Alpha sort
Affiliated faculty
Funder
Project text
PI: Elizabeth Lamont, MD, MS
Medicare spent nearly a billion dollars in 2002 on chemotherapy for newly diagnosed elderly breast, colorectal, and lung cancer patients, but surprisingly little is known about the extent to which cancer chemotherapies actually benefit elderly patients. This unsettling situation is the result of the well-described under-enrollment of elderly patients on clinical trials of chemotherapy. With the dearth of information on treatment outcomes of elderly cancer patients, many researchers have turned to studying observational data sources like Medicare data to fill these information gaps regarding chemotherapy outcomes in the elderly.
Despite its many contributions to date, Medicare data unavoidably lacks certain critically important pieces of clinical information, like the survival endpoint of "disease-free survival" (DFS). DFS represents the time from treatment until the first of two events: tumor recurrence or death from any cause. DFS is an important survival end-point in clinical trials of patients treated with adjuvant chemotherapy (most commonly breast or colorectal cancer (CRC) patients) because (1) they have survival horizons that are comparatively long and (2) DFS is highly correlated with overall survival (OS) which, as noted, requires very long follow-up to measure. For these reasons, DFS is an established survival end-point used by the FDA to make decisions on drug marketing. However, unlike OS, DFS is not an existing element within Medicare-based datasets and therefore is not readily available to researchers who use Medicare data to answer a variety of questions - including the generalizability of the adjuvant therapies developed in clinical trials, and the comparative effectiveness (CE) of distinct adjuvant therapies in patients treated in the community.
Our broad aim is to establish the ability of Medicare data to measure DFS in elderly breast cancer and CRC patients who have been treated with adjuvant chemotherapy using a variety of data sources. Our prior work in a small cohort (N=45) of elderly breast cancer patients treated on a clinical trial suggests that it may be possible to reconstruct DFS using Medicare claims. Here we seek to extend this prior work by (1) studying a much larger sample of elderly breast cancer patients treated on clinical trials; (2) broadening our sample to include elderly patients with CRC treated on clinical trials; and (3) applying the CMS-based algorithm to a sample of elderly breast cancer and CRC patients treated with chemotherapy outside of clinical trials (i.e., in the usual care setting).