Internet technologies in hospital information systems

Citation

Farkas, Z. (1999), "Internet technologies in hospital information systems", Internet Research, Vol. 9 No. 2. https://doi.org/10.1108/intr.1999.17209baf.003

Internet technologies in hospital information systems

Internet technologies in hospital information systems

Researcher: Zoltan Farkas.Postal address: Department of Information Systems, University of Veszprem, 8200 Veszprem, Egyetem u. 10. Hungary. E-mail: farkas@almos.vein.hu

Along with the development of new information technologies, there is an ever-growing need for Hospital Information Systems (HIS) that cover all possible areas of medical information processing. Existing HIS systems cannot solve this problem. Their server side is mostly based on a traditional relational database that cannot store and supply multimedia data (X-ray images, etc.) effectively, while their client side often uses peculiar user interfaces that are difficult to govern and are usually not platform independent. To overcome these problems, our research team has been developing a new HIS, that is based on Internet technologies to make it more effective and user-friendly.

A Hospital Information System is, in essence, a database application, that stores patient information and is modular in structure. This modularity is not a necessity, it just makes the system easier and cheaper to implement. The core of our HIS (called MediNet Millennium) is a new generation object-oriented database that can effectively handle multimedia content as well as traditional information (characters, integers, dates, etc.) This is a critical issue, since most of the patient information is actually multimedia data ­ X-ray and other images. We have investigated extended relational and object relational databases as possible competitors, but found true object-oriented databases the best solution for our HIS. Some post-relational databases with an object-oriented interface already exist on the market but, as its name suggests, this interface is just a link to the underlying native database, giving the product an object-oriented look, but making it actually slower (according to our first performance measurements). Object-oriented databases currently on the market have been targeted to the Internet or, more precisely, to the World Wide Web, where large amounts of multimedia information must be stored, and data access must be as fast as possible. From this aspect, WWW and modern HIS seem similar, i.e. both handle the same type of data, and end-user data access characteristics are similar, too. As a sub-project we are now trying to verify this theorem by comparative analysis of existing WWW and HIS systems.

We designed a three-tier architecture in which the database (or databases) is linked to a Web server that is responsible for managing queries and supplying requested information back to clients. Web servers are the logical tools for doing this, because they are inexpensive, are implemented on most hardware platforms, can connect to multiple data sources and can work with common browser clients (like Netscape Navigator or Internet Explorer). Apart from this, they are easy to set up and maintain and some of them can drive teleconferencing and multimedia broadcast, thus giving a foundation to telemedicine and teleconsultation. By using Web servers to link clients and databases, a high level of modularity and scalability can be achieved. This is well documented; the most obvious examples for this scalability and modularity are the Web sites of well-known, large companies (like IBM or Microsoft) and Worldwide events (like the Winter Olympics at Nagano, Japan).

The third tier in our system, the client side, is implemented via the use of common Web browsers. Using them has many advantages: since they are designed to communicate with Web servers, they are the logical client-tier choices in our HIS architecture. Because they were designed to be used by "any" Internet user, Web browsers are indeed easy to use and are not linked to a specific Web server product. Apart from accessing and retrieving data from the HIS databases, they can be used as the client side for teleconferencing and can, of course, access data (including medical resources) on the Internet. Web browsers' free availability and platform independence are also very important factors because, besides usability, the most important factor affecting the implementation of an HIS is its Total Cost of Ownership (TOC). In case of our HIS, the TOC is greatly reduced, owing to the use of inexpensive but reliable Internet technologies. Web servers and browser clients are free or cost virtually nothing (compared to the price of a single PC for instance), while their continuous development is guaranteed for all popular hardware platforms. Certain versions of Web browsers can be run on older PCs, which is a significant benefit in Hungary, where most hospitals are not equipped with state-of-the-art hardware. In fact, most of the technology needed to set up a three-tier HIS like our MediNet Millennium already exists, freely available and well documented; moreover, these technologies are being continuously developed by various software companies, so long-term state-of-the-art availability and inexpensiveness seem to be guaranteed. Investment in software is only needed for the object-oriented database application.

As for investment in hardware, instead of buying robust and expensive database servers, Web server technology in the middle tier allows for distributed data storage, making the entire system decentralised ­ thus making it more error-tolerant as well. Another important issue with HIS is security ­ data access must be controlled and monitored carefully. In addition to local security, since the Web server based database access allows hospitals to link their patient databases, we must pay attention to Internet security concerns as well. Using the technologies developed for secure electronic commerce on the Internet (and basic Web security itself) such as Secure Socket Layer (SSL) and Secure Electronic Commerce (SET) security requirements defined by HIS standards can probably be fulfilled, but further research is certainly needed in this area.

Project status

Our HIS project consists of two parallel development directions: design and implementation of the object-oriented database and of the Web-server based client access components. We are running two subprojects as well: one for comparing WWW and HIS traffic characteristics and another for investigating security issues. We are almost ready with the Web-server based client access components and with the first subproject. The second subproject that deals with security issues is scheduled to start early 1999.

The most complex part is, of course, the design and implementation of the object-oriented database. The data model was expected to be ready by the end of January 1999. We have two options for the implementation: either we build the final product completely on our own, or we cooperate with a business partner. In the second case, we should rebuild an existing relational data model on an object oriented basis. In 1999 Hungary will adopt the European Union's new Hospital Information System Architecture standard (HISA) and companies that would like to remain on the HIS market here must upgrade their products to conform with HISA. This will, among others, mean that their data models must be revised and probably changed to fulfil HISA requirements for modularity and interoperability, which opens a window on possible cooperations. In either case the implementation of the data model must be done by mid-1999. Once operational, a detailed evaluation of the system and a comparsion to market leading traditional systems will follow. Our research team at the University of Veszprem, Department of Information Systems works on this project in close cooperation with the local Veszprem County Hospital and with HIS provider companies.

European Health Telematics Observatory (1998) http://www.ehto.be

IBM Co. (1998) IBM at Nagano http://www.olympic.ibm.com/building/ns_jar_e.htm

Kékes, E., Kincses, Gy. and Várhelyi, T. (1993), Egészségügyi informatika, Springer Hungarica, Budapest.

Klein, G.O. (1997) "Standards ­ a link between R&D and the real world", Ehto Journal, Vol. 1 No. 5.

Recommendation No.R(81)1 (1981) on the regulation for automated medical data banks, Committee of Ministers of the Council of Europe, Strasbourg.

Recommendation No.R(86)1 (1986a) on the protection of personal data used for social security purposes, Committee of Ministers of the Council of Europe, Strasbourg.

Recommendation No.R(86)20 (1986b) on the harmonisation of laws relating to the requirement of written proof and tote admissibility of reproductions of documents and recordings on computers, Committee of Ministers of the Council of Europe, Strasbourg.

Simon, P. (1990). "The question of development of the information systems on public health", MIE'90 Proceedings, Spinger Verlag, New York, NY.

Simon, P. and Vincze, I. (1989), "Informatics in the hygiene network in Hungary", Proceedings of the International Symposium of Medical Informatics and Education.