The total medical costs of cancers are about 5% of national health care expenditures and 10% of Medicare outlays. Much of what we know about cancer costs comes from the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) cancer registries linked to Medicare claims (SEER- Medicare). While the SEER-Medicare link represents the experience of 85% of Aged Medicare beneficiaries enrolled in the fee-for-service (FFS) indemnity option, the link omits the experience of the 15% of seniors enrolled in Medicare HMOs. These groups differ: 1) HMO providers face different incentives;2) HMO Medicare beneficiaries generally have better benefits;3) given that FFS vs. HMO choice is voluntary, the 2 populations may differ in their health status and preferences in ways that are difficult to measure. In addition, utilization and costs for Medicare Working Aged beneficiaries may be under-reported in SEER-Medicare. These factors may cause selection and omission biases in cancer cost estimates based on ether group alone. Building on the Cancer Research Network, we will develop a multi-site, multi-payer database to support analyses extending and complementing the SEER-Medicare link by addressing 3 aims:
Aim 1 - Descriptive Analyses: Estimate the incremental medical care cost of all cancers, and selected cancers, broken down by cancer site, phase of treatment, stage at diagnosis, cancer type (fatal vs. non- fatal), patient demographics, co-morbidity, and source of health insurance for the years 2000-2007 in 4 large integrated health care systems.
Aim 2 - Omissions Bias: Estimate the costs of each cancer care component for Aged Medicare HMO beneficiaries that is omitted from SEER-Medicare: a) non-Medicare covered services;b) effects of health care for seniors whose cancers were diagnosed before age 65;c) cancer screening services;d) cancer prevention services;e) changes in benefits, if any, associated with switching from private health insurance to Medicare;and f) employer-covered benefits for Working Aged beneficiaries.
Aim 3 - Selection Bias: Estimate the incremental medical care costs of all cancers, as well as selected cancers, for SEER-Medicare over 2000-2007. Model the determinants of cancer costs across HMO and FFS systems for Medicare Aged beneficiaries, correcting for selection and omissions biases. We hypothesize that HMO/FFS system differences, benefit differences, and omissions biases will account for a larger proportion of cancer costs that those from selection biases. This study will update the information on the medical care costs of cancer for Aged Medicare beneficiaries. We will also learn how cancer care varies between Medicare's FFS and HMO options.
Showing the most recent 10 out of 16 publications
