Presentation (CTI)

OBJECTIVES: Patient input is considered in health technology assessments (HTAs) by the UK’s National Institute for Health and Care Excellence (NICE), Canada’s Drug Agency (CDA), and Australia’s Medical Services and Pharmaceutical Benefits Advisory Committees (MSAC and PBAC respectively). This research explored the frequency of mentions and impact of patient input, including both patient-reported outcomes (PROs) and supplementary patient inputs (SPIs), in HTAs of cell and gene therapies (CGTs).
METHODS: HTA reports of CGTs assessed by NICE, CDA and PBAC or MSAC since 2020 were identified. Natural language processing techniques, including the use of an OpenAI o4-mini language model, were used to gather relevant quotes and generate ratings of direct, indirect or no impact of PROs and SPIs on the HTA outcome, which were subsequently validated via human input. Counts of predefined relevant keywords were also extracted and analysed using a negative binomial generalized linear model, to explore how wordcount varied by market.
RESULTS: Across the 3 markets, 4 CGTs were identified which have undergone an assessment for at least 1 indication. A total of 21 HTA reports were analysed (7 per market). The variation in the total wordcounts across markets was not statistically significant (NS) (Australia: 99; Canada: 92; UK: 66). Where PRO outcomes were available, robust, and significant, they directly influenced the HTA outcome through their incorporation in economic analyses across markets. Where available, SPIs were acknowledged and had an indirect impact across all HTA reports, however none had a direct impact.
CONCLUSIONS: The variation in frequency of mentions of patient input in HTA reports across markets was NS. PROs were consistently found to have a direct impact on HTA outcomes. While SPIs only had an indirect impact, it is promising to observe their consistent consideration in CGT assessments across markets over the last five years.