The Fellowship of the Australian Academy of Health and Medical Sciences includes expertise across the spectrum of medical research from laboratory and clinical research to health services and public health research. We believe that access to data is key to initiating and to conducting the research undertaken by many of our Fellows. Indeed, it’s difficult to contemplate undertaking research on how the environment, health practice, and individual behaviour impacts on the health of the community, without access to public sector datasets.
In the Academy’s view, there are several ways that access to public sector data sets can, through research, improve health outcomes for Australians:
- Delivering quality assurance in health: to evaluate the effectiveness of how we are delivering healthcare across the country, and whether we are doing well in comparison to other countries.
- Measurement of research impact: to determine whether we are achieving significant change in health, from changes we make in the way we deliver healthcare.
- Defining new research objectives: to understand, from the problems that data show we currently face, what should be a priority for what research we may do in the future.
For all aspects of health research, access to public sector data is increasingly important. Health systems are becoming more complicated, as we acquire more options in the way we deliver healthcare, and unless there is access to reliable data, we cannot determine the impact of changes we make on community health.
Public sector data can help solve health problems in our society
Two examples where public sector data sets can be shown to impact on health.
We are seeing an increasing trend in travel by bicycle. Data gathered over the last 25 years shows that injury and accidents rates have not changed much amongst younger cyclists, but there are increasing numbers of older cyclists who are sustaining injuries at a higher rate than young cyclists. Data on injuries can be used to start to understand why this is the case, and then to argue for facilities for cyclists on the road that minimise accident risk, including barriers between vehicular traffic and cycles, and dedicated cycle lanes.
In the United Kingdom, in the 1970s a new health problem, dialysis dementia, was emerging but it was not evenly distributed across the country. By accessing public datasets about water treatment, and about dialysis dementia incidence, it became clear the use of aluminium salts to clarify the drinking water in some reservoirs was associated with increased incidence of dialysis dementia in patients on home dialysis. This finding would not have been possible without access to the appropriate public datasets, and lead to a practical intervention to resolve the problem.
Privacy must be maintained
Publicly gathered datasets have a huge potential impact on health through research but there are challenges caused by statutory regulations, a risk averse culture and lack of understanding of how datasets can be used without compromising individual privacy. Privacy concerns can be particularly difficult to allay, in the case of datasets concerning rare diseases. We have to ensure that we do not lose the value of the dataset but at the same time the data must not be so specific that you can identify an individual in a community simply because they are a rare data point in a data collection.
We recognise that legislation may be required to control the use of aggregated data to protect the privacy of the individual. Data for the most part is used in aggregate and de-identified; and it is best practice to do so when this is possible. It is very hard to explain to people that we have these datasets where, in principle, individual people are identifiable, but we can use them so long as identification is not revealed as part of the process. I think that managing this is going to be quite challenging for the Office of the National Data Commissioner.
An example of where public sector data can produce useful results in aggregate is in cervical cancer research. For a while, cervical cancer vaccine registries were being kept by the Federal Government and cervical cancer registries were being kept state by state, so the ability to align those datasets through individual identity was important to determine whether cervical cancer vaccines are working. In this case the data user has to aggregate these data by people’s name—even their Medicare number is not unique—but still be able to explain to people that we are not able to identify any individual when looking at a set of data to work out whether the vaccines are working to prevent cancer.
A clear framework will enable better data sharing
The principles that the Office of the National Data Commissioner promote in the recently released Data Sharing Principles are going to provide a basis for what investigators can and cannot do with public sector data.
My personal feeling is that data use for research should be underpinned by enabling legislation. We need to provide a framework acceptable to the public that defines uses and processes by which people can access and use public data for legitimate research. At the moment it’s a bit of a minefield – there is no consistency when you need to request access to datasets. We need a process that is not overly complex.
Currently, barriers that can arise that are so complex that data users can’t really find a way through them, so having a clear legislative framework in place is critical. While privacy is important, the public is generally comfortable for their data to be used for medical and health research – they understand the potential life-saving outcomes, and indeed expect that we should be using the available data to help with research. Whether their data is being used appropriately is also front of mind. The new legislation should clearly define appropriate purposes of data sharing.
Looking to the future
If access to public sector data on health were improved we could start to get to the bottom of some of the really important problems in health service delivery. For example, there is a substantial discrepancy in life expectancy between the better off and the less well off, even when they live in the same general location. This discrepancy in health would likely be more open to investigation with better access to public datasets on factors that might determine health. Indeed, virtually every aspect of research on health service delivery requires access to public sector data.
We have many data sets about how we deliver health care to the elderly in the public domain. We need to be able to analyse these and look at the outcomes of what we are doing and see if we are getting the best possible outcomes for individuals through investment in their care: data sets from across the country help to examine these issues. Some very innovative approaches to healthcare re becoming available – we can anticipate more use of online consultations, artificial intelligence, and other new methods of interaction between health care professionals and health consumers via the internet; we are going to need innovative ways of evaluating these new innovations. These will doubtless involve access to publicly gathered datasets, to determine whether the new ways of patient care are as effective as the old ways, and how to use them better.
We have a really good health system in this country, and the data we access tells us that we get better value for money than most countries in the world; better access to public sector data will play an important role as we continue to improve and evolve our medical and health system to keep us in better health into the future.
DISCLAIMER: The comments in this article are personal to the author and do not represent the views of the AAHMS or of the University of Queensland.
The Office of the National Data Commissioner wants to foster discussion on data issues that are relevant to all Australians, by collaborating with experts in the private, civil society, research and government sectors to publish their posts and linking to their articles on our website. The views expressed in the posts and comments are not necessarily those of the Office of the National Data Commissioner.
Contribution Data Sharing