OBJECTIVES: Cost-effectiveness analyses consist of a baseline model (BM) representing absolute natural history under a “standard treatment” in the comparator set, and a model for relative treatment effects (RTE). Estimates from the two models are then combined to obtain the natural history under the new treatment. NICE Decision Support Unit Technical Specification Document 5 recommends using BMs alongside RTE models for evidence synthesis. The guidance also highlights the importance of using population-specific evidence in BMs and the need for sufficiently detailed reporting justifying the evidence used. This study explores reported implementation of BMs in network meta-analyses (NMAs) supporting NICE technology appraisals (TAs), focusing on their use, and if used, the evidence used to populate the BMs and NICE opinion on applicability to the UK population.

METHODS: TAs posted on the NICE website between May 2020 and April 2023 were considered. Multiple technology appraisals (MTAs); TAs with changes to marketing authorisation, recommendation updates, or revised patient access schemes, rather than resubmissions; and TAs without a supporting indirect treatment comparison (ITC) were all excluded.

RESULTS: Of the 257 TAs, 202 (78.6%) non-terminated TAs were reviewed. After excluding 5 (2.5%) MTAs, 8 (4.0%) minor updates, and 66 (32.7%) TAs without an ITC, 123 TAs remained. Of these, NMA was performed in 64 TAs. In most TAs (92.2%), the implementation of BMs was not specified. Only 5 TAs reported the use of BMs, 2 of which appeared to include all reference arms from all trials included in RTE models, and 3 where the evidence source was unclear. Justification for choice of evidence was not reported. Only 1 TA was criticised for not selecting evidence reflective of UK clinical practice.

CONCLUSIONS: Despite recommendations for transparency, reporting of the evidence informing BMs remains unclear across NICE TAs, thereby hindering assessment of the suitability of the BMs for the decision problem.