Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Senate Select Committee on Health
11/12/2015
Health policy, administration and expenditure

ELLIOTT, Dr Julian, Senior Research Fellow, Australasian Cochrane Centre

JORM, Prof. Louisa, Director, Centre for Big Data Research in Health, University of New South Wales

PEARSON, Prof. Sallie-Anne, Head, Medicines Policy Research Unit, Centre for Big Data Research in Health, University of New South Wales

WATSON, Dr Diane Elizabeth, Chief Executive Officer, National Health Performance Authority

[09:20]

CHAIR: Welcome to our first roundtable today. I appreciate the time that you have taken to make yourselves available to talk with us. I remind committee members and officers that the Senate has resolved that an officer of a department of the Commonwealth or of a state shall not be asked to give opinions on matters of policy and should be given reasonable opportunity to refer questions asked of the officer to superior officers or to a minister. This resolution prohibits only questions asking for opinions on matters of policy and does not preclude questions asking for explanations of policies or factual questions about when and how policies are adopted. Please let us know any further capacity in which you appear.

Dr Watson : I would like to let the committee know that I have two other capacities that can be engaged to shed light on this topic. One is that I am an advisory board member for the National Population Health Research Network, which represents a large investment on behalf of the Commonwealth and the states and territories to support data linkage capacity in the country. Prior to that role, which is about a year and a half in duration, I was, as an individual appointee, a reviewer of the National Population Health Research Network. I was probably asked to be a reviewer because, prior to my six years in Australia, I spent 10 years in Canada building the population based data linkage architecture in that country. I heard your question earlier about examples of health and outside of health, so I can give you a few of those later if you wish. I will try to preface my comments about things which relate to my statutory office holder position for the National Health Performance Authority versus that which you would otherwise like to engage.

CHAIR: Thank you very much. That is very helpful.

Prof. Pearson : I also chair the New South Wales Population and Health Service Research Ethics Committee. The background might be useful for our discussions today because that committee oversees all of the health data linkage that is undertaken in New South Wales.

Prof. Jorm : Another role of relevance is that I am a board member of the New South Wales Bureau of Health Information, which is a statutory body responsible for reporting on health system performance in New South Wales. I have tabled a report from the Bureau of Health Information.

CHAIR: We have received that, thank you.

Dr Elliott : I am also a doctor, a clinician, and I work at the Alfred Hospital in Melbourne, as well as being a clinical researcher.

CHAIR: We are very excited to have this opportunity to hear from you. Would each of you like to make an opening statement? That will give us some information from which I know we will have many questions. Would someone like to kick off?

Prof. Jorm : Thank you very much for inviting me to appear. The Australian Atlas of Healthcare Variation was released by the Australian Commission on Safety and Quality in Health Care last month. Many of you may have seen that, and it is a great example of the use of Commonwealth big data for health system performance measurement, but I would note that basically it relies completely on unlinked health data—so individual data collections.

The atlas presents, really, a quite striking picture of very substantial variation in healthcare use across the country, covering areas such as antibiotic prescribing, surgical interventions, mental health and diagnostic services. One message that comes through very clearly in the atlas is that, now we have demonstrated all of these variations in care, we need to do a lot more work to actually assess the outcomes of interventions so we can help to identify appropriate treatment rates and what level of variation is warranted. We know there is variation, but we do not know what the right rate of many of these interventions is.

Doing this will require that we bring together data about characteristics of patients themselves, with details of the health care we provide to them and with data about the short- and long-term outcomes of this care. Basically, this means integrating Commonwealth health data with the rich data that are held at the jurisdictional, local health district and hospital levels, as well as private sector data and research datasets. These include, for example, data from clinical trials, general practitioners' data, and data from pathology and imaging services. Many of these datasets are not currently standardised across jurisdictions, and often they are not even standardised across health districts within jurisdictions.

Just getting access to health data used to be a challenge and a major roadblock for researchers like me, but with the increasing flood of electronic data that are available, and with the very robust frameworks in place currently for governing data linkage and managing the associated privacy risks, the current bottleneck has moved somewhere else: the bottleneck is our limited capacity for data linkage. Currently only the Australian Bureau of Statistics and the Australian Institute of Health and Welfare are accredited as integrating authorities for Commonwealth health data, and yet we have state-of-the-art health data linkage facilities operating in states and territories, including New South Wales, South Australia and the Northern Territory, and Western Australia. National capacity for data linkage would be dramatically boosted if these state and territory based units were accredited to link Commonwealth as well as jurisdictional data. In particular, it makes a lot of sense for these units to service projects that involve linkage of data from a single jurisdiction with Commonwealth data, rather than these projects adding to the linkage logjam and linkage queue that currently exists at the Australian Institute of Health and Welfare. Accrediting these additional units will facilitate research that uses, for example, data from both public and private imaging services linked with data about surgical procedures and longer term patient outcomes and costs to provide the evidence we need to tackle the twelvefold variation between local areas in CT scans for lower back pain that was identified in the Australian Atlas of Healthcare Variation.

This research is best done through collaboration between academic researchers like our centre, policymakers and clinicians, but we need more investment in programs of partnership research into clinical variation using these large-scale linked data and we need long-term investment and long-term programs of research. In the end, these will yield a big return for Australia in terms of better care for patients and less waste in our health system. Thank you.

CHAIR: Thank you very much, Professor Jorm. Professor Pearson, would you like to make an opening statement?

Prof. Pearson : Thank you for the opportunity to speak today. The basis of the submission that I co-authored is that establishing whole-of-population data linkages to monitor the outcomes of our health investment is a public good. I am actually going to use the Pharmaceutical Benefits Scheme today as an exemplar, but the issues that I argue can equally apply across other portfolios in health.

Last year the Commonwealth spent more than $9 billion on prescribed medicines dispensed through the PBS, yet the value proposition and real-world safety and effectiveness of these medicines is not well understood. The Drug Utilisation Sub Committee of the PBAC, of which I am a member, analyses PBS data to examine real-world use and costs of selected medicines and also has the capacity to track person-level patterns of use, but the analyses are essentially restricted to PBS claims alone.

However, pharmacological treatment does not happen in isolation. The linkage of PBS, MBS and other Commonwealth collections, such as those held by the Department of Social Services, can expand our opportunities to explore value, real-world use and pivotal issues such as equity of access. Despite medicines being tested extensively in clinical trials, when they are PBS subsidised there is significant uncertainty about how they would perform in routine clinical care. When I talk to consumers, they are surprised to learn that comprehensive postmarket surveillance research does not occur routinely in Australia. Why is this the case? Activity of this kind actually requires Commonwealth and state based data holdings to be linked.

This is a very different scenario to Canada, which Barbara talked about earlier, because Canada basically has provincial health care and the data is linked within the provinces. We have a cross-jurisdictional situation that we need to tackle. This was identified as a key issue in a recent report from the Expert Panel Review of Medicines and Medical Device Regulation. The report stated that a reliable assessment of safety and efficacy requires data from a range of different sources. The currently fragmented data systems in Australia make it difficult, if not impossible, to systematically capture these impacts.

My own research has identified that fewer than 30 studies have examined drug safety in the last 25 years. This needs to change. Australia is actually well-placed to deeply understand our return on PBS investment, and also other health programs. The data already exists. We have information that covers our entire population. The 10 per cent PBS sample, in my view, is not sufficient to do this. We have the capacity to link the necessary data from different health jurisdictions; therefore, we strongly recommend that the Commonwealth mobilise and support cross-jurisdictional data integration by creating updated datasets, rather than one-off linkages for specific projects. Funding long-term partnership programs, rather than relying on uncertain grant funding, will also ensure that the research community and health agencies will work hand in hand to undertake robust and timely monitoring, surveillance and research.

CHAIR: Thank you very much. Professor Pearson, you referred to a report—what was that report?

Prof. Pearson : Yes. That was the Expert Panel Review of Medicines and Medical Device Regulation—the Therapeutic Goods Administration review. It was chaired by Professor Lloyd Sansom.

Dr Elliott : Thank you for the opportunity to present today. My main point is that big data is the starting point, not the end in itself. We have more data than ever before, but to get the most value from these data, we need the systems and the national capacity to use the data effectively. These systems enable us to convert data into information and knowledge, and we can then use that knowledge to drive action in health and then monitor the effects of these actions. These cycles already exist, where we go from information to knowledge to action, and then monitoring of that.

But with the advent of an increasing array and diversity of data sources, there is the opportunity to create much more effective and cost-efficient cycles of health data, knowledge, action and evaluation in which we make data available, findable and usable, as the other speakers have highlighted, but also invest in our ability to synthesise these datasets using appropriate and trustworthy methods and systems. We would use this then to inform the creation and updating of evidence-informed policy standards and guidelines, use effective mechanisms to then translate those recommendations and policy into action, and then, importantly, use these systems to monitor their effects in actual healthcare practice, resource use and outcomes, so creating a new cycle, a new ecosystem of the use of data in Australia.

So the Commonwealth has a central role in establishing the framework for the use of big data in Australia, going forward. This should include investment in the operational capacity of government and non-government organisations not only to generate and link data but also to effectively use and act on the data that is generated. It should also include an investment in our human resource capacity in Australia to use these large and complex datasets. Again, that should include not only the generation and linkage of those data but, importantly, the ability to synthesise those data into a meaningful understanding of what the data is saying and then use those data to generate action in health care and monitoring of those actions.

CHAIR: Thank you very much, Dr Elliott. Dr Watson?

Dr Watson : I did not prepare a statement, except to speak about the background which I bring to the table, both from Canada, sitting on the population health research board, as well as its independent review, which engaged people right across the country on data linkage, as well as my statutory position and also knowing that there are so many people on the panel. I will just give you a couple of sentences on the 10 years in Canada building information architecture and the six years in Australia.

On my arrival, in my global observations of the differences of the community and their maturation in building linked information systems, I was quite astonished—and, now that I know the Australian language, I would say it might even qualify for the appropriate use of the Australian word 'shocking'! It had nothing to do with views across the government and the research community around the need and the potential benefits and little to do with goodwill. It had more to do with the complexity of your system, the complexity of legislation, more particularly the authorising environment in which this work occurs.

Six years later, as a Canadian in Australia observing this community, both within government and within the research community, I guess my observation today is that demand is really high, and that relates to the tremendous goodwill in government and in the academic community to use information to drive quality improvement on behalf of patients, as illustrated by previous speakers. I also see that, in the last week in the newspaper, the policy environment is very ripe around innovation, the use of data to drive that agenda. We also see tremendous goodwill in the Commonwealth Health portfolio, with the speeches that have been made in the last few weeks by our secretary.

I would also observe that, in the Population Health Research Network, with tremendous funding support over the years by the Commonwealth government and state and territory governments, the technical architecture and the partnerships and relationships and knowledge base are there to accomplish the technical work. Taking these key ingredients that you have around goodwill, legislation, authorising environment, need or demand for this type of information, the process to reach resolution to expedite the speed at which the volume and the productivity of this type of work occurs will be complex and will take a commitment probably of time and money and an organising framework to bring the community together to address each of the hurdles. In my experience, you will find that the barriers are numerous, but they all relate; depending on what you link, this is going to be the problem, and if you want to link this, then this is the problem. So you will find that problems are common. It takes time to find the common threads. To me, those threads relate to legislation, in some instances, depending on what linkage you want to do. They relate to guidelines that govern the implementation of legislation, depending on what you want to link. They relate to an authorising environment that becomes very complex very quickly depending on what you want to link.

CHAIR: Thank you very much, Dr Watson.

Senator MOORE: Can I just ask whether you know each other?

Prof. Pearson : We just met Julian.

Senator MOORE: But do you know each other by email? This is a relatively new area for me, I am interested in the range of submissions we have in this inquiry and whether people talk to each other across the board. I am sure they do, but it is just an open question to the people on the roundtable.

Prof. Pearson : The answer is yes.

Senator MOORE: Good.

Prof. Pearson : And of many of the other people who will be appearing today—

Senator MOORE: That would be what I would expect. It is a really big issue, but it seems that the people working on it are in different areas. You can identify on a page the places that this work is being done.

Dr Watson : If you wanted to talk about the whole community, both in the government and in the academic communities in Australia, I would say that there is a core of people who know each other, and have known each other for a very long time. I knew Louisa for years before I even moved to Australia. I have known Barbara for decades. That community is actually very large. That is why you see conferences and why there are international data linkage conferences—because, internationally, those people know each other. It is why I knew Louisa for years before I came to Australia. Just in Australia, that community of government and researchers, for whom this is their core work and where there are core opportunities, is actually very large. There are people on the periphery of that who spend their time doing other things with other research methods because they cannot easily access and do that work. So the potential to expand that core base of people will happen relatively rapidly should the authorising environment become much more simple. There are lots of people who spend their time doing other things because of the authorising environment. Personally, let's say, as an academic, or as a government department, you spend your time doing other things because of the nature of the complexity of the authorising environment.

Senator MOORE: So it is actually a blockage?

Dr Watson : That is correct—in my opinion.

Prof. Jorm : As someone who has worked in two medical schools, at Western Sydney University and University of New South Wales—where, obviously, there is a lot of opportunity for interacting with clinicians—I know that in both cases at the moment clinicians become aware of the possibilities of data linkage and of what they might be able to do in linking their local clinical data sets with longer term patient outcomes they get very excited. So I think there is a big group of clinical academics out there who, if things were made easier for them, would also be leaping into this research, as Diane points out.

CHAIR: Dr Elliott, do you want to make a comment there? It sounds like they might be talking about you!

Dr Elliott : Absolutely. I just support the comment. I completely agree. There are a lot of people who are essentially working on very small datasets. They are trying to answer questions with datasets that are underpowered, essentially. They just do not have the data in order to answer the sort of questions that they are interested in. What we are talking about here is enabling people to get to those more definitive questions much more easily.

Senator MOORE: Throughout all of the submissions, the issue about the authorisation process keeps coming up. The term 'custodian' comes up a lot. I think in just about everything that I have read there is a very similar recommendation that we have to do something about the authorisation processes. In your submissions, you have actually said that that has to happen. What needs to be done about the authorisation process?

CHAIR: I think that was the heaviest side, Professor Pearson. So you get to have the first go.

Senator MOORE: I am not sure whether Hansard picked up the collective sighing from all of the witnesses!

Prof. Pearson : The custodianship of the data is a very important position to be placed in. I think traditionally custodians have been very risk averse. There are different interpretations of legislation guidelines, et cetera, about the nature and the type of data that can be released. I have actually seen a massive change over the last few years where custodians are now thinking about arguments as to why data should not be released—

Senator MOORE: Turning the onus around.

Prof. Pearson : rather than thinking, 'We have to stop this.' I really have sympathy with the position, but there is inconsistency. Starting to think about an even playing field and getting custodians to talk to each other around their policies and release processes are, I think, useful. We have started that in New South Wales, actually. As the chair of the ethics committee, we convened a lot of the main players in the data custodians. Louisa was at this meeting. It was amazing—sitting around and talking to custodians. They share a common concern, but once we started fleshing things out a lot of the challenges that were faced seemed to be put aside. People started to harmonise the way they were thinking about this.

Senator MOORE: Who are the custodians? If we got all of the custodians together, are they designated positions?

Prof. Pearson : There are, literally, hundreds of them. In NSW Health, for example, there are custodians of individual datasets and there are delegates at the Commonwealth level who are responsible for the release of data. It depends on the dataset. They can be state based. They can be hospital based. So there are literally hundreds, which really plays into the variation issue that I have discussed before.

CHAIR: You said that, when they get together, they share a common concern, or maybe it is a set of concerns.

Prof. Pearson : Yes.

CHAIR: Could you articulate what they are?

Prof. Pearson : It is about the responsibility to ensure that data that is released is beneficial but also minimises the risk of identification. That is one of the things that they really have spent a lot of time thinking about. But I think now, given a lot of the infrastructure that has been put in place, that concern and those risks are minimised—because of facilities like secure unified research environments et cetera, where data sit and researchers basically cannot take the datasets and have them on their individual servers at universities and so on. So there are lots of things—

CHAIR: Can you tell me about that practical reality? One of the things that always come up is privacy, privacy, privacy. What is there now technologically that makes that less of a risk? You have almost all dismissed it immediately as something that was a big mountain of concern but that has been bypassed. What has changed that you feel that way?

Prof. Pearson : Louisa was pivotal in establishing one of those environments, so I think it is probably more appropriate for her to comment.

Prof. Jorm : I was actually involved in developing the Centre for Health Record Linkage, which is the New South Wales data linkage unit, as well as the secure unified research environment, and they are sort of tackling—

Senator MOORE: That is the Sax Institute?

Prof. Jorm : Yes. They are tackling different ends of this issue. But you are probably aware that the model for data linkage that is used in Australia is one where the identifying details for individual people—names, addresses, dates of birth and so on—are linked together, but the health information about those people is held completely separately from those identifying details and held by the individuals that were just referred to as the data custodians. So all of the identifying details and all the health data are never together in the one place at the one time. It is called the separation principle, and it means that what is given to researchers just has a unique number on it that the researchers can use to connect the records together but does not have the identifying details such as name, address and so on. Now, that model has been in place since data linkage has been occurring in Australia. In New South Wales, it has been in place since 2000, for example.

The newer developments include ways of holding the data in secure environments that limit the possibility that they can be downloaded or shared or copied. That is what the secure unified research environment is. I think you will hear more about that today. Basically, the data are stored in a secure server in a bank-vault-like atmosphere of security. In fact, data from banks, federal agencies and all sorts of other things are in the same data centre. The researchers themselves never hold any data locally. Everything is done in a virtualised environment. Everything that goes in and out of that environment is logged, tracked and audited. Researchers are not able to download and hold local copies of those data. The data that they are using are de-identified; but, even so, this adds an extra level of security so that the data cannot be downloaded and copied and shared.

Senator MOORE: In other committees I have been involved in, privacy issues always come up and will continue to come up—and I think that is a community issue as much as anything else. Given what you have just said to us about the privacy aspects being addressed, the argument can be made, what other limitations are there for the custodians? What are the other problems? Would you know, Dr Watson? This seems to be something you have been looking at.

Dr Watson : Yes.

Senator MOORE: We have put privacy in a very secure box over there. What else is a problem?

Dr Watson : Again, as a Canadian I think the context in which historic decisions around legislation, guidelines and the authorising environment are made is a culmination of events that have occurred over time, and I would say there must have been a very risk averse environment in Australia to create the history and the trajectory that brought you to today—very, very risk averse relative to other countries.

How do you reset that agenda? You as Australians could learn from other countries that had less risk-averse environments that created their legislation—legislation which is very similar, I must say—and their guidelines and authorising environments, which are different, as you would understand. There are countries now that are some 10, 20 and 30 years ahead of Australia in implementing guidelines and authorising environments that have resulted in a very, very low prevalence of privacy breaches. I just read a quote this week, in preparation for presenting today—I think it was in the bureau's report—that at Fiona Stanley in Western Australia in 20 years there had been no privacy breaches.

I worked in a location that has been linking data in Canada for almost 30 years now. There have been no privacy breaches. I will say that while I was there there was an employment behaviour breach—one that I know of in 10 years—and that resulted in no breaches of privacy. But there are guidelines, processes, policies and procedures in place to ensure security. When those are breached, employees, for example, are let go. I do not know of any substantive examples. I have heard of one recently, but over the 20 years I know of no other privacy breaches that have caused harm in this kind of work in Canada.

Coming back to my first point, the culmination of what has happened in Australia, both in your legislative framework and your guidelines process that comes from the privacy commissioners as well as your authorising environment, in your history, must obviously, to me, be a very, very risk-averse environment. Revisiting some of those guidelines and some of those authorising environments, in light of the events that have not occurred in other countries and have not occurred in Western Australia, might be one of the avenues of cutting through the existing hurdles in the authorising guidelines and authorising environment.

Senator DASTYARI: Thank you for that, Dr Watson. The point it seems you are making is that the concerns that obviously exist are not necessarily borne out in the facts. I think the challenge becomes—and, again, I think you are touching on this—how you win over that community support for the collection of information. I suspect that it has more to do with an emotional visceral reaction to the collection of information rather than necessarily a fear that that information is going to appear somewhere. You are right: at one extreme there is the Ashley Madison thing of, 'Hey, this is my private thing and now it is public.' I was not on that, by the way. But there is the idea that someone is going to break into the data and it is going to get leaked and my neighbours are going to find out that I have some embarrassing condition. Maybe it is not even embarrassing; maybe it is just information about me that I do not want out there. That is the more extreme example.

On the other hand, you just have that sense of unease about other people being able to collect that type of information on you. I know one of our fellow senators Senator Leyonhjelm, who very much has that libertarian streak, would argue that the whole collection of the information itself is an infringement on civil liberties and rights. I do not agree with that per se, but there is that sentiment out there, especially when you tie that into other debates around national security and other types of information collection. It all gets conflated into this world of: 'They are gathering all this information on me. What does that mean?'

Dr Watson : Can I make two comments about that?

Senator DASTYARI: I would not mind if you could give us a comment on that, yes.

Dr Watson : Those views exist in Canada. Have no doubt that those views and those discussions exist in Canada. I can give you maybe two practical examples of events in Canada in 2003 that altered the trajectory of the authorising environment very practically. If you use my analogy that you have your legislation and in reviewing your legislation, as I have done in my Commonwealth office holder position, the principles and legislation are very, very similar. I do not see a lot of differences in your legislation. The next tier up from that is some of the guidelines that you have in interpreting your legislation—for example, guidelines around the custodianship responsibilities around PBS and MBS being in separate databases et cetera. In many instances, those are designed by your privacy commissioners. There was a process of engagement of privacy commissioners in Canada between 2001 and 2013 led by the privacy commissioners in tandem with the Canadian Institute for Health Research, which is equivalent to your NHMRC, with an equal number of researchers and privacy commissioners in the rooms. In the jurisdiction I was in at the time, at the advice of another privacy commissioner I went to talk to my privacy commissioner who did not even know that we were doing this type of work, the benefits of it and the protocols, policies and procedures under which we did the work. That process created much more altered guidelines in ways that were respectful of privacy and communities that wanted privacy for all the reasons that you have described and it resulted in guidelines that were enabling, while respecting privacy and confidentiality.

The other thing that happened at the same time in that three or four year period in the early 2000s was the Canadian Institute for Health Research funded national assessments of the public's views on the authorising environments and the capacity to use weighed against the benefits of data for those purposes. There was Diane's reflection back then—the summary of all that work. It led us to the fact that Canadians were extremely surprised, if not shocked, that the information that governments held about their use of health services was not used to improve the strength of the healthcare system and was not used to improve their health. For example, as Dr Mintzes said earlier, when you are prescribed a medication in British Columbia and you go to a pharmacy, they know the other kinds of medicines that you have, because they are looking for potential drug interactions. Canadians were surprised that we were not necessarily doing that and it was not occurring. If it is occurring in British Columbia, why is that not happening elsewhere in the country so that you can use information to protect me?

My reflection on that time is that the authorising environment became much less risk averse by understanding Canadians' views through those engagement undertakings and it became much more an environment of a reflection on existing guidelines, as articulated by privacy commissioners when they understood the potential to benefit and the practices that can be put in place to ensure privacy and confidentiality of information.

Prof. Pearson : I would like to clarify: the term 'concerns' probably should be replaced with the word 'challenges'. There clearly is an enormous responsibility on the part of custodians to protect the identity of individuals, but the challenges that they face relate to the interpretation of legislation, which can be different with different individuals. The other great challenge that they face is that oftentimes this role is done on top of a highly powered, high-level and highly committed role. It is really about goodwill and trying to serve the research community. One of the greatest challenges they face is that this is incredibly under-resourced, and designated roles and having people responsible to do this work would be a very pragmatic and significant advance in getting some of those current logjams freed.

Prof. Jorm : I would second that. In fact, several data custodians have said to me over the years: 'Providing data for research is not core business for our agency.' It is viewed as an add-on. As Diane pointed out earlier, there is a lot of evidence at the Commonwealth level that the policy environment is changing and there is a bit of a sea change. If the leaders of their agencies are saying, 'This is your core business and we're going to resource it,' then things will start to move.

CHAIR: In your view, what drives that change? Is it increasing awareness, HR requirements or KPIs?

Prof. Jorm : I think it is leadership from the top of organisations and the specific allocation of resources.

CHAIR: So innovative individuals who are reading the data nationally or internationally? You have just talked about a sea change. What was it that has driven that?

Prof. Jorm : I think a number of appointments at the Commonwealth level and the new Commonwealth Secretary of Health. I think a report was released by Prime Minister and Cabinet last week—sorry, I cannot remember the name of the report—which was basically about making better use of Commonwealth data and setting out a whole series of projects and initiatives to do that. I was very encouraged to see that report being released.

Dr Elliott : A point I would also make is about linking it to the innovation agenda beyond health—looking at the way that data are driving new business globally. That is as important in health as it is in other sectors, and Australia has a good opportunity to join the new industries of the next century, but that is going to be dependent upon the policy environment available in Australia to access and use data for those new products and services. That is, I guess, just a commercial element of the general point that needs to be in the mix around the benefits or the use of the data. If you talk to people about their data and the benefits that can come from using those data, most of those surveys show that people are very supportive of that.

Senator MOORE: Is there anywhere that links all the datasets that there are in Australia? Is there a point that you could go to where someone could say what data is available?

Prof. Pearson : No. A lot of it actually relies on oral history currently.

Senator MOORE: Which is funny in itself when you think about it, isn't it—datasets and oral history?

Prof. Pearson : Yes, it is really quite incredible. One of the things that I firmly believe is that this certainly advantages people that I call 'inside the tent'—people that know people who are in this smaller group that we discussed earlier. I think that to really advance this agenda there needs to be a place where people can go to find out information. I probably field at least three or four calls a week from people who are trying to understand how to navigate the system, so this is inefficient.

CHAIR: How does the atlas assist, or not, in this case?

Prof. Jorm : The atlas is basically a report publishing analyses of data, and it is not linked data that were used. In fact, Diane's agency did the analyses.

Dr Elliott : One point I would make there is that it is not really about the data being in one place but at least the information about where the data sits.

Senator MOORE: Where it is.

Dr Elliott : Yes, like a clearing house. It could be federated in some ways. There could at least be an ability—and this is, I think, what other countries have found. The US has obviously been much further ahead in the open government type agenda. There was initially this wave of release of data—it was the default position of the new Obama administration—but what they have found over time is that it is not just about making the data available; it is also about making it findable and usable—so structured—and having mechanisms in place in which people can navigate and find the datasets that they need.

Prof. Jorm : There is also the need for data about data, otherwise known as metadata. For example, with MBS—Medicare data—the metadata that are available to researchers are extremely patchy, and again it is one of these oral history things: you go and ask another researcher, and we gradually build up knowledge over time, but you cannot go to a single place to actually find out what the changes to the dataset have been over the years and how to interpret a lot of complexities within that dataset.

CHAIR: Would such a clearing house as we just talked about in terms of data also be a site at which metadata and that oral history might be recorded so that it could be shared a little bit more fully?

Prof. Pearson : Yes.

CHAIR: That would be one place. Can you just dream for me for a moment about what such a clearing house might look like.

Dr Watson : I will describe the historic approach in Canada to this clearing house, because the architecture exists in Australia but it is used a bit differently to the similar architecture in Canada. A clearing house idea is essentially a central location that keeps the repository of data and is the repository about how to use it, so it does not rely on an oral history. In Canada, it is also the repository of knowledge about what the authorising environment would be for your project. It is different for different projects, depending on the information systems that you would like to link. You are using the term 'clearing house'. In Canada there are centres in each of the jurisdictions and nationally that are the repository of the data, the repository of the knowledge about how to use it and the repository about the authorisation process that you would need to undertake for your project.

Let us bring that to Australia for a minute. We have two authorising locations in the area of health—the Australian Bureau of Statistics and AIHW. We have a location in each one of the states and territories that is the state location and they are linked nationally, because they use similar national information systems like hospital files, so they share from each other. In each location they do not need to be the repository for the oral history for hospitals, because that is a national repository. So Australia has the locations. What was interesting to me was that those locations were built and funded under infrastructure, under NCRIS, but in Canada the repositories were not just built for an infrastructure; they are also funded to be the authorising environment and to be the repository of knowledge. I will say that Canada just did not come up with money to pay for the authorising environment infrastructure. I would use the example of the Manitoba Centre for Health Policy, because I was in the leadership team there—and it may not reflect their environment today. When I was there we were funded for five or seven projects every year. It was agreed the project was a policy relevant or a practice relevant deliverable. About 20 per cent of that money was used to keep the oral history, after that project was done, about how you use the data in a unique way. To do that documentation we made it available in a web environment—as it was in those days—so that the architecture was built as a legacy of the production of a deliverable for government.

CHAIR: Constantly self-renewing.

Dr Watson : Over the course of 20 years. That is how the Manitoba centre—an international standout—has been built. Every year, year on year, for almost 25 or 30 years now, it has been funded for deliverables. Each one of those deliverable adds a legacy to the architecture. When I was working in the management structure there, one deliverable every year was an exploratory deliverable. I remember the year we were given access to the pharmaceutical files; I remember the year we were given access to the home care files; and I remember the year we were given access to the education, family services and housing data. That first year there is a deliverable, and that is where I have examples for you. What we did in the first year—it is a deliverable that was pared down in terms of policy relevance, because we had to do all the linkage, and then we did some descriptive information that they would have never seen before that was now available because that new information system existed and was being developed from a very immature point to a mature point, bringing together, for example, education and health data for the very first time. So one deliverable a year was always an R&D type deliverable.

Transporting that back to Australia, you actually have the locations that hold the repositories that are the centres for the oral history and that are asked to be the authorising environment but are not necessarily funded to be the place to learn about what the authorising environment is for a researcher for their particular project, or a government official who needs to work and would like to link data for a particular project and needs to go through this oral process about, 'How do we do this?'

Prof. Jorm : There is a very major difference though between the Canadian situation and ours. The Manitoba centre that has just been described is actually a research centre that sits in a university, while all of the record linkage units in Australia are actually based in government agencies—the AIHW, ABS and state government health departments; they are not based in research groups.

Dr Watson : There is one difference in the legislation, though, in Manitoba. Universities are considered a public body. In Canada, the health data, the education data can be held as a custodian in government or a public body. The university, under legislation, is considered a public body, and that is why the repository could sit at a university.

Prof. Jorm : What I am trying to get at is the current data linkage units would not be able to operate under the sort of model that Manitoba does, where they are analysing the data and producing policy relevant deliverables, because they do not analyse the data. They link the data. So we have this sort of separation thing going on. The other difference is that the Canadians have a unique health identifier. The way linkage is done is actually different there, as well, because we do not have unique health identifiers—at least not ones that we use routinely for linkage in Australia. Therefore, a probabilistic matching—or fuzzy matching—is the main way that linkage is done here. But I would actually argue that the model where there is a much stronger relationship with an academic research group and the linkage units would be extremely beneficial. I think I have referred to that in my statement earlier. Creating these partnership approaches is what we really need to do to get the best of both worlds working with these data.

CHAIR: If you can come back a few degrees for my non-experienced mind in terms of these areas, you talked—

Unidentified speaker : You have to get linked!

CHAIR: I have to get better linkage going on! My data linkage is failing me! Separation and then the unique health identifier—these seem to be very important differences that might go to the issue of privacy for Australians, in my view. And there other things I am thinking about with the universities. Imagine if it went to one university in Australia. This would be a big event, wouldn't it? Can you talk to me about how the Manitoba model can shed some light on a path forward? Are they going to be political and cultural impossible barriers that would prevent us moving to that kind of a model? What alternatives do you see?

Prof. Jorm : Yes, it would be difficult to have a single university. In the case of Manitoba, there is only one university in that province.

Senator MOORE: Such as back to the province issue?

Prof. Jorm : Yes. In Australia, clearly, we have a lot of universities that are in competition. What we would really be looking for would be coalitions of researchers from across institutions operating in, perhaps, virtual ways—but, again, to fill deliverables for government contracts. Governments could decide to call for proposals from researchers for, say, a five-year program of work for a certain amount of money, with the deliverables to be negotiated every year, similar to the Manitoba model. Then researchers are very good at organising themselves into coalitions that contain the necessary mix of skills. Something like the SURE environment, which is virtualised and allows access from anywhere in the country, would be enabling infrastructure for that more sort of virtualised research centre.

CHAIR: Can you fill me in on what the SURE environment is?

Prof. Jorm : That is the remote access computing environment.

Senator MOORE: From the Sax.

CHAIR: Yes. We were talking about that one earlier. I was reading 's', 'h', 'o', 'r', 'e' when you said, and I was thinking, 'What is this?'

Prof. Jorm : They are called safe havens. They are used internationally. This may come up: the SURE capacity will soon be exceeded. There is probably a need for larger numbers of these safe havens. Most other countries have multiple safe havens operating.

CHAIR: Is that related to population size or data growth?

Prof. Jorm : Partly, but it is often good to have more than one option out there, as well. A little bit of competition often helps in terms of driving innovation and keeping prices down for researchers.

Senator MOORE: Also, you mentioned under the current system we only have two authorising places, and it gets to be a logjam.

Prof. Jorm : Yes.

Senator MOORE: If everything is having to be channelled through one or two points, and you have a lot of research going on, there could be—and I just quoted that before—linkage logjams and linkage queues. If there is only one place, that could be a problem. I read somewhere that we do not maintain linkages. Currently, when you do the work with the PBS and the MBS, in particular, that work could be done for a particular purpose, but that is not maintained. I was trying to find it—there is a term for it.

Prof. Jorm : 'Link and destroy.'

Senator MOORE: Correct.

Prof. Jorm : It is a great term.

Senator MOORE: That does not seem to be the way things operate in Canada. It goes back, and we have to look at it, to this bit in the health legislation that says MBS data and PBS data cannot be linked, which is a long entrenched position. It would be really good, when we get the department here, to get some history about that and about why. Certainly it has not come up in discussions that I have had until I was reading this evidence. It seems to be a really big issue, for health in particular, that the two largest pieces of data—there is a legislative reason that people cannot link them. Do you know what that is, and what the impact of that is on your work?

Prof. Pearson : There is a significant impact. I would like to clarify that the integrating authorities have linked these datasets and they have been released for research purposes. The agencies, I believe, are prevented from linking their MBS and PBS data, but if you actually think about it—if you want to understand something as basic as does a person go to a specialist to get a particular medicine and then how is the medicine continued, is it continued by a specialist or by a general practitioner—we cannot find that out because we do not have the visits linked to the prescription. Very basic things around navigating through that system are really actually important—are people monitored after they are prescribed a medicine? We do not know that because that information cannot be linked. There are some really practical impediments to doing some very basic work in this regard.

Senator MOORE: It seems core, doesn't it?

CHAIR: Yes, it does. If you go to Dr Watson's issue about changing the nature of community awareness around—'Oh, what? You don't know this stuff? You actually aren't figuring out whether this is good for us or bad for us, even though you have the data?'—that struck me as a critical moment of community attitudinal shift to data: 'It's not just my data and my privacy, it's my data as part of how do I improve things for my health and the health of my family and my community et cetera'.

Dr Watson : I do think those things are worth re-exploring with Australians, I truly do. Even within your MBS files, we have no understanding of the health of the person; there is no diagnostic information. In Canada, the diagnostic information of the equivalent file that you call MBS has been in there since the 1980s. That information has not—to my knowledge, in the 15 years I worked in Canada with those files, there have been no breaches of privacy related to the diagnostic information that is in there. That diagnostic information—

Senator MOORE: Dr Watson, can you tell us what 'diagnostic information' means?

Dr Watson : Okay. If you went to visit the doctor today, the doctor in Australia would send a bill to the Commonwealth for bulk billing. What we would know from the MBS data is that it would be used for policy relevant analysis. We do not need your name, so we do not know anything about you; we just know that it is a person of a certain age, who went to see a doctor for a general medical reason because of the billing number. That is it. We do not know—in Canada, you would know the diagnosis or the medical condition principally responsible for that visit. So if you went today and you had diabetes and you went for diabetes related care, we would know you had diabetes and that would be the reason for the visit. This year the National Health Performance Authority released a report that looked at Australians' use of GPs. We could string that information together, MBS, to understand how many times each Australian visited or once or twice or three or 20 times to a GP. We could describe the population in a small area of geography in Australia and the percentage of Australians who visited 12 or more times, or 20 or more times. We cannot tell from the MBS data whether those people were there because they had a fever or they went because they had multiple chronic conditions.

So that information, from my understanding just from 'oral' history—a fantastic word—with colleagues about why we don't have information about someone's health in those files is that people would not want to release that information. If that is true, that must have been a historic context in which the framing of the MBS dataset was created, which I think is worth revisiting, especially in an engagement way. Researchers have ways of doing engagement research with populations to say, 'What is your position with the disclosure of information?' and, 'If it could be used in this way, what would your permissions be?' and then, 'If you knew that this had existed already in another country that has 30 million people, Canada, for 30 years and there are no known significant breaches of privacy, would you change your mind?' Researchers can tell you that by engaging with Australians.

The other part is the MBS and PBS data linkage. That is part of the guidelines for interpretation of legislation, which, as you know, are easier to change than legislation. But changing guidelines means, from the privacy's commissioner perspective, in my view, would mean engagement with privacy commissioners and engagement with Australians because those guidelines must have been set at a point of time in your history that reflected the views of the people at that time. I think that that is worthy of revisiting as Australians views will shift over time on privacy and experiences with breaches either domestically or internationally.

CHAIR: I am really keen to hear from a practitioner about what the implications of such a change in practice would be and how likely that would be positively received.

Dr Elliott : Maybe I can illustrate it just by an example. My clinical area is in HIV. People with HIV have very significant concerns about confidentiality and privacy of their health data. We have been working with people with HIV and their representative organisations for many years and there has been a real shift in the understanding of data and data use. Now that HIV is essentially a chronic, manageable condition in which people can obtain care from multiple different services over a longer period of time, in order to effectively manage someone with HIV there is a huge advantage of bringing data together from many different sources. The specific cases of it being different to, say, linkage of MBS and PBS, are just an indication of the way that that community view has shifted. Those community organisations have been advocates for, essentially, a form of data linkage—of bringing data together from multiple different sources and making that available and useful for both clinicians and people with HIV. So we have a research project now where we have done that. We have created a shared health record with data from many different clinical sources that is pooled and then made available to patients and clinicians. That, I think, is just an example of how the community has shifted.

CHAIR: It is very fortunate to have you here with that particular issue where privacy in the past has been significant in terms of discrimination.

Dr Elliott : It is absolutely at the forefront of people's minds, but it is just that people can see the benefits that can come from bringing data together.

CHAIR: Say I am a practitioner; I am under huge pressure. Now you are asking me to put additional information into the public place. In addition to putting down my item number and processing that through, I am going to have to create more information to hand over. How am I going to receive that as a—

Dr Elliott : For many of these systems, it does not usually require that. It is really about how we use the data we have already collected. We have heard lots of great examples of some of the significant barriers of bringing existing datasets together.

CHAIR: Would the dataset be just a click away? In addition to item number 5753, say, I click and see this is a diabetes related chronic illness. Another click and boom—the information is gone.

Unidentified speaker: It should be.

CHAIR: It should be that simple?

Dr Watson : I would say no, not just a click away. Let me qualify my remarks. Some data should be just a click away, as in Dr Mintzes' example earlier. When you go to a pharmacy to get a prescription for a government funded program, the pharmacist should be able to know and understand the other medications that you are on, for the purposes of harm prevention and optimising the mix of pharmaceuticals for a person. That should be just a click away.

At the other end of the spectrum are things that should not necessarily be a click away. To uphold our responsibilities to citizens not to use their data in ways that they did not give explicit permission for, there should be an authorising process by which we can get their consent for other uses—for example, to take their information from MBS and PBS and put it together for purposes that relate to improving quality of care and better interpreting appropriate and inappropriate variation. As long as it occurs within those authorising processes or authorising environments—those centres we were talking about earlier—you are living up to the responsibility and meeting the conditions under which citizens have globally told you they are willing for that information to be used.

CHAIR: To be shared.

Dr Watson : That is correct.

CHAIR: Professor Pearson?

Prof. Pearson : I want to highlight that I think we need to walk before we can run. There are massive gains in linking the data that we already have. That will only drive an important agenda to understand the health outcomes that we buy from our investment in health. There are two things here: linkages and enhancements to existing datasets. That is what we have been talking about in the last few minutes. Every single dataset—MBS, PBS and all of the other datasets that we use—could have enhancements. I think linking what we already have and then driving that agenda as an important secondary piece is where we need to be going, rather than thinking about enhancing stuff before we do the linkages.

CHAIR: So we should remember that Christmas comes every year and not try to do it all in one.

Prof. Pearson : That is exactly right.

CHAIR: Professor Jorm?

Prof. Jorm : I would agree with that, but I think the 'one click away' conversation got a bit confused. What I was referring to is that I think it is virtually one click away for a GP to have a drop-down list of a set of diagnostic codes that they could choose for a patient they have just seen that would get processed along with the Medicare claim. I think it is probably quite a high priority to think about that, but I totally agree with Sallie—just making it happen with what we have now is an essential prerequisite and we can think about the enhancements after that.

Senator MOORE: Can what you are suggesting happen?

Prof. Pearson : The linkages?

Senator MOORE: Yes. It can happen?

Prof. Pearson : Yes, it is definitely possible. It is happening.

CHAIR: So what would be required to make it happen more efficiently and more effectively? What investments need to be made?

Prof. Jorm : It would need investments in a number of places. I think we have already discussed the need for greater investment in the data custodian agencies themselves, because they are the source of the data that is being linked. There is a need for additional investment in the data linkage units and infrastructure. As I pointed out in my opening statement, I think we are not efficiently using the infrastructure that we currently have. We have these logjams because we have accredited only two agencies to link Commonwealth data when we in fact have highly functional data linkage units in the jurisdictions that could also be linking the data for a specified set of projects.

The other area I think we need to invest in is probably the use of the data and the research using the data. Again, we have touched on the idea of longer term programmatic funding approaches to building research partnerships between policymakers, clinicians and academics to answer the important questions—and not in a piecemeal way where you are funded for two years by an NHMRC grant and then you pack up and go home. Data are linked and destroyed. We really do need to have enduring datasets that we continue to enrich and enduring programs of research using those data.

CHAIR: How would they be established and how would they function in the future? How would you see that happening? How would it start off?

Senator MOORE: Who can make that happen?

CHAIR: Exactly.

Prof. Jorm : The Commonwealth Department of Health and each of the state jurisdictional departments of health could decide that they wish to support a joint program of research to address their particular priorities for health care and clinical services and they could decide to—

CHAIR: Having different people deciding, what is that going to do? Does it need something? I think the word 'framework' was used earlier. Do we need a framework to enable this or do we just need to say, 'Here's the money. Let it rip,' and see what happens? What really needs to happen if we are going to move forward efficiently?

Prof. Jorm : 'Here's the money and let it rip' would work for me!

Dr Watson : Earlier we talked about the centres that exist across the country. Our colleagues are here now from CPR and they can tell you more about that. That history was funded under NCRIS funding. Remember that we also talked earlier about having the locations and having the repositories. That is considered infrastructure.

CHAIR: That already excess.

Dr Watson : Yes, but it is eligible for infrastructure money. It is about coming to work that relates to enhancing and funding the authorising environment and the oral history, and how to use the information systems, and a centre where you could go to find out how to get the authorisation process for your project. That starts to fall out of the eligibility criteria of what is deemed to be infrastructure money and starts to become more operational money.

CHAIR: Or innovation money.

Dr Watson : That is a good idea: innovation money. That is a very good idea. I will have to remember that one. The eligibility for money on the other side, the operational grants for research, cannot necessarily be pooled to create the funding in the middle, because it is getting too close—'Sorry, it's starting to sound a bit like infrastructure to me.' Earlier, in the Manitoba Centre example, I was saying that it would be operational money, but it was always agreed that around 20 per cent of that money—we never counted it, but we knew that it was very explicit—was going to be used to build that infrastructure. One of the deliverables was like an R&D, because that would add to the infrastructure.

CHAIR: So a bit of bridging goes on between the discrete units that are currently completely siloed?

Dr Watson : That is right. Those units are in the ABS and AIHW in each of the states and territories. Where do they go for that money? They would go to the health department or they would go to their own department. I would imagine that, if you are a health chief executive or something in a department, and there is money that you are going to provide to that, they would say, 'What's the sell or return on investment when you want to capture the oral history? I think we'll take our money and put direct it to patients.' Do you see what I mean?

CHAIR: Yes.

Dr Watson : It is a space that we have talked about. There have been threads throughout the conversation today. It is an unfunded space that is part of your coordinating centre. A coordinating centre needs funds. We have those centres, but they do not necessarily have funds for serving that oral history, coordination, smoothing the authorising environment, interacting with privacy commissioners and interacting with Australians to keep a pulse on their changing views and perspectives on privacy and use of sensitive health information. It is an unfunded location. It does not sound like infrastructure, it does not sound like operating grant money for research, and: is it a priority for a government who is here to deliver services to Australians?

Dr Elliott : I could think of three categories of work. The first is essentially addressing the framework by which data is made available from different agencies, which is a lot of what we have been talking about. It is about improving the efficiency and the availability of data and then linking that to some of the other discussion we had about encouraging awareness, definability and use. That is, stimulating a broader culture of engaging with that data, making more and more people aware of that data and making it more and more useful. That is one big block of work. The second is for the Commonwealth government and other governments to really invest in the way that Australians, and particularly public agencies, use that data. We need to keep in mind that there is incredible value in the data that we are talking about, but of course it is not the only data relevant to making policy or clinical decisions about health. There are other similar sorts of datasets around the world. There are other data sources—importantly, data generated through research programs.

So we need to have a capacity in Australia, particularly public agencies, to use the datasets that are becoming available to really drive effective decision making, evidence-informed policymaking or evidence-informed clinical practice—for example, in development of clinical practice guidelines. At the moment, the commission, the NHMRC and the Department of Health have launched a guideline reform process in Australia. That, to me, links quite closely to the discussions we are having here. It is really about how we take data and then transform that into evidence-informed, up-to-date recommendations across the whole of the health sector, whether it is policymaking at a Commonwealth level or right down to the decision of an individual clinician. We have a huge opportunity to improve that cycle. Just last week, I was in Norway and in the UK, where people are establishing those sorts of systems.

The third part of it is really about how we monitor the effect that these systems then have on the outcomes that we are interested in. Is it actually changing healthcare practice? Are we getting a return on investment for our healthcare interventions? Ultimately, what effect is it having on patient outcomes? Those elements can also be collected and understood within these data systems.

CHAIR: I would love to come back to the Norway and UK reference that you just made, but I am trying to think about what you are saying in terms of something that is often discussed at the moment in a public place, which is arthroscopies. There has been doubt about whether that has any efficacy or not. If we had in place the sorts of structures that we have been dreaming of today, in terms of the framework, there would be data collection and data access that would have been in the system many, many years ago and that would have been pointing to whether this is having an impact or is not having an impact. In terms of people—which is your second framework, the people interaction dimension of it—people would know that this work is going on. They are not going to end up doing something that does not help them, so they would think that was probably a pretty good thing if we were able to be aware of it. Finally, in terms of the Commonwealth and the government investment, the early triggering of whether this is an efficacious treatment or a non-efficacious treatment would mean that there would be a saving of dollars or an investment of dollars for health and wellbeing that would be more likely to be effective and of good value.

Dr Elliott : That is right.

CHAIR: That is an example of the way in which it would work and why there might be benefits in a range of ways: to the researchers, to the people receiving the services and to the nation, the taxpayers funding this.

Dr Elliott : Yes, I think that is a great summary. Effectively we have mechanisms for translating data into knowledge and then action that are built for the last century.

CHAIR: It is in a circle, yes.

Dr Elliott : They are incredibly slow, and there is enormous waste that is generated from that. In terms of an innovation agenda, we have innovation generated through the data, through the research, but then the mechanisms by which that is then translated into action are currently based on very slow and patchy systems that hamper the ability to obtain a return on that investment.

Dr Watson : Could I just make a comment on the return on investment area. I liked the conversation this morning with Dr Mintzes and the pharmaceutical example, and I like the Vioxx example, because it is a drug that had class action lawsuits from Canada, Australia and the United States. In the United States, for example, it is my understanding that the class action lawsuit was $4.5 billion. This is a drug that we started to understand from traditional research but that we could have more easily understood from population-based data, which at the time, I remember, was the foundation of the rationale for the establishment of the Canadian network that Dr Mintzes was talking about. It is not just the money and the people with the heart attacks et cetera; it is a clear and easy example of not just waiting for clinical trials and clinical studies with patients on the ground for us to watch our health system to make early detection. In the Vioxx example we had the capacity in Canada and the United States and we started to estimate, particularly after the drug was pulled off the market, the extent of harm that was happening to patients.

We can also link data for pharmaceuticals to understand the benefits. In the atlas that Louisa was talking about that we did the analysis for the National Performance Authority—for the Australia commission—there are very wide variations in the use of antidepressants in the country. What is behind that? As Louisa said, we do not know. I suggest that until we know what the availability of mental health counsellors, GPs and psychiatrists and the prevalence of domestic violence et cetera are from social services data we will not know whether those variations in prescribing rates for people with mental health conditions are reasonable and appropriate. They could be extremely appropriate or extremely inappropriate. We will not know until we put that data together—data about access to mental health services, either from people in the community and doctors, and simultaneously understanding the social context of what is happening in those communities, which we have the data for but it is in another database over there.

So some of those stories were behind the rationale and experiences now almost 10 or 15 years ago of bringing together education, health and social services data in a number of jurisdictions in Canada. I have some of those early examples if you are interested about what for the very first time could be seen about health and care, education and social services outcomes when those information systems were first put together in that country, in Canada.

Senator MOORE: The Rural Health Alliance and the Aboriginal and Torres Strait Islander communities that have given submissions to this inquiry make the point that the basic data without being enhanced and analysed can give you a completely skewed picture. They were pointing that out. So it is interesting you said that. We know the government made statements earlier this month about a new public data policy. Were any of you involved in any consultation or discussion about the development of those policies?

CHAIR: Was that a yes or a no?

Senator MOORE: They were head shakes.

Dr Watson : Maybe I should abstain from that answer relative to my statutory position.

Senator MOORE: Was your organisation involved?

Dr Watson : For what?

Senator MOORE: It was the Australian Commonwealth data policy statement, which was announced earlier this month.

Dr Watson : I will take that on notice.

Senator MOORE: That will be fine. I am just trying to get a general view. The statement has come out and it says lots of good stuff. It talks about optimising the use and reuse of public data, releasing non-sensitive data—all the kinds of things that were involved in recommendations that many of the submitters have made about the need to do it—and making sure privacy is maintained and all that stuff. Now that the government has made that statement what happens next? All of you who work in this field have been talking about this and wanting it. What is the interaction process from now on? You have the government making the statement they want to do this. Is there a link now so you can talk with the Commonwealth ministers and say what you want? Is there a mechanism for you to maintain this conversation?

Prof. Pearson : The pathway does not seem to be that clear right now. It is great news, but we have been speaking for the last hour about some of the impediments to making this happen. I think this is a great step forward but we still have the structures that exist right now and navigating through these things is actually not as clear as it potentially could be.

Senator MOORE: Do any of you have an idea about how this conversation can now be most effective?

Obviously all the years of discussion around the data have got some traction. The government has put out a statement saying that they respect the need for better work in data. That has been done—and this is at the innovation level. I am sure, from your major perspective, it is health. So that would come back through Sussan Ley, and she has been connected with that statement as well, saying that this department will be involved. How do you think the work that you are doing should be best communicated to government and with government so that we can maintain the momentum, link funding to the policy and make sure that we get the changes we need?

Dr Elliott : I just have a brief statement. Again, if you look at what happened with the Obama administration when they drove that open data initiative, it was a statement from the White House, using the administrative mechanisms they had, that the default across government would now be that data will be open. Agencies then needed to scramble quickly to work out how they were going to do that.

Senator MOORE: To change the way they operate, yes.

Dr Elliott : Yes. It was a switch that went, 'We're going from a closed data world to an open data world.' Probably one of the most important things we could do is to say that, from health, there are mechanisms in place that can enable that to occur safely. If we continue to make that sort of statement from a technical point of view—we do not have concerns about moving towards a more open data environment across government—then that perhaps can help reinforce that agenda within government. Obviously then there is a lot of work to do with the specifics of how that is implemented within health, but it is possible.

Senator MOORE: We will try and find out who is going to be doing that through the questions we will ask during the day, and to see what is going to be the action element of the policy statement.

Prof. Jorm : Hopefully some of the innovation funds may also be directed to help make it happen.

Senator MOORE: All of you have mentioned there need to be resources.

CHAIR: And ongoing ones. The unique health identifier in the Canadian context—how do you think that would go in Australia?

Dr Elliott : The Australia Card?

CHAIR: Yes, exactly! That is why I am asking.

Senator URQUHART: So sad, isn't it?

Prof. Jorm : We already have the Medicare number.

Senator MOORE: The eHealth record.

Prof. Jorm : There are conversations around whether or not the Medicare number and its final digit can or cannot be used for data linkage. I am not the one who can give a definitive answer on that, but in some ways we already do have pretty much a unique health identifier. In those countries that have unique health identifiers, they often also still maintain probabilistic linkage—in particular where they are linking health to non-health data, but also to overcome issues of duplicate health identifiers that some individuals have or errors in the data entry of health identifiers. It is a fantastic adjunct to data linkage, but it does not totally replace the probabilistic methods that we are currently using in Australia.

CHAIR: I might ask you on notice, if I can, Dr Elliott—just with your Norway and your UK experiences—if there is anything of usefulness to the committee, in your view. We would really appreciate receiving that from you. Finally, the Medicare Local network was dissolved and replaced by the Primary Health Network. We have spoken a lot about the sharing of information, but, in terms of the Primary Health Network, how do you see that plugging into the systems we have been talking about? How important is data for the effectiveness of Primary Health Networks, in your view?

Dr Elliott : Very important.

CHAIR: Is it connected currently?

Dr Elliott : No, it is not connected. Again, it is about how we use data. You can imagine, of course, data that is collected in local healthcare systems being very useful, and so making that data available in a way, again, that is usable by people within that local area, but then also more broadly, is an important objective that we should pursue. But then there is also how the data is used locally, so how people working at a local level can understand—for example, what is the evidence that I should do X, Y or Z, or what should I do on these things, and then how does that relate to my local circumstance?

I think that is one of the important elements we need to build so that, whether people are working in a hospital, a Primary Health Network or a local clinic, they can understand in their own context what their local data tells us about the way they can apply those guidelines or that evidence. Again, that relates to our ability to utilise the data that is generated at the local healthcare level and then also link people back in to the end of the loop so that they can also contextualise the evidence or the guidelines being presented to them.

CHAIR: Thank you so much for your time this morning and for the time you have invested in this project and in developing your knowledge. After questions today, if you think there is anything that would be helpful for the committee to receive from you—high-level pieces of research or observations that you would like to add to the evidence you have given—I invite you to provide that to inform our reporting, which we hope to be able to do early in the new year.

Pr oceedings suspended from 10:51 to 11:09