Session (CTI)

ISSUE: Indirect treatment comparisons (ITCs) are essential for health technology assessment (HTA) when head-to-head evidence is unavailable. Advances in methods, including population-adjusted indirect comparison, multilevel network meta-regression, and real-world evidence, have increased the sophistication of comparative effectiveness analyses. However, a persistent challenge remains: which covariates should be included in ITC adjustment models? Covariate misspecification can affect the feasibility, precision and validity of ITCs. HTA guidance, including the Joint Clinical Assessment (JCA) Guidelines and NICE Technical Support Documents, emphasise adjustment for treatment effect modifiers and prognostic factors, but give limited direction on identifying, justifying and validating them. In practice, covariate selection varies across submissions, raising concerns about transparency, reproducibility and bias. Regulatory agencies such as the European Medicines Agency (EMA) routinely assess clinically relevant covariates in trials and for marketing authorisation. Whether and how regulatory evidence should inform HTA covariate identification remains an important question, especially as stakeholders seek greater alignment between regulatory and HTA evidence requirements. OVERVIEW: representatives from HTA, regulatory, academic and industry will debate key questions on covariate identification in ITCs. The moderator will introduce current challenges in covariate selection across ITC submissions. Panellist 1 will discuss HTA expectations for robust covariate adjustment. Panellist 2 will examine how regulators consider clinically meaningful treatment effect modifiers and prognostic factors during evidence evaluation. Panellist 3 will present a structured framework using publicly available regulatory documentation to support transparent covariate selection in ITCs and evaluate it through a previous NICE case study. 10-15 minutes for each panellist followed by 20 minutes for panel discussion and Q&A to determine next steps towards more transparent and decision-relevant covariate identification practices in ITC.