Data Integration Along the Healthcare Supply Chain - Semantic Scholar

Proceedings of the 34th Hawaii International Conference on System Sciences - 2001

Data Integration Along the Healthcare Supply Chain: The Pharmaceutical Extranet Gateway Project G.M. McGrath JRCASE, Division of ICS Macquarie University North Ryde, 2109, NSW, Australia [email protected]

Abstract The evidence suggests that information systems in most organisations are poorly integrated. We might therefore expect that inter-organisational e-commerce applications will only aggravate this situation. Here, we report on an Australian e-commerce project designed to improve pharmaceutical supply chain efficiency within the healthcare sector. One interesting outcome was that the standards that underpinned the project seem to have resulted in improved levels of data and systems integration - both within and between partner organisations.

1. Introduction Information systems (IS) in most organisations are very poorly integrated. This can be very costly and, in some cases, seriously inhibits competitiveness. The costs involved can be massive. For example, one large Australian company, with an IT budget of the order of several $100M p.a., recently identified that it had 630 separate interfaces, and that 43% of its total processing capacity and over 50% of its (1,000+) programmers were being utilised in processing, developing and maintaining these interfaces [13]. This, however, is by no means the end of the story. Systems developed (or purchased) independently will almost certainly contain inconsistent data. Apart from costs incurred in trying to reconcile this data, customer service may well deteriorate and managers will face great difficulties in using their databases for decision-making purposes. This was emphasised by Laudon, Traver and Laudon [9] in discussing a survey run out of MIT, where the Chief Information Officers of 50 large US companies were interviewed. Almost all of them said that they didn't trust the contents of their databases sufficiently to use them for important decisions. Lee, Conley and Preikschat [10] have identified fragmented IS as a major problem in the worldwide healthcare industry and, particularly, in the e-health sector. Given that these authors estimate the current

E. More Macquarie Graduate School of Management Macquarie University North Ryde, 2109, NSW, Australia [email protected]

worldwide value of the e-health (products) market at $150Billion and that the e-health index is growing at a much faster rate than other key indicators (e.g. the Nasdaq Composite and the GSTI Internet Index), the potential seriousness of the data and systems fragmentation problem should not be underestimated. Poorly integrated IS definitely constitute a major problem within the Australian healthcare sector. Here we report on some outcomes of a recent study of a major Australian e-commerce initiative which had, as one major objective, improved integration of data (definitions and repositories) along the complete pharmacies/hospitals wholesalers - suppliers/manufacturers supply chain. The project concerned is called the Project Electronic Commerce and Communication For Healthcare (PeCC), and for a comprehensive report on all aspects of the study, the reader is referred to [18]. There are multiple projects within PeCC, but here we focus on the Pharmaceutical Extranet Gateway (PEG), an Internet-based facility, developed to allow the automated passing of common order transactions between all parties - and, in the process, to more tightly integrate their disparate systems. During the PeCC study, interviewees at all points in the supply chain expressed concern at the fragmented nature of their IS, plus a good deal of frustration at the fact that accomplishing effective integration of their own IS suites had proved to be extremely difficult - and, in most cases, impossible. Given, then, the difficulties encountered in IS integration initiatives within single organisations, one might expect that inter-organisational IS integration might be close to an intractable problem. Surprisingly, this is not what we found: certainly some fairly significant problems with inter-organisational collaboration were encountered but, in the main, the PEG architecture, combined with a heavy emphasis on standards, ensured that (within limitations) the project's data integration objectives were achieved. Our paper is organised as follows: in the following section, we discuss IS integration and, in particular, problems associated with the more common approaches. In Section 3, we introduce the PeCC project, and then, in

0-7695-0981-9/01 $10.00 (c) 2001 IEEE

1


Section 4, we focus on PEG, its data integration approach and the implications for e-commerce applications in the healthcare sector. Conclusions are presented in Section 5.

2. The IS integration problem Organisations are currently spending billions of dollars on IT and related services and this investment is growing rapidly, with major implications for national economies. For example, a recent report from Australia's National Office for the Information Economy [19], indicated that, over the next decade, e-commerce alone is expected to increase national output by approximately 2.7% and enhance consumption by around $10billion. Now, the increasing ubiquity of e-commerce applications (and, indeed, even local initiatives such as the recent announcement of a major change in taxation arrangements) means that even the smallest company will find it very difficult to operate without some form of IS support. Later in this section, we identify common areas of waste and inefficiency resulting from poorly integrated systems and databases. Quantifying this, on a national basis, is not an easy task, but total losses are almost certainly massive. Some industry examples of major Australian projects attempting to address these issues, but which did not succeed [4], include for example: • During the mid and late 1980s, a major bank embarked on a significant initiative, which planned to substantially and radically replace its critical product management systems with a new and generic system based on data-centred concepts. After substantial investment, the project faltered in its implementation and was shelved. It is interesting to note that the design products of this project have subsequently been sold to some hundreds of banks around the world to assist in the redesign of their systems. • A major telecommunications company has struggled over many years with integration of its customer data, due in very large part to the use of packaged software for its key operational systems (which did not integrate). These problems were mainly due to a "protection of turf" mentality (particularly related to billing) and, in part, to business mergers and reorganisations. • Another telecommunications company has struggled with the conflict between implementing new systems quickly (in a start-up environment) and building for integration. Their continuing need to replace major systems, after 5 or so years of operation, indicates that the benefits of a "green field" environment, seemingly an ideal opportunity in which to promote

and implement integrated systems, does not guarantee this desirable outcome. • In the early 90s, an organisation in the construction business was presented with the value and benefits of a Road Asset Information System, which took a datacentred approach to integrating information about the history of their road pavement projects. The project was estimated to have a Benefit/Cost Ratio of about 8. The project was not implemented and packaged system alternatives were evaluated. They are still being evaluated in 2000. Poorly integrated IS are largely a consequence of the organisational fact that, historically, systems have been built, sponsored or purchased by individual functional areas or divisions: for example, Finance may have its billing system, Sales its orders system, Warehousing its delivery system and Product Management its marketing system. Each of these systems requires its own database, containing details of the organisation's customers, together with their purchase history. Thus, the following extra costs are incurred: • Development or purchase: four systems instead of one; • Processing: four sets of hardware/software platforms, plus the same transactions must be processed four times; • Maintenance: programmers are needed to maintain all four systems; • Interfaces: each of these systems have to talk to each other. Typically, this is accomplished on a 1:1 basis, which means that the interfaces required increases (roughly) exponentially with the number of systems. This may result in very substantial development, processing and maintenance costs; and • A lack of competitiveness: inconsistent and redundant information makes the organisation less effective and more uncompetitive. That is, poor information conformity leads to the same things being defined differently in diverse systems, making it difficult (if not impossible) to match "apples with apples" and hence information performance cannot be reported coherently – and certainly not accurately. In looking at systems and data integration, we go back some 25 years to James Martin [17], who specified the essential elements of a database as follows: i. all data must be consistent and conform to a common definition (or schema); ii. there must be minimum data redundancy; iii. a common approach must be employed for database updates; iv. there must be a clear separation between the data and the applications that use it; and v. the database must provide a foundation for future applications development.

0-7695-0981-9/01 $10.00 (c) 2001 IEEE

2


In a sense this represents nirvana, in that organisations have vainly been trying to implement this type of DataCentred Information Management Environment (DCIME) for over 30 years (see e.g. [16]). Classically, a DCIME is established by: i) using formal conceptual modelling techniques (e.g. entity-relationship or UML based approaches) to develop a Corporate Data Model (or CDM - a logical model of all key data within the company); ii) constructing a set of Subject Databases (SDBs - groupings of data sets, with close affinities, derived from and consistent with the CDM); and iii) building applications around the set of SDBs. Theoretically, this approach makes a lot of sense and, from the late 1970s through to the early 1990s (at least), it was the favoured approach of companies attempting to truly integrate their information systems. However, the problem with the DCIME approach (as discussed above) is that it doesn't seem to work or, more accurately, is perceived not to work. Consequently, organisations seeking to integrate their data and systems have turned to a variety of other methods that are seen to offer short-term success or to cost less. These include suites of packaged applications, new applications development and database paradigms (notably objectoriented approaches - see e.g. [20]), federated heterogeneous databases [7], data warehouses [2] and enterprise-resource planning (ERP). All of these may help to realize integration to some degree and, certainly, may be useful in specific situations. However, there are inherent limitations and costs associated with all these more recent approaches and, despite much commercial hype, none of them represent the "magic bullet" claimed by their proponents. In particular, our observations are that applications of these approaches are plagued by many of the same problems that have caused so many DCIME initiatives to fail. We group these problems into the following five categories: • Strategic thinking: A lack of long-term corporate thinking leads organisations to respond to short term horizons, driven by shareholder demand for early action or a 3 year political cycle that inhibits longterm solutions reaching finality. • Technical/Administrative: A number of key problems have yet to be adequately addressed by the information management community - specifically, how can we allow multiple parties (particularly in larger organisations) to develop applications for, and access, the same set of databases without unduly compromising data and system integrity, security and performance. Migration from a fragmented to an integrated IS environment also represents a serious challenge in many organisations. • Economic: Many organisations have a massive investment in their existing IS. Naturally, they would

like to maximise that investment (or, at least, recoup as much of it as possible). • Timing: Most executives want immediate solutions to their IS problems and are not prepared to sit around for months (or even years) while the necessary SDB infrastructure is established. • People-Related Issues: For example, organisation parties often gain considerable power from their stakeholdings in existing information systems. In general, they are liable to be less than impressed with any proposal that threatens these power bases. We draw particular attention to the last of these problem categories as we believe there is a tendency to ignore or, at least, underestimate these "softer" issues. In our opinion, this is a major mistake and, indeed, we have spent considerable research effort over the last few years focussing on precisely this issue [14]. The ever-growing range of computing and communications technologies generally plays only an enabling role in the development and utilisation of the systems designed to achieve an enduser's business objectives. The building of IS involves, at its core, the still emerging discipline of software engineering as well as other more traditional areas of engineering and science and a number of non-technical management, social and organisational disciplines. In essence, systems are not built in a vacuum, but within organisational environments where outcomes are heavily influenced by a myriad variety of internal and external softer factors. IT managers who ignore these critical softer factors, do so at their peril. For an entertaining overview of the many people-related issues that can have a major impact on IS development efforts, the reader is referred to [3].

3. The PeCC project PECC is the first project of its type in Australia where an entire industry supply chain is being revamped to take advantage of the Internet and web-based technologies. It promises to create major savings to all the participants in the supply chain, particularly the publicly funded hospitals [12; p.5]. Initiated in 1997, PeCC emerged from Federal Government concern over burgeoning costs in Australia’s $37 billion health sector. This multi-stage project was developed and has received support from a number of Federal government departments, but is a joint activity of both government and industry. PeCC was developed to introduce e-commerce practices into the health sector with almost 700 suppliers, automating pharmaceutical and other supplies to hospitals. Supply chain communication is facilitated by an Internet-based platform, allowing more efficient interaction between the pharmaceutical industry’s

0-7695-0981-9/01 $10.00 (c) 2001 IEEE

3


outlets (retail and hospital pharmacies), wholesalers, suppliers and manufacturers. Promoting and demonstrating e-commerce for the pharmaceutical industry supply chain, the project connects manufacturers, wholesalers, suppliers and hospitals. Barcoding every consumable and streamlining the supply chain relating to pharmaceuticals and other healthcare items supplied to hospitals is the basic focus of change in the project. The approach is based on that already utilised for over a decade to increase efficiencies in modern warehousing and retail systems such as databases, barcoding, and having suppliers and customers linked electronically. It is grounded in common numbering systems for products (e.g. European Article Number (EAN)) and in electronic distribution of orders by wholesalers and acknowledgment by manufacturers, using the Internet. The pharmaceutical industry is one of the first industry groups to have adopted a standardised approach to ecommerce. The project’s impact, however, is significant within the broader healthcare industry. As one authority put it: “The project heralds a global transformation of many aspects of health industry administration, putting barcode scanners into the hands of nurses and even replacing the doctor’s hand-scribbled prescription. Every item used in hospitals, from cornflakes to soap, will eventually be covered” [1; p.75]. PeCC is one of the leading edge innovative examples of Australian Internet commerce. This business-tobusiness e-commerce project, as indicated earlier, has been driven by initiatives from both government agencies and industry partners. PeCC reform of the health sector supply chain commenced with the private sector (pharmaceutical companies and private hospitals) but is now spreading to public hospitals. It initially targeted pharmaceuticals but has now also incorporated a wider range of products. PeCC was designed to: • accelerate the uptake of e-commerce, Internet connectivity and the use of the EAN standard numbering system in the health sector manufacturing, professional community care and distribution environment; and • demonstrate supply chain improvements that will become best practice for the management of product, inventory and allied services in the healthcare system [21; p.3]. While initially focussed on the pharmaceutical sector, eventually the supply chain will be extended to include end-users (i.e., hospitals) which will allow pharmaceuticals to be optimally scanned by the bedside on consumption. Once PeCC is fully implemented and with industry products compatible, it is anticipated that those products not complying with the EAN barcoding system and e-commerce/Internet solutions will be

excluded from purchasing panels and electronic catalogues. PeCC itself consists of: a Council – which provides policy direction for the project, meets three to four times a year; an Executive Steering Committee – which decides on budget allocations, provides guidance to the Project Director, meets six times a year; a Project Director; Financial Stakeholders – which includes all PeCC council members and all other organisations that have provided financial support to PeCC; Industry Sponsors – organisations that have contributed financially but are not actively involved in any project; and Advisers. Of the $1 million PeCC budget, 60% has been provided by government agencies and the rest by industry sponsors and project participants. PeCC is Australia’s first industry wide Internet trading community. It has achieved an open standards system allowing anyone to communicate with anyone else, instead of the traditional closed, proprietary networks dominated by IT-strong organisations. Moreover, uniformity across the pharmaceutical industry sector is facilitated through the project. Additional benefits identified include more complete and readily available medical records for individual patients, better understanding of the costs of providing patient care inside hospitals, and improvements in other hospital systems such as patient billing. The most successful of the PeCC projects is the establishment of PEG (the Pharmaceutical Extranet Gateway) whose position in the complex healthcare supply chain is illustrated in Figure 1. PeCC’s initial focus was to link five major competitive pharmaceutical wholesalers (Australian Pharmaceutical Industries (API), Faulding Healthcare, Hospital Supplies of Australia HSA, Sigma Company, and W.H. Soul Pattinson & Co), and the 700 manufacturers from whom they purchase. The CEOs of the five major wholesalers agreed on a handshake and then negotiated buy-in from their Boards for collaboration on developing a common Internet based EDI/e-commerce platform, which would allow them to trade electronically with their suppliers at reduced costs [6]. The partnership that developed has culminated in the five wholesalers collaborating to use standard electronic order forms through PEG. This network has just recently been expanded to include wholesale distributors Clifford Hallam Pharmaceuticals P/L and LJ Cottman (WA) P/L. PEG provides a single common electronic ordering system that allows pharmaceutical wholesalers and suppliers to transact business through the Internet with the use of a common EAN-based bar coding or standardised numbering system. It enables wholesalers and suppliers to send purchase orders and to receive responses across the Internet rather than using the more expensive EDI option. EDI represents an alternative solution but, while satisfactory for large organisations able to invest in technology and skills required for the system, this is not a

0-7695-0981-9/01 $10.00 (c) 2001 IEEE

4


subsidising the program by committing to the bulk of the solution for smaller companies. The newer solution development cost and paying for the operation of the offered by PEG is a single common electronic ordering central facility. system without much implementation time and minimum cost because of Internet utilisation. The wholesalers are Figure 1: HEALTHCARE SUPPLY CHAIN All Manufacturers

Clifford Hallam

HMR

Mayne Nickless

Amrad

DOCTORS GROCERIES Ballarat Pharma

VIC Hos Austin

Upjohn

Gosford Pharma

Ballarat

Baxter

Chemworld

HCoA

Amed

Pharmacist Advice

Gosford

Pharma Pack

API HealthCare

TGH

Chem -Mart

R Hobart H

Terry White Chem

R Adelaide

The Med Shoppe

Woywoy H

Healthsense

Wyong H

HSA

Raw material suppliers

API

SB AstraZeneca

PEG

F H Faulding

SB Sterling

End Users

Sigma Co. Abbott Faulding H.

WS Pattinson

Purepac (US)

PRN-Synergy

Long Jetty

AMCAL

St Vincents

Russells (NZ)

Guardian

Essendon

Com. Drug

Extend-A-Care

Westmead

Andrews Lab

Pharmacy Care

P Macq H

Sigma Pharma

PPS (NZ)

RNS

Soul P. (Man)

SP Chemists

Alfred

Suppliers to the major wholesalers are generally small manufacturing companies with sales of up to $4mill annually and between 20-30 staff. Such organisations are provided with the requisite Internet application software, connectivity, and help-desk support for approximately $50-150 per month. Those already with such facilities pay an annual e-commerce fee for document transmission of between $600-2200, depending on how many documents are processed [15]. PEG is essentially aimed at overcoming problems of disparate databases and the current inefficiencies related to supply chain ordering via fax. These inefficiencies include re-keying orders, lack of confirmation of orders, multiple transmissions, delays and cost factors. PEG involves EDI standard-based forms, in flat file and electronic web browser format. Consequently, it offers accuracy in processing, advanced delivery notification, streamlined payments, and accurate and timely shared business information. It provides a network linking the major wholesalers to manufacturers and suppliers for purchase orders, acknowledgments and payments. Ensuring secure encryption, documents can be tracked through the system. Analysts estimate that the cost of placing an order through the normal manual process is

around $50 to $70. With full implementation of PEG, this transaction cost will be reduced to merely $2- 5 per order. Given a current purchase order transaction volume of 3.8 million per year, this equates to annual savings of the order of $200 million. PEG standards will be used to send orders by hospitals and pharmacies over the Internet; prescriptions may be sent; and PeCC standards will permit pharmacists and doctors to be paid by the Government electronically. PEG allows for replacement of traditional fax transmission ordering. In reality: "It is a single, common electronic ordering system for all wholesalers and suppliers. Small to medium enterprises can trade with their largest wholesale customers without the expensive EDI price tag or a lengthy implementation period" [8; p.2]. PEG’s formation and implementation is leading edge global practice, allowing a group of companies to use the Internet for exchanging messages, correspondence, and product turnover ordering with approximately 700 potential trading partners. Furthermore, the project will enable e-trading between the PEG trading platform and transportation and logistics companies, an Australian first, pointing to the ability to track freight "across docks". The anticipated project completion date is end-July 2000, with the majority of wholesalers and suppliers connected

0-7695-0981-9/01 $10.00 (c) 2001 IEEE

5


A view of the PEG architecture is presented in Figure 2. All parties (at whatever point in the supply chain) communicate with the central PEG databases using common, EDI-based transaction formats and, at the implementation level, through common, atomic-level

(add, delete and modify) database access routines (DBARs). Furthermore, whatever supply systems or packages are employed at individual sites, all transaction details must conform to the PEG Catalog (which is based largely on EAN standards). Sterling Commerce (developers of PEG) assist each of the participants to develop interfaces for converting local data to standard PEG formats. Transactions with data in non-standard formats are rejected and are returned to initiators for investigation and correction.

While not obviously apparent, the PEG architecture presented in Figure 2 has much in common with data warehouses, federated heterogeneous databases and, indeed, the ISO 3-Schema Database Architecture [5]. Essentially, all these approaches to data management are based on interfaces. With PEG, each partner in the venture may employ whatever internal systems they prefer and the interfaces, I/f-1,---,I/f-n, then converts their internal data to a form consistent with the PEG schema (which, in turn, is derived from EAN standards). The interfaces also handle conversion of data travelling in the opposite direction. The great advantage of the approach is that each partner has abundant flexibility in choosing software, hardware and applications to meet their particular needs. For example, if one company decides on a major upgrade of its IS and IT support platforms, neither PEG nor any of the other alliance partners will be

impacted. Required changes will be restricted to that company's interface to the central PEG system. Thus, in a sense, the necessity for inter-organisational coordination inherent in an e-commerce application dictates a system design (based on well-defined interfaces) that may result in a level of data integrity beyond that usually found in intra-organisational information architectures. As observed earlier, on the surface at least, this is a somewhat surprising result. It should be noted, however, that there are some significant limitations associated with this (PEG-type) approach - specifically: • Ideally, as transactions are rejected because of inconsistencies with the central schema and repositories, source systems and databases should be amended to bring them into line with the central system. Experience with data warehouses indicates that most organisations choose not to invest in (the

and able to electronically trade using the complete range of supply chain documents.

4. PEG: data and systems integration

0-7695-0981-9/01 $10.00 (c) 2001 IEEE

6


•

•

admittedly expensive) data clean-up operations of this sort. Costs incurred in developing and maintaining interfaces can be prohibitive (even in an m:1 interfaces environment such as PEG). This is particularly the case for SMEs (such as many of the PEG manufacturers and suppliers). If data clean-up operations are not undertaken,

Finally, the central PEG databases constitute a potentially invaluable information resource that could be used to provide one of more partners with a significant competitive advantage. However, there are risks here as well as potential benefits and these are illustrated in the causal-loop diagram presented in Figure 3. The question the partners must face is should they continue with the current system (albeit, in automated form), where they only have access to their own

•

inconsistencies between partner organisation and central (PEG) records can be a major source of errors and unnecessary work (e.g. reject investigation and other associated corrective activity). The approach does nothing to address data inconsistency and redundancy problems in partner organisations.

transaction data or should they allow each other to access the complete data set. If they opt for the latter course, they will be in an excellent position to use information for true strategic advantage: essentially, because each partner will now have access to information on the total business domain and not just their portion of it. Obviously, however, the risks here are high. So, effectively, those partners that favour data sharing will be backing their own organisation’s IT capability against that of their

0-7695-0981-9/01 $10.00 (c) 2001 IEEE

7


collaborators/competitors. This, in our view, is one of the more fascinating issues still to be resolved among the PEG partners and goes way beyond basic, operationallevel concerns with data privacy and security (important though these are).

5. Conclusion The PEG case, within the wider PeCC project, illustrates one significant effort in IS integration within the health sector - a major Internet commerce initiative, which is especially noteworthy given the level of interorganisational competitive collaboration involved. But the implications go further. First, while internal integration is vital for contemporary organisations, it is not enough in a global economy where partnering and collaboration are becoming the order of the day, and where organisations for survival and growth simply must move beyond their secure internal boundaries in the global marketplace. Second, all sectors in our local and global economies face an increased challenge in such integration, as a consequence of the Net causing “radical discontinuities, catastrophic breaks in the already crumbling façade of business-as-usual [11; p.25].” If ever there was a time for concerted efforts to achieve improvements in IS integration, that time is surely upon us now.

References [1] AH&HCJ, Australian Hospital & Health Care Journal, 1999. [2] Bischoff, J. and Alexander, T., Data Warehouse: Practical Advice from the Experts, Prentice-Hall: Upper Saddle River, NJ, 1997. [3] Constantine, L.L., Constantine on Peopleware, Prentice-Hall: Englewood Cliffs, NJ, 1995. [4] Doll Martin, Unpublished case files maintained by Doll Martin Associates Pty Ltd, Sydney, Australia, 2000. [5] Griethuysen, J.J., Concepts and Terminology for the Conceptual Schema and Information Base, ISO Technical Report ISO/TC97/SC5/WG3, 1982. [6] Head, B., "A Net Cure for Hospital Waste", Australian Financial Review, June 5, 1998, p.45. [7] Hsu, C., Enterprise Integration and Modelling: The MetaDatabase Approach, Kluwer: Norwell, MA, 1996.

[8] Kilbane, D., "Australia Pilots Pharmaceutical EDI Systems to Save $340 Million", Automatic I.D. News, Cleveland, 1-2 Jan., 1999. [9] Laudon, K.C., Traver, C.G. and Laudon, J.P., Information Technology and Society, Wadsworth: Belmont, CA, 1994. [10] Lee, R.D., Conley, D.A. and Preikschat, A., "eHealth 2000: Healthcare and the Internet in the New Millennium", WIT Capital - Industry Report, Jan.31, 2000. [11] Levine, R.; Locke, C.; Searls, D; and Weinberger, D., The Cluetrain Manifesto: the end of business as usual, Cambridge, Massachusetts, Perseus Books. [12] McCrea, P., "Reducing the Cost of Supply Chains", Unpublished paper (available from authors), 1999. [13] McGrath G.M., "An Implementation of a DataCentred Information Systems Strategy: A Power/Political Perspective", International Journal of Failures and Lessons Learned in Information Technology Management, 1997, Vol.1, No.1, pp. 3-17. [14] McGrath, G.M. and More, E., “Forgotten Human Dimensions in Dysfunctional Collaboration”, Proceedings of the 5th International Conference of the Decision Sciences Institute, Athens, 4-7 July, 1999, pp. 1031-1035. [15] McIntosh, T., “Bringing the Health Industry Online,” The Age and Sunday Age, p.13, Oct. 9, 1998. [16] Mackey, J.G., "The Australian Post Office Data Input System", Proceedings of the 3rd Australian Computer Conference, May, 1966, Australian Trade Publications, pp. 139-141. [17] More, E. and McGrath, G.M., Health and Industry Collaboration: The PeCC Story, Canberra, NOIE, AGP DOCITA 8/00, May, 2000. [18] Martin, J., Computer Data-Base Integration, Prentice-Hall: Englewood Cliffs, NJ, 1975. [19] NOIE, E-Commerce - Beyond 2000, AGP, National Office for the Information Economy: Canberra, 1999. [20] OPEN, Object-Oriented, Process, Environment and Notation, URL available at http://www.open.org.au/ [21] PeCC and RCNA, Getting Business Online, Canberra: DIST, May, 1999.

0-7695-0981-9/01 $10.00 (c) 2001 IEEE

8