Doctors are used to ordering tests for patients – such as bloods or MRI scans – but they will now test out a platform where they can order the analysis of large-scale data relevant to an individual patient and treatment decision they face, with results returned within clinical timescales.
The ‘Informatics Consult’ platform – the first UK prototype of its kind – will have access to UCLH data to begin with. Professor Bryan Williams, Director of Research at UCLH, and Professor Harry Hemingway and Dr Alvina Lai of the UCL Institute for Health Informatics will lead the work, which is a beneficiary of a Better Care Catalyst Award from Health Data Research UK and The Health Foundation.
University Hospitals Birmingham, Barts Health Trust and Great Ormond Street Hospital will also trial the system using their own data.
Patients might reasonably expect that their medical outcomes are used to help doctors make treatment decisions for other patients. However, this is not always the case. A common dilemma facing clinicians is deciding on the most appropriate treatment decision in the absence of evidence from randomised clinical trials (RCTs) or clinical guideline recommendations.
For example, a patient who has both cirrhosis of the liver and atrial fibrillation (AF) poses a treatment dilemma for the doctor because AF is usually treated with anticoagulants to lower stroke risk, but cirrhosis is known to increase the risk of bleeding. There are no trials or guidelines to help the doctor, and currently, there is no mechanism in the NHS to learn from the treatment and outcomes of previous patients with both conditions.
The Informatics Consult will allow clinicians to select relevant conditions from a computer interface, and order an analysis of raw data relating to those conditions that is performed within minutes or hours. The platform will return easily interpretable information to the clinician on outcomes involving patients with the conditions of interest.
Professor Williams said: “Data we gather from the platform could transform health and care. There is so much raw data collected by hospitals, and we want to put this data to use, to shed light on the best treatment approaches where there is uncertainty. This will improve treatment for individual patients, and it will also reveal where we don’t have good evidence for treatment approaches, and indicate what trials we need to do. It will also help with recruitment into clinical trials.”
Dr Lai said: "We have taken a 'patients like me' approach to understand how we can improve health outcomes by optimising treatment decisions using population health data. This is particularly useful in situations where randomised trials are not feasible, for example in patients with multiple chronic conditions. Starting from one exemplar (cirrhosis and atrial fibrillation), we hope to use the Better Care Catalyst Award to scale up to look at a wide range of disease combinations where treatment uncertainty persists.”