Descriptive and Quality Analysis of Markov type Decision Analytic Modeling in Economic Evaluations of Metastatic Breast Cancer Treatment Therapies

Background: Breast cancer is the most common type of cancer in women. The assessment of characteristics and quality of decision analytic models of economic evaluations (EE) in metastatic breast cancer treatments has primary importance because EEs provide important evidence for decision-making concerning health care. The decision analytic modeling (DAM) process is central to performing economic evaluations in chronic diseases like breast cancer, but may produce low quality evidence depending on the methods applied. Objective: The objective was to summarize the cost-effectiveness results, methodology and quality of DAMs in MBC with the aim to provide guidance on the cost-effectiveness of current therapy option and to appraise and illuminate the flaws in the quality of DAM so as to provide future recommendations for generating robust evidence on cost-effectiveness of MBC therapy. Methods: A literature review was conducted to identify the EEs of MBC treatment written in English language between 2002 and 2012. Methodological characteristics were observed by extracting data regarding methodology of DAM. A quality appraisal checklist designed specifically for DAMs was applied to appraise the quality of all EEs in MBC treatments utilizing a Markov models. Results: In total thirteen EEs were identified which represented DAM-based cost effectiveness and cost utility analysis, and outcome measures were quality adjusted life years or life years. Studies were conducted in various countries with the health care perspective being prevalent. In most studies lifetime horizon was used, though this differed according to target patient group and was found to be influential on the results for similar drugs. Discounting was applied according to guidelines and decision maker requirements. Comparators in these studies were included hormonal therapies, targeted therapy and chemotherapy. In general, combination therapy including 2 or more types of therapy was more cost-effective than combination therapy. The overall quality score of DAM was 70%. The best performance of all reviewed studies was in structure 80.93% and the data dimension 67%, while the consistency dimension scored the lowest at 35.38%. Model type (S6), cycle length (S9), and parameter uncertainty (D4d) were scored high in all EEs. The sub-dimensions which scored lowest were comparators and strategies (S5), rationale of structure (S3), data identification, (D1) Pre-model data analysis (D2), baseline data (D2a), and treatment effects (D2b) and quality of life weights (D2d). Conclusions: Development of a generic model for MBC is recommended to improve quality of DAM and reduce the variations in final results. A systematic review of the efficacy evidence of all therapies in MBC is needed to identify the heterogeneity which can be integrated into the model by means of a meta-analysis. Such a model could then be used to economically evaluate MBC treatment therapies and adapted depending on patient subgroup.