OBJECTIVES: The purpose of budget impact models (BIMs) is to forecast the anticipated incremental cost of a new treatment by combining the number of patients potentially eligible for the new treatment and the cost of that new treatment. Historically, the number of potentially eligible patients for new oncology therapies have relied on newly diagnosed cases alone (incidence). As improved screening, early detection, and new treatments for cancers have emerged, duration of survival has increased across many cancer types. With this improved duration of survival, consideration of both newly diagnosed patients and existing patients (prevalence) may be a better way to estimate the eligible patient population accurately.

METHODS: All published studies in PubMed relating to oncology BIMs since September 2010 were identified using MeSH terminology. Relevant abstracts from ISPOR conferences since 2015 were also identified using search parameters: oncology; economic analysis; budget impact analysis. All screened studies and abstracts were evaluated for completeness, relevance, and specifically meeting criteria: English language; United States only; oncology (pharmaceuticals); including BIMs.

RESULTS: Of the 958 records screened (886 PubMed, 72 ISPOR abstracts), 50 were identified for full review. Following exclusion criteria, 21/50 studies were removed from evaluation (no BIMs 13/50; no full text 7/50; not United States 1/50), leaving 29/50 studies for qualitative analysis. A diverse number of cancers and cancer-specific pharmaceutical therapies were included in the final analysis (lung cancer 7/50; blood cancers 4/50; prostate cancer 4/50). Methods for calculating population for BIMs included incidence (19/29), prevalence (6/29), mixed (2/29), and not reported (2/29).

CONCLUSIONS: Oncology BIMs are predominantly using incidence data to calculate the eligible population, which could be contributing to an undervaluing of the oncology pharmaceutical market. These findings will inform future research to quantify the potential underestimation of eligible patient population size by relying on incidence alone.