Presentation (CTI)

OBJECTIVES: To summarize the economic modelling approaches used in recent NICE Health Technology Assessments (HTAs), examine how and why these approaches were chosen, and assess alignment with NICE's preferred methods.
METHODS: We reviewed all technology appraisal guidance (TA), highly specialized technologies guidance (HST) and health technology guidance (HTG) published by NICE between January and June 2026, extracting key data from all appraisals that used economic modelling.
RESULTS: Overall, 44 (of 69) appraisals included economic modelling (32 TAs and 2 HSTs for medications plus 10 HTGs for digital/devices). These spanned diverse disease areas including oncology (38.6%), respiratory (9.1%), neurology (6.8%) and cardiovascular (6.8%). The majority (n=27) of appraisals for medications used cohort-level models such as Markov state transition models (STM) or partitioned survival analyses. Fewer used decision tree models (n=4), or individual-level simulation models (n=3). For device/digital technologies, most appraisals used decision tree models (n=9), one used a cohort-level STM. Model structures were chosen based on precedent from previous NICE appraisals or published models, clinical expert validation, or alignment with trial design and data availability. No evaluations reported the use of formal model conceptualization processes or established conceptualization frameworks. NICE committees accepted the modelling approach used in most appraisals but noted limitations or concerns in 28 (64%). In one case, the committee overrode the model structure. In two further cases, the model was initially rejected and only accepted after it was rebuilt to resolve structural issues.
CONCLUSIONS: The modelling approaches used in recent NICE HTAs are typically driven by precedent and data availability rather than formal conceptualization processes. Adopting structured conceptualization frameworks, such as the ISPOR-SMDM guidelines, and transparently reporting how and why model structures were developed could improve methodological rigor in HTA submissions.