The NICE Diagnostics Assessment Programme: Does Assessment Lead to Approval?

OBJECTIVES: The National Institute for Health and Care Excellence (NICE) Diagnostics Assessment Programme (DAP) assesses whether evidence for diagnostics supports National Health Service (NHS) adoption. However, evidence requirements are unspecified and the frequency of positive recommendations is unclear. This study analysed diagnostics guidance (DG) to identify recent trends.

METHODS: Recommendations, committee discussion, and the evidence evaluated were analysed from DGs published between June 2019–June 2024.

RESULTS: 23 DGs were identified, of which 18 assessed multiple technologies and 4 assessed single technologies; 1 DG was terminated early. 1 received full support (FS, 4.3%), 11 received partial support (PS, 47.8%), and 10 were not supported (NS, 43.5%). Of the DGs receiving PS, 6 had positive recommendations restricted to a subset of the technologies assessed, 2 had conditional recommendations requiring further research, 2 were restricted to specific clinical settings and 1 had a restricted patient population. De novo cost-effectiveness models (CEMs) were developed for 20 DGs, 1 DG conducted a cost-comparison and 1 conceptualised a CEM due to lack of data. Despite CEM development, the cost-effectiveness was uncertain in 10 DGs, predominantly due to lack of clinical evidence, and 8/10 of these DGs were NS. In line with this, NICE suggested further research recommendations in 20/22 DGs; of the 2 that did not have research recommendations, 1 received FS and the other received PS for restricted use to specific clinical settings.

CONCLUSIONS: Our analyses highlight a lack of positive recommendations for the majority of technologies assessed by the DAP. A primary reason is a lack of clinical data, leading to substantial uncertainty regarding cost-effectiveness. Manufacturers must collect sufficient data if they are intending to pursue a recommendation via NICE, whilst greater transparency from NICE on evidence requirements may ensure their limited resources are not spent assessing technologies with little chance of support.