New ‘NICE real-world evidence framework’ launched

A new ‘NICE real-world evidence framework’ was launched last week, with the aim to improve the quality of real-world evidence which informs NICE guidance development.

Real-world evidence refers to evidence which comes from data gathered outside of a highly controlled clinical trial. This is particularly relevant to primary care, as we know that patients recruited to clinical trials are not the same population as those we manage day-to-day in practice, who are often older and with more complex co-morbidities than those eligible for recruitment to research studies. Examples of real-world data include the Clinical Practice Research Datalink (CPRD), which gathers clinical information from subscribed GP practices using Vision or EMIS software, and the Zoe app collecting information directly from patients about potential Covid symptoms. Such research platforms mean we can gather large amounts of data from an unrestricted population and which can be used to answer research questions using a more ‘real-world’ sample. Their use has increased dramatically in recent years, with many different ‘real-world data’ sources becoming available. However,  concerns have been raised about data quality, data mining and risk of bias in analysis methods. The new NICE framework sets out core principles to address these issues, and presents Reporting Checklists – a Data Suitability Assessment Tool (DataSAT) and Reporting on Methods used to Minimise Risk of Bias which should be used in submission of real-world evidence for publications.