NICE is now taking comments on the proposed process and methods for its Diagnostics Assessment Programme (DAP). From HEDS, Alan Brennan, Myfanwy Lloyd Jones and Matt Stevenson were members of the Diagnostics Methods Working Group that drew up the guidance.
There are notable differences with the Technology Appraisal methods in literature searching (where filters should be forgone due to the poor indexing of diagnostic studies), in quality assessment (where the QUADAS checklist is recommended) and in meta-analysis (with no mention of network meta-analyses, presumably due to the dominance of measurement against a reference standard). In cost-effectiveness terms, the reference case and methods are very similar with the only two differences being a greater recognition of marginal costs and implementation costs.
Also of note is that much greater detail is given on the use of expert elicitation, with the methods guide recommending the use of formal methods that also assess probability distributions after training responders to recognise possible cognitive biases. Although it does not mention the Sheffield Elicitation Framework, we would consider this to be the most appropriate and have received excellent feedback on courses and workshops that have run using this.
