to what are we doing with all this data and all these different data a secure research data environments how are we going to action those insights here we have a software vendor with up to 50% reach into primary care where we can action those insights that are found in research why don't we talk about what are we trying to achieve with these linked data we've already got insights from research we know that there's a 20 year gap to translate research into application let's try and test some of those findings within a software vendor that's able to action those findings within the software without extracting data outside of the system so I'd really love to see a shift from gathering research insights to actioning those insights and testing whether they're actually working in real reality Thanks Charmaine Dougie, how about you? I think back with that focus with thinking about the RDC it's really challenging from a provider perspective to a a process that works efficiently to get access happening fast so certainly trying to address efficiency is a really difficult thing in this space so that would be really good but I'd also like to barric for what Jim was saying there around interoperability or things like common data models there are other ways we can access data without having to release it but Charmaine sometimes we do have to link it to answer the question I'm afraid but I'd like to think about that Thank you Thanks very much so you've heard some interesting insights I mean if I just try and wrap up some of the conversation that we've been having this afternoon I think we've heard about lots of examples of how institutions are building and delivering trusted research environments there isn't one size fits all there are lots of different options I think we also heard that we don't necessarily understand the pros and cons of one approach versus another we just know that there are lots we've got examples of institutions where there are multiple different trusted research environments that researchers are having to think about how do you integrate across what are some of those interoperability challenges a lot of dialogue around the role of Australia the Australian Government guidelines Australia is a collaborative nation how can we integrate and work more effectively together the need for data privacy the need for informed consent so there are lots of challenges and opportunities that we've heard this afternoon and so as we think about the future the opportunities for us to be able to continue to innovate around these trusted research environments are tremendous but we have to be able to innovate and that does open us up to the conversation around globally what our global best practice is how do we think about sharing data are we willing to share data outside the shores of Australia and if so under what conditions so a lot of really interesting dialogue and rich commentary this afternoon and so I'd like to thank the panellists that have joined us for this great discussion today Dougie, Charmaine, Jim and Bernie and invite all of you to join me in thanking them for their contributions, thanks very much and before we wrap up I'm going to hand over to Rosie Hicks for some closing remarks, thank you Rosie I have to do this bit because otherwise who else thanks Joanna but I'll get to that in just a second I do want to echo the thanks to all of you in the room for this afternoon's discussion my thanks to all of you online whether you're live today or watching this subsequently I really hope that you've heard something this afternoon that changes the way you think about this problem because if you have then we've succeeded there was a comment that was made to me as we were gathering at the start this afternoon that all too often these groups, whether it's researchers whether it's industry, whether it's government come together within their silos and admire the problem and I believe the phrase that was used was an echo chamber I can't remember who it was was it you Dougie, was it Jim? Jim the echo chamber and what we really seek to do with the ARDC leadership series is to break out of our own echo chambers bring all the groups together for a rich and informative conversation that can change what we do next now I know that you're all with me on that journey we give our thanks to the panel a special thank you to Joanna I also want to extend my thanks to all of the ARDC staff that make the series possible, Asher thank you at the back of the room your hand on this is so obvious but also a quiet hand to those that don't get recognition every day but are such an important part of the ARDC team as we move forward with these programs so that we can enable Australian researchers through world leading knowledge infrastructure to transform our lives thank you