Professor issues proprietary e-health warning

Professor issues proprietary e-health warning

Summary: A health informatics professor from Sydney University today said Australia's e-health systems should be strictly open source rather than using proprietary software.

TOPICS: Health

A health informatics professor from Sydney University today said Australia's e-health systems should be strictly open source rather than using proprietary software.

Professor Jon Patrick, a health informatics researcher at Sydney University, reckons Australia is heading down a high-cost and ultimately unworkable path for its e-health systems, which could repeat mistakes of the past.

Patrick pointed to frustrations he claimed clinicians are experiencing in attempting to customise FirstNet, NSW's Emergency Department $80 million information system built by vendor Cerner. "One of their frustrations is that they can't get systems changed to suit local needs and there's no pathway to getting change," he said.

Patrick said the current requirement for "closed vendor" supplied systems cannot support code changes to enact new processes to existing systems. "The only way to get these things changed at will is to have access to source code which implies an open source solution," he told

"They need to be able to get things designed to suit local needs and then changed to local needs," he said.

Patrick's comments follow the release earlier this year of the National Health and Hospitals Reform Commission (NHHRC) which recommended up to a $1.9 billion government investment to kick off e-health in Australia. It had recommended finalising an open technical standards framework by 2012.

However, Dr Mukesh Haikerwal, a former commissioner of the NHHRC, a general practitioner and head of the National E-Health Transition Authority's (NEHTA) clinical unit, said it was too early to debate whether open source was the best fit for e-health. "Before we get to whether open source or proprietary software is better, we need to make sure we get e-health out there. To do this we need the government to commit to the infrastructure first," Haikerwal told

The problem at this stage, according to Haikerwal, was that while, for example, the Prime Minister had blogged about e-health, there was no financial commitment from the government yet.

"If you look at the Prime Minister's blog, he's talking about [health], but we now need someone to get on the white charger like Tony Blair or Barrack Obama did," he said.

But while e-health lacks its Australian white knight, Haikerwal warned the "show stopper" would be if the government got privacy and medical legal aspects of e-health wrong.

Beyond privacy, though a key piece to delivering health care, in particular for remote areas, said Haikerwal, was the government's current $43 billion plan to build a fibre-to-the-home network, which would aid the use of health applications to be developed in the coming eight years as the National Broadband Network was built. Other pieces of infrastructure under development by NEHTA included health provider and patient identifiers, authentication systems, secure messaging and an index of service providers, said Haikerwal.

"As long as you have those components the source of the software doesn't matter," said Haikerwal.

Haikerwal said NEHTA did not want a "captive system", explaining that "we want to use people in our skilful IT industry to give us IT solutions — and we need to make sure that the standards are interoperable."

"If we have the building blocks right, and [across the board] standards signed-off, we then have some surety about a stream of suppliers," he said.

Topic: Health

Liam Tung

About Liam Tung

Liam Tung is an Australian business technology journalist living a few too many Swedish miles north of Stockholm for his liking. He gained a bachelors degree in economics and arts (cultural studies) at Sydney's Macquarie University, but hacked (without Norse or malicious code for that matter) his way into a career as an enterprise tech, security and telecommunications journalist with ZDNet Australia. These days Liam is a full time freelance technology journalist who writes for several publications.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • yes and..

    yes i for one would like to have open source used. However, the development probably would be based upon existing solutions or funded by the government and developed by a community / other based company / entity. The software is not going to be built to comply with standards easily in the free time of developers for a specific industry is probably an issue. (that is alone).

    Of course any system must have an open framework and structure so that it can be interfaced with through other systems.
  • What is the problem with health?

    I have not been following what is going on in this space, however from an IT perspective, is the root cause of problem understood? are we just throwing more money at the wrong problem, more than 70% of project fail because there is no clear mandate, nobody knows what the problem is and everyone wants a piece of the pie. Ultimately tax payer money is wasted, people in the middle make some money, and voters are left with an inferior service.

    A lot more openness is required, from understanding the problem, through to defining solutions including the use of Open Source solutions which by its very nature is collaborative while keeping the end user in mind.
  • Lack of Understanding

    Lack of understanding here - the real issue is not the choice between ‘Proprietary’ and ‘Open Source’ domains it is actually the way we construct applications regardless of the aforementioned domains being used.

    The way applications are developed results in the deployment, functional, massive cost overruns and project failure that we continue to experience with large software projects. Open Source may be ‘cheaper’ but you still have to get all the knowledge into an application which means co-ordinating large number of programmers. Humans and good at design but very poor at coding very large scale applications.

    We need to eliminate the human code effort from the development of large scale applications – applications built this way will be near real, operate on the Web and act as utilities. The outcome is that a health value chain can be designed so that hundreds of entities and thousands of users can share the information – thus eliminating the infeasible ‘messaging’ approach and the hundreds of siloed and stand alone IT systems and support. Also the application could be owned by the Commonwealth rather than a software vendor.

    This can be done today with some vision and commitment rather than being trapped by Proprietary vendors and the illusion of open source.
  • Oh Dear

    There is a very simple mistake being made here. Open Source and Open Standards (what the gentleman actually wants) are two different things.

    Also, the complexity of these applications is such that it would overwhelm the average development shop outside of the vendor.. let alone the problems of code forking, patching etc. You would have a business in a technology dead end in no time.

    Open source is not the panacea the over-excited open source world have us believe. Have an aspirin & a good lie down ...
  • Open Standards are not enough

    I'm afraid that while Open Standards are a good start, they do not address the issue that was raised. When it comes to customising the delivered solution, Open Standards give you at best the ability to completely replace a part of the system. On the other hand, Open Source gives you the ability to make an arbitrarily small change to the system.

    As an example, take an application that has a data entry window that can not be resized or uses a very small font. With a proprietary application that adheres to Open Standards, you have to completely replace the entire software component with something that is compatible. With Open Source, you make relatively minor modifications and continue using the same software without any risk of compatibility issues.
  • More thinking needed

    ...have more of a think... these systems are hugely complex. What seems like "an arbitrarily small change to the system" may have far larger effects that you realise.

    ...and if the vendor can't give you the ability to do this - time to have a conversation ... it's called business..
  • Complex and complexity

    There is now available in Australia a technology known as Sentillion. Clinicians in the Nthn hemisphere love it because it gives you BOB SSO, and is CCOW compliant.
    Healthsmart is snubbing it, yet the clinicians are embracing it, why?
  • Cerner working with Sentillion video
  • Support best of breed Australian solutions

    ISOFT is an Australian company with ehealth solutions with excellent functionality, architecture, interoperability and a published api.

    Why use a solution from a US company "Cerner" when a best of breed solution is available from a local company with an international footprint... and they make great ehealth solutions !!
  • Work for a Govt Health provider and you will understand

    I can only say that they are a wondrous beast that have many heads and of course unlike other workforces the people out on the front line are majority degree qualified and with multiple degrees and post graduate qualifications in some cases. Some interesting egos, power struggles and mix that with the generally high intelligence of the people that drive the demand for applications and you have a volatile mix at times.

  • Yes lets build another unsustainable ehealth model

    Talk to anyone who works in Software Development for a product that is deployed out to multiple customers and one of the key things you want to avoid is multiple branching. Essentially what Jon (who by no means should be considered an expert) wants to do is allow every customer to tweak and tailor the source to their own requirements.

    Then when it comes to maintenance the costs will become non-sustainable.

    What the key issue here is that health providers should not be purchasing products that do not meet their requirements.
  • Place your ads elsewhere loser

    Sentillon has nothing to do with the article. The product is a bogus piece of marketing ploy which has a special eHealth cost tag overhead to it, to do 95% of what industry agnostic products do out of the box much better....

    If you are a purchaser of and you think you need Sentillon, it is probably because you haven't architected your enterpise properly.
  • Bingo!!

    Sounds like someone with experience!
  • How shameful you vendors

    Withdraw your posts now. Follow the rules.
  • Let's put all our eggs in one basket again?

    One of the problems with health is the sheer diversity. One product cannot meet all the needs of all the users in all the health care settings, even for a so-called single business requirement such as ED management across one State. Implementers need to be able to embrace diversity and allow localisation while maintaining standards and the integrity of basic functions. I believe that large scale investment from government should be carefully targeted to a small set of core administration functions for health care services across the board.

    In hospitals, this means a small set of core patient administration functionality, and out of hospitals this means identity management and information sharing using something like an SOA approach. Of course, this core infrastructure needs to be very open so that applications developers can add functionality in a modular fashion. It may even be that an open source approach for this would be the best way to ensure openness and avoid vendor capture. This approach is a better match for the diversity of large scale and cottage enterprises that is health care. As an aside, I think it is fair to say that most hospitals are actually a rather loose collective of much smaller enterprises.

    Government should also provide infrastructure to manage secure communication between health care providers, resources to support development of standards and, importantly, verification and accreditation of software and of IT service providers. It should provide services to allow consumers access to their own health care information, and to investigate and mediate on complaints about privacy breaches. In other words, government should support governance.

    My 20+ years of experience in health IT tells me that government should emphatically not try to purchase and implement large scale clinical systems. The "economy of scale" benefits touted for this approach have proven to be illusory. Far better to encourage local health care authorities and providers to spend a reasonable proportion (I would argue that 5% is required) of their budgets on a major part of their core business - managing information.

    Government should focus on providing an environment that will allow software developers of all sizes and health informaticians to get on with their work, while providing an appropriate level of governance to ensure that appropriate professional standards are maintained. Oh, and clinicans need to develop some respect for people who have devoted their careers to their chosen field of expertise and stop thinking that information technology and information management is easy.

    Open source - closed source - I don't really mind. I have used both in health care applications. What matters is spending the big dollars on the right things - and large scale clinical applications are not the right thing. Leave those decisions to the business units such as hospitals and give them the freedom to choose their product(s).

    Large scale infrastructure to support the technical and social (governance and managing change) aspects of identity management, security and privacy and secure information sharing is what governments should be focusing on right now.

    Jon is an expert in his domain. So am I. CoenBros is clearly not an expert in health care IT because he or she seems to believe that government can successfully purchase and implement large scale health care systems in Australia and this flies in the face of substantial evidence to the contrary.
  • Let's get real here!

    What I find remarkable is that NSW Health when last reported had already got more than 30 sites live with Cerner just over the last year or so, and so far the only real noise has come from a few open source cranks and self-styled “experts”. Perhaps government really can purchase and implement large scale health care systems - much to the irritation of those who continue to believe that doing everything differently at every single hospital is a good idea, and that rolling out a customised system to each and every hospital could be achieved this side of the 22nd Century.
  • Oh yes

    You got that right - it's hard bloody work!! I'm experiencing that right now.. much tougher than any other industry I've worked in to deliver solutions.
  • Old US programs dressed up

    I know the team involved with e-health, they are the same people who ran our insurance systems which were packages written in the US in the 1960s and cheaply maintained.
    The Health Insurance Commission the operators of Medicare/Medibank are absolutely NOT involved in eHealth. A shame as they developed one of the few systems that has come on on time and within budget.
    I have modified insurance software and it appears to have been designed for 1960s tape systems and run on disk. The eHealth systems staff have no exposure to good design or robust coding from their insurance, banking or superannuation experience.
    EHealth is modified Amercian code replete with 50 state flags and 10 province flags that was designed for American style health systems and customised for Australia.
    medicare was written in 6 months.
    There is a saying in IT that if you can't impletemnt a system in 2 years, you never will so throw it away and start again
    Also know Jon Patrick.
    I would be happier with Australian written open source code.
  • Software Development is Broken

    The points made in this discussion thread validates the observation that way we build software applications is broken. The vendors love having everyone buying point solutions.

    Having humans in the software development process means that you will never be able to construct applications to the functional richness and scale required. Trying to wire together disparate ‘point’ solutions across health via messaging and data interchange is technically infeasible and costly – you will always have a legacy system you dare not touch – so much more easy to maintain a single properties file running as a ‘utility’ rather than many millions of lines of code and disparate systems. This is not wishful thinking as it can and is being done today.

    Open Source and Open Standards will not address the fundamental issue of a 'cottage industry' approach to application development.