Systematic review of cost-effectiveness modelling studies for haemophilia.

Abstract

Aims: Haemophilia is a rare genetic disease that hinders blood clotting. We aimed to review model-based cost-effectiveness analyses (CEAs) of haemophilia treatments, describe the sources of clinical evidence used by these CEAs, summarize the reported cost-effectiveness of different treatment strategies, and assess the quality and risk of bias.

Methods: We conducted a systematic literature review of model-based CEAs of haemophilia treatments by searching databases, the Tufts Medical Center CEA registry, and grey literature. We summarized and qualitatively synthesized the approaches and results of the included CEAs, without a meta-analysis due the diversity of the studies.

Results: 32 eligible studies were performed in 12 countries and reported 53 pairwise comparisons. Most studies analysed patients with haemophilia A rather than haemophilia B.Comparisons of prophylactic versus on-demand treatment indicated that prophylaxis may not be cost-effective, but there was no clear consensus. Emicizumab was generally cost-effective compared with clotting factor treatments and was always dominant for patients with inhibitors. Immune tolerance induction following a Malmö protocol was found to be cost-effective compared to bypassing agents, while there was no consensus for the other protocols. Gene therapies as well as treatment with extended half-life coagulation factors were always cost-effective over their comparators.Studies were highly heterogenous regarding their time horizons, model structures, the inclusion of bleeding-related mortality and quality-of-life impacts. This heterogeneity limited the comparability of the studies. 19 of the 32 included studies received industry funding, which may have biased their results.

Limitations: It was not possible to perform a quantitative synthesis of the results due to the heterogeneity of the underlying studies.

Conclusion: Differences in results between previous CEAs may have been driven by heterogeneity in modelling approaches, clinical input data, and potential funding biases. A more consistent evidence base and modelling approach would enhance the comparability between CEAs.