OBJECTIVES: Considerable variation is observed in studies presenting model-based cost-effectiveness analyses of systemic therapies for metastatic colorectal cancer (mCRC). This review provides a comprehensive and detailed discussion of structural and methodological assumptions in such analyses of systemic mCRC treatments, and how those might impact health economic outcomes.

METHODS: Five databases (EMBASE, MEDLINE, Cochrane Library, Health Technology Assessment, and National Health Service Health Economic Evaluation Database) were searched for model-based health economic evaluations of systemic mCRC treatment. Study selection, appraisal using the CHEERS checklist, and data extraction was performed by two reviewers independently. Data extraction included general study characteristics, economic methods and assumptions, model structures and modelling technique, evidence used, extrapolation methods, validation efforts, and analyses performed. Economic outcomes were indexed to 2019 US$.

RESULTS: The search yielded 1,418 publications of which 54 were included, representing 51 unique studies. Most studies focused on first-line treatment (n = 29, 57%), followed by third-line treatment (n = 13, 25%). Model structures were health-state driven (n = 27, 53%), treatment driven (n = 19, 37%), or a combination (n = 5, 10%). Cohort-level state-transition modelling (STM) was the most common technique (n = 33, 65%), followed by patient-level STM and partitioned survival analysis (both n = 6, 12%). Only 15 studies (29%) reported some sort of model validation. Health economic outcomes for specific strategies differed substantially between studies. For example, survival following first-line treatment with fluorouracil, leucovorin, and oxaliplatin ranged from 1.21 to 7.33 years, with treatment costs ranging from US$8,125 to US$126,606.

CONCLUSIONS: Model-based cost-effectiveness analyses of systemic mCRC treatments have adopted varied modelling methods and structures, resulting in substantially different outcomes. As models generally focus on first-line treatment without consideration of downstream treatments, there is a profound source of structural uncertainty implying that the cost-effectiveness of treatments across the mCRC pathway remains uncertain.