OBJECTIVES: The Joint Clinical Assessment (JCA) is a crucial procedure designed to standardize and accelerate the evaluation of new technologies across various regions and markets. Recently methodological and practical guidelines for quantitative evidence synthesis were published to respectively support JCA by describing methods for direct and indirect treatment comparisons (ITCs) and to provide more granular advice to assessors. This work aims to provide guidance on how to conduct ITCs considering both JCA and National Institute for Health and Clinical Excellence (NICE).

METHODS: We compare methodological recommendations from the practical guideline for quantitative evidence synthesis with those from NICE, as summarized across the Technical Support Document (TSD) series.

RESULTS: The recommendations from both sources significantly overlap, particularly regarding the assessment of key assumptions of a (network) meta-analysis, i.e., similarity, homogeneity and consistency. However, the JCA guideline provides revised thresholds from Cochrane Handbook for assessing heterogeneity based on the I² and presents different recommendations on inclusion of trials with zero cells, advocating frequentist approaches, e.g., the beta-binomial model, while NICE follows a fully Bayesian approach. Other differences lay in the issue of missing data and in recommended methods for survival outcomes with non-proportional hazards with NICE advocating fractional polynomial models or restricted splines whereas the JCA guideline adds the restricted mean survival time approach. Lastly, the guidance discusses the ‘researcher degrees of freedom’ regarding the multiplicity issue arising from multiple population adjustment methods and suggests these methodologies as more suitable for exploratory rather than primary analysis.

CONCLUSIONS: The JCA and NICE guidelines share significant overlap, and an updated version of the NICE TSDs incorporating more recent methodologies will likely increase this alignment. However, some context-specific differences remain, necessitating careful planning for different sets of analyses to meet requirements in an efficient manner.