Doctor Doctor Give Me The News

Since the days of Hippocrates doctors have jealously guarded the learning which underpins their status and privilege. If ever there was a profession that lived the phrase "Knowledge is Power", theirs is it. But the same clinicians that are so loath to share their wisdom with others typically value knowledge and expertise almost above all else, while the health system that supports them rewards them for learning on and off the job. Now a highly successful project using the Internet to make up-to-date medical evidence available at the point of care is turning clinicians' attitudes around. In addition, it encourages widespread sharing of information while improving the quality of patient care.

Implemented in very short order, with a relatively low budget and using simple technology, the system represents a world-class example of how knowledge delivered using the Internet can be used to support knowledge workers. Not bad for a project that was originally seen as no more than providing a low-cost method of getting information to rural clinicians, and not as a knowledge management strategy at all.

A clinician's access to knowledge can spell the difference between life and death, an accurate or erroneous diagnosis, early intervention or a prolonged and costly stay in hospital. Yet medical research findings are slow to change medical opinion and practice, not least because the sheer volume of medical literature makes it impossible for clinicians to keep up with all the advances in medical research, or even to make sense of often-conflicting research findings in particular areas. Empirical studies show that there's commonly a time lag of between eight and 13 years (depending on the specialty) between a treatment being proven to work and its adoption in common practice. Worse, there's scant evidence 70 per cent of treatments currently in use are any more effective than doing nothing.

It is recognition of problems like these that led to the discipline of evidence-based medicine (EBM), an application of knowledge management principles in the medical field that predates current knowledge management literature by more than two decades. EBM aims to bring research and practice closer together and thus reduce the time lag between the development of clinically proven treatments and their use in everyday medical practice.

But synthesising research evidence is only the starting point for the use of research to improve practice. Knowledge isn't knowledge until information is disseminated and used. That means making systematic reviews readily available to medical practitioners, and encouraging their active use in everyday clinical practice. That's what has been achieved at NSW Health, which had created a world first in using the Web to bring clinical decision support information to the point of care for 78,500 doctors, nurses and allied health professionals working in hospitals, community health centres, general practice, ambulance, and Corrections Health.

NSW Health's Clinical Information Access Project (CIAP) provides a comprehensive range of knowledge databases, clinical practice guidelines, protocols and policies to support evidence-based clinical practice, research and education. The system operates using a Web server located at head office North Sydney with management of medical reference databases outsourced to an external organisation. Usage commonly peaks at more than 25,000 hits an hour.

The project has proven so successful that in 1998 CIAP received the Data Management Association (DAMA) Australia Achievement Award for Excellence in Information Management, and the Australian Library and Information Association (ALIA) NSW Branch Merit Award for Services to Rural and Remote Users and the Community.

According to David Gates, general manager (CIO) of Information and Asset Services, at NSW Health the project was initially conceived as a way of bridging a significant gap in clinicians' knowledge base, particularly in rural areas.

"We never saw it as a knowledge management project, to be truthful. That came to it later," he says. NSW Health has its own knowledge management strategy. According to Gates, CIAP was never part of that strategy. Rather it was [meant as] a low-cost method of getting information to rural clinicians.

Initiated by the Clinical Systems Steering Committee, CIAP has revolutionised clinical practice by bringing a comprehensive range of quality, peer-reviewed information to the point of care. The information is available 24 hours a day, seven days a week via a Web site established on July 4, 1997.

Clinical systems manager Dianne Ayres says use has increased exponentially since its inception. Clinicians universally praise CIAP and demand for education in how to use the databases is continuous. "CIAP has also proven to be a catalyst in developing a clinical information culture within the NSW public health system," Ayres says. "If feedback from clinicians is any measure, the value of this information resource as an essential tool of modern medicine is now embedded in clinical culture."

Share and Share Alike

The project had two objectives. The first was to create a comprehensive organisation-wide repository of medical knowledge - a virtual "library" accessible 24 hours a day, seven days a week to clinicians across the state. The repository's main component was external knowledge, in the form of medical reference databases and full-text journals and textbooks, but it also contained structured internal knowledge in the form of clinical policies and protocols. The second objective was to encourage sharing of knowledge between clinicians across the organisation via e-mail discussion groups (listservers) and posting of clinical policies and protocols for peer review.

Gates says the award-winning project has been so successful largely because it was developed by clinicians to meet clinicians' needs. Clinicians also embrace it because it is so easy to access through the Internet; however, this has proven rather a mixed blessing. It's a plus because it means clinicians can be connected both at work and at home. It's also a minus, since it opened up a range of issues surrounding security and access into the internal networks.

In addition, at many participating organisations it was the IT managers who put up the strongest barriers to change, largely due to the culture and structure of the organisations. "In Area Health Services with strong clinical leadership, the value of the Internet as a tool to deliver information to decision-makers to influence health outcomes was well accepted," Ayers says. "Conversely, in some organisations the control of this resource is often put into the hands of IT managers and directors of finance. A common perception of these managers was that the CIAP is a 'nice to have', rather than an essential, tool of clinical practice. One manager stated that 'the CIAP is just a toy for clinicians to play with'."

Resistant IT managers also said clinicians would access inappropriate information on the Internet, or waste time "surfing the Net", instead of caring for their patients. "This was an inappropriate response and demonstrated a lack of understanding of clinical processes and the critical nature of the decision-support information provided," Ayers says. "Clinicians are professionals, accountable for their own clinical practice and professional conduct."

IT directors were also concerned about infrastructure issues, not only in regard to the need to support additional PCs, but also when it came to issues of gateways and security of internal networks. These issues had to be progressively resolved before IT managers were prepared to fully endorse the project.

"Another pitfall was just the inadequacy of our infrastructure base, particularly in rural areas, which can't be underestimated," Gates says. "All health information is going through an explosive period, with client/server technology coming in and a whole lot of new systems, both admin and clinical, competing for bandwidth. That explosion is adding to the load on both our IT skill base in the country and our infrastructure," he says.

Part of the answer was to form a rural IT forum, to develop methods of collaboration between more advanced metropolitan areas and their rural counterparts, and to gradually move to frame relay technology. There's also a growing investment in rural infrastructure. The project team also devoted considerable effort to creating a communications network to market CIAP and funds to teach clinicians how to use the Internet and search Web-based knowledge databases. Since effective communication between the NSW Health Department, clinicians and management was pivotal to the project's success, a CIAP representative in each Area Health Service liaises between clinicians, management and the Department of Health.

Another barrier was cost, with most funding for IT directed towards organisation-wide applications and maintenance of legacy systems. With a few notable exceptions, penetration of PCs located at the point of care was low, with most applications that provide clinical information being mainframe-based. "Clearly, the value of knowledge used by clinicians to influence decisions that will improve the quality of patient care is not inculcated into organisational culture," Ayers says. "It is also difficult to provide a cost-benefit analysis in this type of project as benefits are qualitative and few measurements exist to communicate the value of Internet technology."

Yet the system proves itself every day. For example:n A clinician in the North West Area Health Service used treatment information from MedLine to save the life of a patient in a critical condition suffering from the Lyssavirus. A relatively new disease acquired from contact with bats, Lyssavirus as yet appears in no medical textbook.n In the Far West Area Health Service, a clinician was able to save a patient with meningitis who was not responding to treatment. A MedLine search revealed a new drug treatment, which he applied with positive results.n In the North Coast Area Health Service, a clinician was able to save a child with a spider bite by looking up the poisons database in Micromedex.

Clinicians That Get IT

Gates says the system's implementers had to work hard to develop an information culture among clinicians. Part of the answer was to set up a network of IT-literate clinicians to introduce the system to their peers.

"The key difficulty we've had in clinical information really is the ability to get clinicians to use the technology. And it's not the question of not having the tools there; it's really a question of take-up. And what this did was provide a very simple method of clinicians seeing real value, so it was a very easy method of developing that culture change," he says.

Proof in the Pudding

Volumes have been written about knowledge management (KM). Yet there are few detailed empirical studies of KM projects on which CIOs can draw in implementing such projects. This is especially true of public sector KM projects, despite the fact that the public sector comprises up to 65 per cent of the economy in some industrialised countries (40 per cent in Australia).

That is what makes so significant the report by the University of Melbourne's Daniel Moody and Graeme Shanks entitled "Using Knowledge Management and the Internet to Support Evidence Based Practice: A Medical Case Study". This landmark study into NSW Health's Clinical Information Access Project (CIAP) not only provides valuable guidance to CIOs looking to implement similar projects, but has also developed a framework for analysing knowledge management projects, based on previous empirical research in this area.

In the course of doing so the study has also identified the factors behind CIAP's success; factors that could prove useful for all CIOs interested in emulating it.

Technical and Organisational Infrastructure. CIAP used a simple technical infrastructure incorporating existing hardware and communications networks and the World Wide Web. However, the existing technical infrastructure is far from adequate, due to the lack of PCs and Internet access in the clinical workplace. Moody and Shanks in their report say this was identified as one of the main barriers to the continuing success of the project, and more investment in this area is clearly needed.

Flexible Knowledge Structure. CIAP has different levels of structure of knowledge content: from medical reference databases (external knowledge) with sophisticated indexing structures and search engines, to the clinical protocols and policies (internal knowledge) that are simply posted on the site in five categories. As the number of protocols and policies posted grows more sophisticated, classification and searching mechanisms (for example, categorisation by specialty) may be required.

Knowledge-Friendly Culture.Knowledge and expertise is highly valued in the medical profession, and the health system rewards clinicians for learning on and off the job. It is also part of their professional accreditation to keep up with the latest medical research. Ironically, however, because knowledge is so highly valued and related to power, there is often reluctance on the part of those currently in power to make knowledge more widely available.

Clear Purpose and Language. CIAP had a clearly stated purpose: "to provide clinicians with access to online medical information to support clinical practice, education and research at the point of care". All members of the project team had clinical rather than IT backgrounds, so understood the appropriate terminology to communicate with clinicians.

Motivation to Share Knowledge. With no incentives or rewards for sharing knowledge there was initially reluctance on the part of clinicians to post clinical protocols and procedures for peer review. However, clinicians are highly competitive, and professional pride seemed to provide the primary motivation for doing so.

Multiple Channels for Knowledge Transfer. The project provided only two channels for knowledge transfer, both electronic:n The Internet for access to the latest medical research and for posting clinical policies and protocolsn Listservers for the sharing of knowledge and experiences between individual cliniciansThe lack of face-to-face communication did not seem to present a problem, as there is a great deal of this in everyday clinical practice.

Senior Management Support. Although management support is frequently identified as a success factor in IS implementation and in business change projects, this project was more the result of a "grass-roots" movement by clinicians (through the Clinical Systems Reference Group) than a top-down initiative of senior management. On the other hand, the former chief health officer and several Area Health Service CEOs were early supporters.

Buy Rather Than Build. The project provided access to repositories of medical knowledge developed by third parties rather than creating new repositories, as is the case in most knowledge management projects. This allowed a lot to be achieved in a short time frame.

Use of the Internet. Making the system available via the World Wide Web rather than through intranet technology meant clinicians could access the system from rural and remote locations as well as at home. In the future, it opens up the opportunity for access by private practitioners and consumers.

Project Champion. The manager of the CIAP project team has a clinical background and clearly had a high level of personal commitment to and enthusiasm for the project. She was active in getting resources for the project and promoting the system both within and outside the organisation.

Level of Consultation and Feedback. There was a high level of consultation with clinicians before implementing the system, to find out what information they needed. This was done through the pre-implementation survey and the Clinical Systems Reference Group. Now the system is in operation, feedback from clinicians using the system is continuously obtained through an online survey and via CIAP representatives in each Area Health Service.

Project Team Skills. All of the project team members had clinical rather than technical backgrounds. This gives them first-hand knowledge of what information clinicians need, and how they use it in clinical practice. The authors of the study believe that one of the main reasons for the success of the project is that it is a system built for clinicians by clinicians. This suggests that in knowledge management projects, it is more important for team members to have expertise in the domain being supported (the knowledge domain) than in the technological domain.