Presentations from:
Mr Philip Davies, Department of Health and Ageing
Mr Greg McAnulty, Microsoft Australia
Dr Louisa Jorm, NSW Health
Discussion and debate

Jump to powerpoint and audio presentations

Government Perspective
Mr Philip Davies
Australian Government Department of Health and Ageing

Over the past two days we have heard a lot of the vision of e-health. This paper is to consider from a Government perspective how this vision can be financed. How will we as a nation can build, implement and maintain the sophisticated information systems that we all agree are needed to support delivery of modern safe and effective health. This paper is the view of a bureaucrat, not that of the government and to some extent anticipates my view of the direction of government policy.

It is clear that there will be some essential components of e-health infrastructure that can only be provided by the public sector. However when it comes to the systems that providers use to record and communicate information the case for Government involvement is less compelling.

Consider this outline of a possible national e-health network. With reference to the slide, down the right hand side there are systems used by service providers on a daily basis (eg GP desktop, pathology laboratory, and a hospital information system), the pathology and GP systems are clearly privately owned, and the public hospital system would be a public asset and responsibility. So systems coloured in blue indicate the private sector and green, government funded services. There would be a whole host of other systems on the right eg pharmacies, private hospitals, specialists.

All of these operational systems would need to communicate to exchange information on patients. However to know who the providers and patients are will need a national database, but not one that contains clinical information. They serve as an index. Without them there would be the potential for confusion and misunderstanding. These are a key element of the national e-health infrastructure and would be publicly owned and operated.

The third component of the system (shown on the left) are the electronic health record repository. Any provider can access these store information from multiple sources and with patient permission. Although some may see these as part of the public infrastructure, I believe that these could be privately owned and operated, i.e. a contestable part of the system. Health funds have expressed interest in these as a service for their members or owned and operated by trusted intermediaries such as perhaps the banks or even the Australian Medical Association.

Another core component of infrastructure is the rules and standards that govern the interaction between these components. By definition there can only be one set of rules and standards. This is the task of the National E-Health Transition Authority, a public body to define and promulgate these rules and standards - this is essential to ensure the interoperability of the components in this system.

The vision is a system that contains a diverse range of public and private components working together harmoniously within a common framework of standards. Just as with the road system: Government builds the roads, defines the road rules yet there is a diversity of vehicles who drive on the roads, most privately owned with a few types eg fire trucks and buses owned by the public sector.

Who should pay for these components, both capital and operating costs? We should aim to allocate costs in accordance with the benefits. The benefits of e-health are spread between health professionals, government and patients. For health professionals there may be financial gains through improved efficiency, increased patient satisfaction and greater professional satisfaction. How much should health professionals be willing to pay for the benefits of e-health?

For Government there are economic benefits from a more healthy and productive workforce via improved health outcomes, lower healthcare costs through less duplication of tests and better population health information to support planning and stronger surveillance capacity to combat epidemics and bioterrorism.. Patients appreciate better health outcomes, more responsive services and less risk of error due to poor communications. How much should we expect patients to pay from these benefits of e-health system?

I would be appealing to undertake some sophisticated analysis of the spread of benefits and attempt to apportion the costs in accord with these and ensure that the costs are spread fairly across the components of the sector. I don't think this is technically possible due to issues such as measuring the way consumers would value specific benefits or the variations in efficiency gains among service providers. If the government were to attempt such an apportionment exercise, I would predict it would generate more heat than light and ultimately be a futile exercise.

What are the alternative approaches to sharing the costs of e-health? One approach with superficial appeal, especially to vendors, would be for Government to pay vendors for the cost of system development and provision of systems to providers. Government does not have a track record of success in picking winners and this approach would distort the market, not least by limiting innovation. Another approach, which has been used, is to allow the providers to choose what they wish to buy from vendors, and to subsidise this. This approach has had limited impact, and after all Government does not subsidise pens, telephones and the like, why should it be different with IT. Government assumes that the costs of these items are built into the professional fee and reflects these costs in patient subsidies via Medicare. While this audience may disagree with the way rebates reflect costs, there is a principle that costs are included in the fee structure providers set, based on their assessment of costs and income expectations. Providers could try and work without a telephone however most if not all find the cost is outweighed by the benefits. In the same way the use of powerful IT systems may result in a net reduction of costs or an increase in income. Only the provider can make these judgments, Government cannot. The correct way to handle the costs for professionals in using e-health is to build these into professional fees. We cannot however assume that professional fees will rise as e-health is adopted across the sector. Fees may well fall. There will need to be national infrastructure, which will be funded by government.

We have to recognise that health professionals are not motivated by money alone. Others may find the ideas of netting off costs and benefits and adjusting fees too forbidding and continue with their established model of practice. In a situation of shortage of supply, as we are seeing in GP sector, the quality of service may not be a significant factor in those calculations.

Given the complexity of this why would health professionals bother and not just carry on with current practice. I will propose a few hypothetical hierarchies of motivating factors. Firstly health professionals will be driven by professional price and curiosity. Secondly before too long. Colleges may determine that use of e-health is required to maintain professional standards. Medical indemnity insurers may require e-health to support quality and safety to reduce claims or load premiums for doctors not using e-health. Patients themselves may demand it seeing doctors who can't summon up their hospital records or send an electronic script as outmoded or possibly incompetent. Perhaps a future Government may not offer the same level of patient subsidy for services not delivered using e-health. This could lead to higher Medicare rebates or incentives for services delivered by doctors who are e-health enabled. . Finally if we are convinced that e-health is fundamental to safe, efficient and effective healthcare then governments may decide, at some future date, not to use taxpayers money to subsidise services not up to appropriate professional standards, i.e. it will be required to obtain access to Medicare subsidies. This is a hypothetical hierarchy of sanctions and incentives, bearing in mind that they are all used currently within our system to influence the professional behaviour of providers. Why should we handle e-health any differently?

To recap on the key points:
We can't predict whether e-health will reduce or increase the overall cost of health to the economy;
It would be foolhardy to attempt to apportion the costs of e-health among patients, providers and Government based on benefits received;
Health professionals need to treat the costs of e-health in the same way as other input costs to their practice and use their own assessment of these costs to determine the fees charged;
There are a range of incentives and sanctions already used in the system and these may need to be invoked to ensure that eventually Australians will receive the benefits that modern information and communication technology can bring to healthcare.

Industry Perspective
Mr Greg McAnulty, Microsoft Australia

Microsoft has a vision for the citizen getting seamless healthcare. Many of the building blocks are already in place:

District Nurses use pocket PCs
Doctors at Royal Melbourne access electronic records on a pocket pc
Physiotherapists are using video streaming to demonstrate safe work practices.

These are diverse solutions to a common goal of improving safety and quality of healthcare.

Three core principles:

Connectedness and sharing knowledge
Information driven software
Rich interfaces and new experiences.

Standard and security are critical. An adaptive platform needs to implemented.

Microsoft platform has multiple layers. Benefits extend to improving outcomes, minimising intrusiveness of care and more efficient systems.

Overall Microsoft has a vision of seamless healthcare.

Public Health Perspective
Dr Louisa Jorm, NSW Health

E-Health is having a major impact on public health work and the benefits gained contribute to the case for public funding of e-health services.

The Internet not only provides early warning of epidemics, but also the size of problem, speed and direction of spread. Avian influenza is a case in point. Daily maps are produced by the WHO of laboratory confirmed cases, to help countries like Australia. In the event of an outbreak daily reports of cases are required to be submitted by the Internet. The Internet has made a major difference to the ability to report infectious disease. Such systems are vital for the response to a potential pandemic.

A second example is the Sentinels network of 1200 GP volunteers across France who report on 14 conditions (eg flu, hepatitis, suicide attempts). This commenced in 1984 with electronic collection based on a national network of phone linked computer terminals (MINITEL) and has now progressed to using the Internet. There are no signs that such a useful system can be rolled out in Australia.

Some local innovation is occurring. NSW has a real time surveillance system (PHREDSS) involving 32 public hospital emergency departments initially in Sydney and soon to expand to 8 rural areas. Building on a system developed for the Olympics, funding was provided as a result of the 2003 Rugby World Cup. It operates in the background taking data routinely collected in the ED department using HL7 messaging. It has no work impact. The central data management system analyses presenting problem and nursing and medical staff assessment. It has provided early warning of the influenza peak. Work is underway to improve its functionality. Technically it automatically classifies the free text collected by staff and using "Na ve Bayes Algorithm" for text classification. This allows real time analysis of the information entered by staff and correlates well against formally coded data, although producing higher absolute counts.

In the SARS epidemic electronic communication played a significant role. In addition to the use of IT by health professionals, a range of self-help early diagnosis websites sprang up to assist the public. A similar service has been developed for flu. These types of services point to ways we can manage to help people exposed or under home quarantine, limiting need for labour intensive phone follow-up and keep people out of doctors surgeries minimising potential for spread of infection. In a recent simulation, daily follow-up of workers exposed to avian influenza proved to be a major problem and overwhelmed resources once more than one poultry farm became involved. From what we have seen in examples like this the best possible public health outcomes may not occur if such monitoring services were developed in the private sector. Another service, which popped up in the Hong Kong epidemic, was a mobile phone text alert for infected buildings, so people could keep away. The impact of this service is not known. What is remarkable is how quickly services like this can be created and launched.

The recent US Hurricane Katrina lead to the loss of medical records and since there has been an intensifying of the marketing on online patient managed records. We have heard of several Australian examples. One risk of the move to private sector systems of this type is the potential lack of access to this information, or inability to link these systems to form a comprehensive record if consumers decide not to participate through opt in or out mechanisms.

There has been a lack of focus on medical research in this meeting. The EHR has potential to capture outcome and follow-up information, however patient access components could assist with publication of the existence of trials and recruitment.

Confidence in consent and privacy and security is critical if consumers are to agree to the use of electronic information systems in research. Paradoxically with the ability to link data in WA the number of projects which can use "de-identified data" is growing with a parallel decrease in the use of identified data.

It is clear that many of the public sector benefits that could emerge from e-health may not happen with a proliferation of private sector services. The idea floated by Philip Davies of private sector EHR repositories begs the question about the ability to access the data contained for public health purposes and research to augment individual patient care.

The design of systems must consider the public health and research components as well as those of direct patient care. Public health practitioners must become more involved in this area. We found with the PHREDSS system that major workarounds and compromises were required which could have been overcome had some public health expertise been available during the design phase. Secondly there is a lot of free text information contained in electronic records and we might be able to make more use of this using sophisticated text analysis techniques as seen in the PHREDSS services.

As with system design public health and research interests have to be considered during the development of frameworks for consent and security. There are some major advantages for "opt out" models to support data collection for research or public health.

Discussion and debate

There is clearly a need for strong Government - industry partnership in these early days of e-health particularly around the infrastructure requirements as has been seen in telecommunications, energy and transport sectors. There are clear signs of market failure due to early adopters being unable to harness the benefits of their investment with the benefits accruing to those who enter the market later. (Michael Georgeff) This view is absolutely right, it is just a question of where do we draw the line or the degree to which Government gets involved. (Philip Davies)

Banking has had to report to central bodies to allow research and also protect privacy. There are ways of working with public-private cooperation. There has to be a framework to allow this reporting to occur. (Peter Roeleven)

Given the degree of uncertainty about Government direction and funding of e-health, there is the issue of sustainability of expensive proof of concept projects such as the Eastern Goldfields Reference Site. What happens afterwards? (Julia Nesbit) At one level this is an example of inefficient Government involvement, however these are trials and by their nature trials have to end. The ideal trial would be one that could be self-sustaining at the end. If it were not, then to replicate it around the country would require a comparable ramp up of public investment. Hence it is important to be clear if a trial is about demonstrating self-sustainability or some other objective. It is hard to stop such pilot projects when their funding ceases. (Philip Davies)

It was interesting that the GP Research Network was not cited as an example of an existing private sector approach to harnessing the data in GP EHRs. (Tom Bowden)

How ready is the public health system to communicate daily with GPs in the event of a pandemic? (Tom Bowden). Not nearly as ready as we would like. This is one of the downsides of not having a comprehensive provider index. Using commercial services it took a week to get a list of fax numbers during the SARS epidemic. (Louise Jorm)

The statement that it is impossible to measure and apportion the benefits of e-health is at variance with international work on this issue. Studies from the USA (accepting the use of caution in comparing with Australia) find that the benefits of e-health are measurable and significant and may justify an investment of up to 5% of health expenditure. Who gains from this has been worked out and as you point out is distributed across patients, doctors, government and the community. Who gains from less duplication of testing, not the laboratories or referrers. It is Medicare or the public hospitals. Who gains when people are healthier? The economy gains - and the tax revenue base. Who gains from the reduction in public hospital admissions? The jurisdictions and indirectly the Federal Government. It is a strange idea that patients should pay more to have their doctor use e-health so that they have better health outcomes or their doctors will do less tests. (Peter Garcia -Webb)

It was suggested that one role of the Government is to provide the infrastructure. All we heard about was the national provider and patient index (these things have largely been done already and making then available adds little cost to Government). Likewise with Broadband. It appears that the level of investment required by Government is out of proportion to the public benefits of e-health. (Peter Garcia-Webb) The Government does not provide other equipment or means of production; they are covered in the fee structure. The AMA has been a strong supporter of the ideas that the relationship is between doctor and patient. The Government role is to support access to needed services by subsidy which in the case of general practice is around 80%. The costs of e-health are just as likely to be negative as positive, and they should be built into the costs of your services, and may eventually flow through to the federal government. (Philip Davies)

Lots of software and resources developed in projects with public money and would be beneficial to the whole industry and support standards. The software industry can't get access as the public servant has changed and it is impossible to get decisions made. (Vince McCauley) This is surprising and I will look into it. Support the idea as long as the Commonwealth has the IP and you are prepared to take responsibility (Philip Davies)

I have heard of figures that 0.5 - 1.5% of GDP may be saved by the improvement in health information and communication systems. My costings support these. It would be preferable to work together not use compulsion. There is a lack of suitable IT systems for specialists (Tom Weinkhardt) Compulsion is the last resource and hopefully wont come to that. There is somewhat of a standoff between IT interested but wary health professionals and vendors who have to account to their shareholders and are uncertain about the market. We need to have some way of providing confidence to the market that e-health is the direction we are moving in hence the idea of compulsion as a last resort as the government in 5 or 10 or 15 years wont tolerate the delivery of services without e-health support. This is a strategy to try and overcome what we see as a market failure. (Philip Davies)

The system diagram proposed by Philip Davies simplifies the system past the point of usefulness and misrepresents the reality. We lack "enterprise wide" systems in Australian healthcare which can provide the 24/7/365 level of service availability and interconnection which would be required to participate in a HealthConnect. The public hospitals and large organisations eg lab and radiology practices may be moving there but the vast majority of GPs and specialists are not. (John Rimmer)

The costs and benefits of change fall unequally. This happens in all industries and the degree to which an industry sector can advance in ICT uptake is dependent on whether this issue can be solved. The component of the health system that is least likely to see the benefits are the private medical practitioners. The advent of electronic reporting provides huge efficiencies for the diagnostic imaging and pathology industries, but little in efficiency gains for the referrer. The distribution of benefits and costs in a change process needs consideration.

A few ways to move forward include:
Compulsion - if you want to deal with us then do it!
Collaboration - public and private partnership as seen recently in transport and logistics areas lead by the Deputy Prime Minister and industry and promoting change management . (John Rimmer)

Do patients in NSW public hospitals consent for their information to be put into PHREDSS? What is the impact on privacy (Rod Pierce) No identifiers are collected and this information is heading there anyway, now just fasts. Clearly privacy is a key factor, however we have better ways of doing surveillance and research without identified information. (Louisa Jorm)

There is a huge disparity between the uptake of EHR in GP sector as compared with public hospitals. (John Frost) There are huge systems problems with regard to information management and e-health in the state public hospital systems, could this be dealt within the Australian HealthCare Agreements (AHCAs) (Jill Maxwell). Until NEHTA get the standards delivered it is understandable that states may be holding back. The AHCAs are not by their natures purchasing agreements they are funding agreements. It would be one way if the Commonwealth wanted to go down that path but it would be preferable for the States and Territories to come to their own decisions about moving forward. (Philip Davies)

The NEHTA work program is in the public domain (Philip Davies) The work program is established and is on our website. It was established at the time of our establishment in July 2006 and worked on over the preceding 12 months. It has been taken around to over 20 conferences around the country and contains few surprises. There are 12 initiatives and significant progress will be made in 2006. Some things have been delivered eg Hospital Discharge Summaries, which will be trialed in 2006. (Lisa Smith)

There have been some outstanding successes in government funded IT programs. Medclaims (electronic claiming between doctors and the HIC) was one case, very successful rollout to over 6000 practices with real benefits to the government in efficiency. Compare with HIC on-line as the benefit was not clear. With some appropriate initial seeding incentives the initial slow progress was much accelerated. Seeding exercises have worked well. Eclipse has only engaged with 200 specialists, presumably the other tens of thousands do not see any value as yet. (Paul Doman)

The role of the Government is to adjust the result of the cost-benefit decision for each practitioner. Using e-health is not a "no-brainer" for the average medical practitioner. We would rather increase the benefit to practitioners rather than supplying supply side subsidies. Driving by stimulating demand not subsiding costs (supply) (Philip Davies)

Government policies can also confound uptake of e-health eg the need for physical signatures. The policy environment needs to be analysed and corrected if it is not to be a further barrier to e-health. (John Youngman/Philip Davies)

To achieve the benefits of e-health requires a certain critical mass, just as happens with the "herd immunity" for disease prevention by vaccination. Unless sufficient people are online there will have to be a continuation of paper and duplicated systems. (John Youngman). This is also seen in the early vs late adopter issue. If everyone knew that e-health would be all pervasive in 10 years then the risk of being the first cab off the rank would be lessen than it would be without a defined end point. (Philip Davies)

There is significant evidence from RAND study and others that significant benefits will occur in the longer term in getting evidence into practice. To achieve these requires immediate action and short term investment. (John Youngman). Access to guidelines and evidence in the practice may be cost increasing or decreasing. Hopefully it will lead to quality increase and this may justify additional costs if present (Philip Davies)

There is an implied problem with doctor resistance. The AMA believe this is a myth and that the profession that instead deserves recognition. The profession established the GPCG, the profession (via the AMA) that formed the specialist IT group to link up the Royal Colleges. The profession has been driving the health informatics workforce/capacity issues and who drove forward security and privacy guidelines and their implementation. The profession drove forward the lifting of requirements under the Practice Incentives Program. The profession has responded to behavioural change programs. Is compulsion likely to be necessary.