What facilities should the Telemedical Information Society (TIS) provide to support a global 21st century health care system?

Dr. Simon Smith, European Commission, EC

The use of high technology in medicine is striking yet another dimension, namely the field of legal responsibility with regard to diagnostic clinical decisions. What is the actual legal situation in the United States, for instance, where new technologies have already entered the hospitals? What is the position of the surgeon on the spot?

Professor Richard Robb, Director Biomedical Imaging Resource, Mayo Clinic, USA

Truly a relevant question. At Mayo Clinic, we are very interested in this matter which plays an active role in telemedicine. We're specifically concerned about the legal and commerce related constraints. We're trying to generate a strong lobby in Washington to advance the telemedical cause. If we are facing a concept of regional medicine in multiple institutions which are fairly independent, how can the law be modified so that the various hospitals can exchange data, make diagnoses, and send patient data?
However, once the practitioners decide to implement an innovative technology, we have the basis for creating rational laws to support this. For the moment, there is still a lot of overexcitement about issues concerning who owns what, who controls what, and who pays who for which service. First the technology needs to be in place. Mayo has several facilities in different states, as well as various hospital affiliations. We are practising medicine as a single system line entity. We have figured out various prototypes to answer these questions of control and payment. This is the most effective way to increase health care delivery to everybody.
Legal issues will not stand in the way. Lobbying in any case is necessary towards politicians and health insurance companies to install the different paradigms.

Dr. Simon Smith

What worries me is that telemedicine as a system may produce artefacts or give cause to incorrect diagnoses. Who is responsible in case of error? Don't you think the legalibility issue will form a block?

Professor Richard Robb

This is part of the same problem of who controls what. If something goes wrong in telediagnosis or teletherapy, everyone is responsible as part of the regional practise. Yet right now, the insurance companies are a little reluctant to say that they are going to be responsible if someone in Jacksonville, Florida makes a mistake and somebody else in Rochester, Minnesota pays for it.

Professor Sherrilynne Fuller, Head Division of Biomedical Informatics, University of Washington, USA

One of the other parties in the USA is the Food and Drug Administration (FDA). They care about devices, applied in telemedicine, and make requirements for their use as well as evaluations about the efficiency of telemedical processes. A lot of discussion is going on at the moment about how everything is going to work out.

Mr. Tony King

The mere fact that we are using networks to view images for telemedicine, makes us cross different states and boundaries of countries. How are we to deal with this reality?

Dr. Massimo Luciolli, European Commission, EC

There is still another issue concerning legalibility, namely the various compression techniques of data and images. It is quite common to reduce the size of data using lossly compression. In the medical field, a wrong diagnosis can be made because the full size data has not been consulted. This constitutes a serious problem that needs to be addressed.

Dr. Andy Marsh

You have a high profile telemedicine aspect on the one hand and a lot of underground areas, such as compression, archiving and networking technologies on the other. These have to be brought in as other players in the field, otherwise they will never be integrated. Surely, there are still some others which will have to be considered.
We have several colleagues here from Eastern European countries, such as Hungary, Poland, Yugoslavia, ... What do they think about telemedicine? Is it too futuristic an idea? In one country, I was told: "This is never going to be for us." One didn't even want to enter into the discussion.

Mr. Adam Koprowski, Head of Computer Department, John Paul II Medical Centre, Krakow, Poland

We experience a big discussion about the transmission of medical records and how to include telemedicine in this traffic. Is it possible to connect high-performance technology to simple patient data? We try to construct a network to exchange medical information. How can we exchange data between countries? We ask the European Community and Commission to offer us tools and help us with the development of telecommunication protocols.

Dr. Sanzio Bassini, Cineca, Italy

The problems in Eastern Europe are quite related to those, faced by the majority of the other European countries. The regional hospitals should provide the same technical resources as the large institutional resort centres. We should approach the problem from two sides. First, we have to find a way to let the new facilities be used by basic physicians, performing normal activities, on their own platform. Some applications can immediately be made available to the physicians in regional hospitals. Second, concerted action is needed between the European Commission, the ESPRIT partners and others, to set up successful research projects which have to be regarded as real research projects, as not to confuse the final users, who are health professionals. They have to deal with different options.

Mr. Tony King

What we are actually talking about, are standards. If we look throughout history, the establishment of standards consistently emerges as a service because of mass production. Unfortunately, new standards are very expensive to implement. In the case of TETRA for example, we see mobile users continuing to purchase the existent analogue type of technology due to the huge cost of TETRA high speed communication. It is very nice to have standards and this obviously has to be the goal but they need to be subsidised in some way.
If we consider the great success story of IBM with the SNL standard, it took many, many years before it was introduced. New products, because of their innovative nature, are manufactured at low quantities, and as a result, are bound to be expensive.

Dr. Andy Marsh

We had one example of this. We had a project in the European Commission for 3D heart model views and were working together with a major hospital in Greece. They bought a new machine which has a DICOM output. When we wanted to use the DICOM output, the hospital staff told us, they didn't buy this module because it cost another $100.000 which was too expensive.
The manufacturers don't want to have the standards because if you make a standard DICOM output, why should you produce software on top of that?

Mr. Vladimir Cerny, Institute of Physics, Bratislava, Slovakia

In our country, we have one single laboratory to process image digitalisation for storage in a database. In the new machines which are purchased, we could use the images in a digital form for long term storage of traditional X-ray images. If you are speaking about large scale projects, this is very futuristic for us. What is the situation of a typical hospital? It simply doesn't store images for long term use. Representative experts of DICOM and imaging technologies do not promote digitalisation. Instead, they offer traditional tools.

Dr. Andy Marsh

This is a critical point with a lot of underlying questions that need to be answered. It implies the compression issue and data archiving.
We also have a few good speakers from the USA, so this is an excellent opportunity to listen to their point of view with regard to telemedical facilities.

Dr. Simon Smith

What about the general decision support for the physicians, which is focused on the possibility to get data from Medline? And how is the use of artificial intelligence coming on for the analysis of patient data?

Professor Sherrilynne Fuller

Physicians in our cities don't like us to do things for them, they prefer to do themselves, such as diagnosis. Artificial intelligence based decision support systems are hardly developed. They are not being used, even if they are really good. Guidelines are needed for the embedding of artificial intelligence into the system but not to make the decision for the physician.

Dr. Andy Marsh

In Japan, there exists a system in which the diagnosis is presented to the computer which has to come up with a list of suggestions. The doctor decides which suggestion applies to his patient.

Dr. Denis Mrejen, Member of the board of the French National Health Information Technology Industrial Association

If we turn back to the financial aspect of the matter, is it possible to explain who pays how much for what?

Professor Sherrilynne Fuller

The funding for two projects, dealing with the telediagnosis and teleconsultation parts, is hard to segment out. The funding for the broad health project which initially started out with the teleconsultation programme, was approximately 1.5 million dollars over three years. The second project, referred to as "bench to bedside" initiative, including infrastructure, knowledge resources and security, has received 2.3 million dollars for three years. The development of medical records was performed by the universities over a short time and amounted to 200.000 dollars. We obtained university resources, as well as federal and state resources.
The heart of the problem are telediagnosis and teleconsultation for which we receive some insurance funding to support the specialist consultation. The infrastructure is not being funded at the moment. We don't know exactly where the money has to come from but it is part of the business plan we are developing. In 10% of the cases, the specialist, the primary care provider, and the patient need to be connected simultaneously. If we can do more with basic instrumentation and sharing facilities asynchronously, we can save costs in those types of consultation.

Dr. Gudrun Zahlmann, GSF - National Research Centre for Environment and Health, Germany

Are the medical patient records combined with and linked to specialised information resources, such as Medline? Which standards and guidelines are used for this sort of resources? Does the University of Washington set the standards for the whole region?

Professor Sherrilynne Fuller

From the electronic record, you can always refer to health links. There are calculators for certain kinds of body measurements, which are directly accessible. You can even obtain lab data and information on how to collect samples.Sometimes, the answers you need are directly integrated in the patient record which provides efficient links from one source to the other one.
The guidelines were developed by the physicians at the University of Washington, including some of their colleagues who work outside the physical boundaries of the campus. These standards were considered as being the best practise. Other guidelines, which were approved of elsewhere, were taken into consideration but they are not clinically used in the established practise. The formal process has been developed very carefully to achieve a norm of standard care at the University of Washington.
The hospitals in the areas, which are part of the regional testbed, are totally independent from the university. There is no official hierarchy, only a friendly relationship, to the extent that their physicians might have graduated at the university or that they accept university students. They do not reside under the legal responsibility of the University of Washington.

Professor Jean-François Moreau, Chairman of Department of Radiology, Hospital Necker of Paris, France

Do you have Health Management Organisation (HMO) systems or something different?

Professor Sherrilynne Fuller

We are funded by HMOs. We accept patients from fifteen to eighteen different kinds of insurance plans, including those with HMO allocations. They come to us voluntarily and we collect their payment. The problem is that some plans do not cover certain products. Physicians sometimes prescribe a drug that is not funded by the plan. As a result, the patient probably can't afford it, so we are dealing with a very complex funding situation with all different kinds of rules.

Mr. Tony King

I just like to make certain that we cover all items which we talked about earlier on, to round this forum. We were dealing with the question of standards.

Dr. Simon Smith

In Europe, the information about medical records is totally scattered and this problem has to be addressed in the European Commission. Within the ISIS programme, we have two specific projects concerning standardisation, namely MEDSEC and SEMREC. Both are dedicated to the development of one standard for data and patient record access.

Dr. Andy Marsh

Has anyone considered the problem of languages with relationship to the medical records? Many doctors use professional jargon language.

Mr. Tony King

This is the subject of clinical coding and various clinical specialities.

Professor Sherrilynne Fuller

The differences in communication between physicians and nurses, between medical specialities and various levels of terminology is something we are working very hard on in our academic medical centre by means of a dictionary project. We also have to address the naming and different levels of human organisms, groups of organisms and human relationships. This is a fascinating problem which deserves our full attention.

Mr. Tony King

There is a second issue as well, relating to a unique patient identifier that can be understood and made by a computer. As part of the G7 country group's initiatives, a subgroup 5 has been set up for "enabling mechanisms". It has been charged with the creation of standard interfaces for security, confidentiality and legal issues.

Dr. Massimo Luciolli

We face two problems, namely the storage of data and how the costs of services can be kept sustainable. In other industry sectors, one has stopped to develop ad hoc packages to turn to commercial off-the-shelf components. In the medical area, pioneering is nice but you have to look at the problems of the service and define the guidance to run the service as well as to control the legal and standardisation aspects. For this purpose, it is best to use the mass markets and not to create your own niche market.
The communication costs are growing for all types of services. The security aspects in electronic commerce are defined by the use of cards. We have to investigate which new possibilities this technology can bring.

Professor Sherrilynne Fuller

We are taking anything we can get off-the-shelf but if you're trying to architect an integrated system, you're dealing with the problem of having to purchase packages which are compliant. We have tested a wonderful integrated televideo consultation workstation on our NT-based platform but since it was Windows-based, it simply didn't work and it doesn't right until this day. The pieces all have to interlink in ways that ordinary people can use them.

Dr. Massimo Luciolli

There we come to our second point of the storage problem. What we try to do in the European projects we fund, is looking for several end user applications in the related area. In the automotive industry, they are performing more and more simulation before creating prototypes. Since the data volumes are increasing, they are confronted with visualisation and data storage problems. One project is dealing with hospital storage systems. Each single hospital cannot afford to set up a large database. If hospitals are using standard programmes available on the market and work together, data storage becomes affordable.

Professor Sherrilynne Fuller

We have a standards based open architecture but we are very depending on Microsoft products. We are very lodged with vendors. We are not developing anything if it currently exists in the marketplace. We are working together with computer and science experts on the architecture and the infrastructure. We are not developing as a medical organisation. Still, not everything we need is available. We have certain demands and we cannot afford to wait so we build these pieces and try to link them together.

Mr. Vladimir Cerny

The need of regulations for certain facilities has not been mentioned yet. If you set up a synchronous teleconsultation service, everybody would like to have his case consulted. If you control this service with the budget of the insurance companies, quite a lot of people might be able to afford it, but the price of access may be too high for those who really need it. How can we regulate the service to make it payable for everybody?

Professor Sherrilynne Fuller

We have problems with too many busy specialists who just cannot handle one more case literally. Some of them are totally overbooked for the next year. This difficulty also exists in other parts of the USA. People prefer the very best specialist if it concerns their health. It really gets complicated if the whole world wants to consult the same expert for one single niche problem.

Dr. Tuomo Kauranne, University of Joensuu, Finland

Referring back to the standardisation, the roll-out is going to take something in the order of five to ten years. You have to plan for industry standards five years ahead. It's very often the case in this rapid transition of disciplines that most of the current industries always tend to loose business. So, they're ending up being partners, like for instance Microsoft and Java. Java will be the winner in five years time whereas Microsoft will be the loser. An industry focusing on medical applications, doesn't exist yet. We really have to be working at a kind of careful and ingenuous way to be in the right position in five years. There will be a lot of alternatives in the meantime, just like it happened with computer networks. Yet, all of a sudden, there were these gateways converting one protocol to another and the networks were linked. A standardisation committee decides which gateways are the easiest to handle.

Dr. Andy Marsh

End of round one. Thank you all for sharing this session.