Personalized Treatment Hierarchies in Bayesian Network Meta-Analysis
Augustine Wigle, Erica E. M. Moodie
TL;DR
The paper addresses how to construct personalized treatment hierarchies within Bayesian Network Meta-Analysis (NMA) when treatment-covariate interactions (TCIs) are present. It shows that, with TCIs, the relative effect for a treatment g versus a reference depends on covariates: $\psi_{g0}+\sum_q\psi_{gq}x_q$, so hierarchies must be tailored to specific covariate profiles and summarized as $\text{E}[\text{rank}(g)\mid x_1,...,x_Q]$. The authors outline methods to compute covariate-specific SUCRA-based rankings from IPD-NMA models and illustrate them with an MDD network, demonstrating that different patient profiles yield different treatment hierarchies even when TCIs are not statistically significant. This work highlights the practical value of personalized hierarchies for clinical decision-making and suggests integrating such tools into IPD-NMA outputs and online decision aids.
Abstract
Network Meta-Analysis (NMA) is an increasingly popular evidence synthesis tool that can provide a ranking of competing treatments, also known as a treatment hierarchy. Treatment-Covariate Interactions (TCIs) can be included in NMA models to allow relative treatment effects to vary with covariate values. We show that in an NMA model that includes TCIs, treatment hierarchies should be created with a particular covariate profile in mind. We outline the typical approach for creating a treatment hierarchy in standard Bayesian NMA and show how a treatment hierarchy for a particular covariate profile can be created from an NMA model that estimates TCIs. We demonstrate our methods using a real network of studies for treatments of major depressive disorder.