A Systematic Review of Modelling Approaches in Economic Evaluations of Treatments for Inherited Bleeding Disorders.

Abstract

Objective: The aim of this review is to identify and assess modelling approaches in published model-based economic evaluations of treatments for individuals with inherited bleeding disorders.

Methods: A literature search was performed on seven electronic databases, from database inception until 30 May, 2024. Inclusion criteria were cost-effectiveness or cost-utility analyses using decision-analytic models. The approaches from included models were identified and assessed, and these approaches were compared across bleeding disorders and treatments.

Results: This review included a total of 47 decision-analytic models. The identified models primarily evaluated treatments for severe haemophilia A and B. For haemophilia without inhibitors, factor concentrates were the most evaluated intervention (n = 21, 68%), followed by gene therapies (n = 6, 19%) and emicizumab (n = 4, 13%). For haemophilia with inhibitors, assessed interventions included emicizumab (n = 8, 50%), immune tolerance induction with factor concentrates (n = 5, 31%) and bypassing agents (n = 3, 19%). Markov models were often used as a model type (n = 27, 57%), followed by decision trees (n = 9, 19%), Markov decision trees and decision process (n = 5, 11%) and individual-level models (n = 5, 11%). Regardless of the model type, most authors used a lifetime horizon, a 1-year cycle length, and bleeding events-particularly joint bleeds-as key health states of the models.

Conclusions: As the reviewed decision-analytic models mainly assessed treatments for severe haemophilia, the identified common approaches may only be generalisable to evaluating these treatments. Further research is required to evaluate their relevance for evaluating treatments of milder forms of haemophilia or other inherited bleeding disorders.

Systematic review protocol registration: PROSPERO registration number CRD42023416560.