Reviewing the quantity and quality of evidence available to inform NICE diagnostic guidance. Initial results focusing on end-to-end studies and the implications for Cochrane Reviews




Poster session 5


Thursday 27 October 2016 - 10:30 to 11:00


All authors in correct order:

Hyde C1, Byron S2, Nixon F2, Albrow R2, Walker T2, Deakin C2
1 Exeter Test Group, University of Exeter, UK
Presenting author and contact person

Presenting author:

Christopher HYDE

Contact person:

Abstract text
Background: The National Institute for Health and Care Excellence (NICE) has produced guidance on medical diagnostic technologies since 2011. This has resulted in 22 pieces of guidance on wide-ranging topics. As part of the process of reviewing its methods, the pieces of guidance and the underpinning evidence are being examined to inform thinking on potential future developments. The expectation in diagnostics assessments is that end-to-end studies, such as comparative outcome studies - like RCTs - are rarely available. This study reports on the availability of end-to-end studies. Anecdotally, the experience of the NICE team is that several pieces of guidance have been informed by end-to-end studies contrary to our expectation that this would be very unusual. Therefore, we wanted to examine the frequency and nature of this phenomenon in detail, and considers how these studies informed the considerations and decision making of the NICE Diagnostics Advisory Committee. We also wanted to see whether Cochrane Reviews of these studies were available.
- To identify how many pieces of NICE diagnostics guidance were informed by end-to-end studies.
- To describe the nature of the end-to-end studies identified.
- To describe how the end-to-end studies informed committee discussions and the final guidance.
- To assess whether Cochrane Reviews could have been used.
Methods: The approach will be a document analysis of all pieces of published diagnostics guidance and the underpinning evidence. A data extraction form will be developed and piloted. Extraction will be performed by one researcher and checked by a second. Data will be tabulated and conclusions derived from the tables produced. Where results are quantified, such as the frequency of reports with end-to-end studies, 95% CI will be calculated.
Results: The analysis will be available at the Colloquium. There are at least two pieces of guidance where RCTs compare the impact of introducing a strategy involving a new test with the impact of an existing strategy.
Conclusions: Once completed, this work will inform discussions on potential future developments for the assessment of diagnostic technologies.