gms | German Medical Science

Purpose: To provide an overview of published decision-analytic models evaluating treatment strategies in Multiple Myeloma (MM) focusing on the structural and methodological modeling approaches and to derive recommendations for future Multiple Myeloma models evaluating different treatment regimens and sequences for Myeloma and Myeloma complications.

Methods: We performed a systematic literature search in the electronic databases Pubmed, NHS EED and the Tufts CEA Registry to identify published studies evaluating MM treatment strategies using mathematical decision models. To meet the inclusion criteria, models had to compare different treatment strategies, be published as full text articles in English, and comprise relevant clinical health outcomes (e.g., responses, progression-free survival, life-years gained or QALYs) over a defined time horizon and population. Inclusion of costs was optional. We used evidence tables to summarize methodological characteristics, such as modeling approach, simulation technique, perspective, time horizon, model validation and uncertainty analysis.

Results: We identified eleven [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11] different decision-analytic modeling studies. All studies included an economic evaluation. The modeling approaches varied considerably. The authors applied a decision tree model, Markov cohort model, discrete event simulations, partitioned survival analyses and area under the curve models. Analytic time horizons ranged from seven years to lifetime. Six models adopted the perspective of the health care system, three a third party payer, two the government payer and only one the societal perspective. Health outcomes included (overall-, median-, progression-free) survival, number needed to treat, time to discontinuation of treatment, life expectancy, and QALYs. Compared treatment strategies included lenalidomide, dexamethasone, bortezomib, melphalan, prednisone, thalidomide, haemodialysis, bone marrow transplantation, zoledronic acid, and clodronate. In most cases, model validation was only mentioned in the discussion when comparing the results with other cost-effectiveness studies. All authors performed deterministic sensitivity analyses. Additionally, seven articles reported a probabilistic sensitivity analysis.

Conclusions: We identified several well-designed models for different multiple myeloma treatment strategies evaluating relevant health outcomes as well as economic parameters. However, the quality of reporting varied considerably and in some cases the models were not sufficiently described. For the future, we recommend a clear model description including all relevant parameters and model validation using independent data.

*This work was supported by the COMET Center ONCOTYROL, which is funded by the Austrian Federal Ministries BMVIT/BMWFJ (via FFG) and the Tiroler Zukunftsstiftung/Standortagentur Tirol (SAT).