Accurate projection of the costs of cancer care in the United States on a regular basis is critically important for decision makers to understand the economic impact of cancer and to develop and assess cost containment policies and interventions. We propose to systematically develop new methodology for cost estimation and projection that addresses analytical challenges in the phase-of-care costing method, which was developed more than two decades ago and has since been adopted by the National Cancer Institute in its periodic cost reporting. Our methodology includes new methods to (a) model the longitudinal cost trajectories without artificial partitioning of initial, continuing, and terminal care phases, (b) identify more comparable matched control cohort in the estimation of net cost attributable to cancer, (c) to extrapolate the cost trajectories estimated from SEER-Medicare database to the general United States population and validate the extrapolation algorithm for the purpose of projecting national cancer care costs. We will apply the proposed methods to SEER-Medicare data to project the costs of the top five most costly cancers (breast, colorectal, lung, prostate, and lymphoma), plus leukemia, in the United States between 2020 and 2030. User-friendly software will be developed and disseminated for routine use.
Accurate projection of the cost of cancer care in the United States on a regular basis is critically important for decision makers to understand the financial burden of cancer and to develop and assess cost containment policies and interventions. This project aims to develop new methodology and software for routine use in the projection of cancer care costs.