Safe and Sound
Safe and Sound was a "grand challenge" project in information and communication technologies for healthcare of the UK Digital Economy research programme (a Research Councils UK programme aimed at realising the transformational impact of ICT in all aspects of business, society and government). The project was funded by EPSRC/MRC and ran from 2008-2009. The partners in the project were the University of Oxford Engineering Science department, the University of Edinburgh Informatics department and the Centre for Patient Safety & Service Quality at Imperial College London.
The starting point for Safe and Sound was the belief that there is enough information available to substantially improve patient care but that, for many varied reasons, this information is not being used as effectively as it could be.
Optimum use of medical knowledge in clinical practice is extremely difficult to achieve, not least because it is growing at such a rate. For example, in 2004, the National Library of Medicine indexed more than 575,000 articles from 4,800 journals, and both numbers are increasing every year. Medical knowledge is also becoming more complex and therefore harder and harder for clinicians to keep pace with it. For instance, more than 600 drugs require adjustment of doses for multiple levels of renal dysfunction.
Safe and Sound aimed to provide technology to integrate the vast amount of existing knowledge relevant to providing high quality healthcare and deliver it in an effective and coordinated manner at the point of care.
Information technology can provide effective support for clinicians; the adjustments described above can be readily managed by a computer program. More generally, Clinical Decision Support (CDS) systems can be usefully deployed in most areas of medicine, addressing many of the most important causes of avoidable medical harm.
Systematic reviews of the impact of prototype CDS systems have confirmed the intuition that they can dramatically improve the quality of care that patients receive, and dramatically reduce the cost of providing that care. Given what we now know about the prevalence and consequences of medical error, this technology should be a high priority for deployment throughout the NHS. Yet for all that work, and despite the urgent need that they can address, CDS systems have not been widely adopted in practice. To make that step, we need a dedicated strategy for translating the field's promise into real, practical results.
In the short term at least, clinical decision support may develop on the basis that knowledge is controlled from a central location. Health centres that adopt the technology first will have to curate their own knowledge services, or employ one of a limited pool of third parties to do so. But in the longer term, it should be possible to take a much more flexible approach.
The long term goal is therefore a progressive shift from centrally controlled knowledge services through flexible knowledge distribution, towards a fully distributed market in knowledge services. Broadly, this shift is desirable because the clinical workflows that CDS systems support are themselves distributed – complex webs of interacting staff and systems - which must be choreographed to best effect. By reproducing that kind of organisation directly in CDS systems, we increase the probability that those systems will “fit” with the institutions that use them. One other more concrete benefit of this distributed approach is that it can help health centres to place the burden of knowledge management and maintenance with those best suited to carry it – the specialist professional bodies on whose approval that knowledge already depends.
The results of Safe and Sound consist of a document, a video and a prototype. The document illustrates the shift in approach required to move from the current state of the art in CDS systems. The video gives an intuitive understanding of the impact of the shift in the medical environment. The prototype shows what is already feasible integrating state-of-the-art technologies.