Every week, hundreds of papers by expert medical researchers are published in high-quality journals and this research fuels continuous improvements in treatment and evidence-based medical practice. The knowledge produced by the research is typically disseminated through the preparation of Clinical Practice Guidelines (CPGs) which give summary recommendations of best practice for particular conditions. NICE, for example, has a world-leading programme in authoring and publishing CPGs for hundreds of common clinical conditions – from cancer and heart disease to diabetes and mental health. With over 20,000 different medical conditions identified, CPGs are generally considered to be critical to disseminating up-to-date recommendations for high quality and safe care.
However, there are important difficulties for achieving all the intended benefits; CPGs take time to read and absorb, and they only provide general guidance not patient-specific recommendations. It is also difficult to manage and access the huge volume of content that is available in CPGs. A specialist team in the US may have written a CPG that would be of great value to an Oxford doctor in a particular situation – but it may be hard to find and not entirely appropriate in the NHS context. Consequently many organisations around the world duplicate the work on CPGs that NICE and similar organisations do.
OpenClinical is a publishing technology that is directly addressing these challenges. It is a web-based ‘knowledge-sharing platform’ designed by Professor John Fox in the Department of Engineering Science that offers a radically new approach to the process of creating and using CPGs (www.openclinical.net). Starting from the traditional text form of a CPG, an OpenClinical ‘publet’ can be created that captures the knowledge described in the CPG but in a form that is computer executable. When a publet is used during a patient’s care it can record and interpret information about the patient, apply the knowledge and reasoning described in the CPG, and give personalised recommendations for diagnosis and treatment.
Publets can also guide a clinician through just the parts of the CPG that are relevant, summarising the pros and cons of each decision option for the patient when required, and providing evidence and relevant background knowledge for each recommendation. Clinicians are protected from material that is irrelevant to the individual case, and they can also act on or reject the suggestions as they see fit in light of the patient-specific rationale that is available.
The key innovations in OpenClinical are a set of artificial intelligence techniques for modelling medical decision-making and care plans, and authoring software that empowers healthcare professionals to create and maintain pathways that are consistent with their professional experience and expertise.
Authors from around the world prepare and submit publets to OpenClinical in much the same way that researchers submit a paper to a journal; each one peer-reviewed by independent experts before it is published. This enables qualified professionals in any country to publish high-quality pathways in any area of medicine, and their colleagues in other countries to download and adapt them to meet local circumstances.
OpenClinical’s ‘computer interpretable models of care’ do not seek to replace the judgement of a specialist, but they provide robust, up-to-date support to improve the consistency, quality, safety and speed of decision-making. The Oxford group thinks the concept has the potential to completely change everyday patient care across the globe. Over time it is envisaged that the OpenClinical repository will become a source of high quality knowledge of best practice and powerful tools for personalised decision-making in every area of medicine.