File(s) not publicly available

Reason: The SEER-MEDICARE data used in this study, are not publicly available but can be obtained by researchers, by following the process described on https://healthcaredelivery.cancer.gov/seermedicare/obtain/requests.html

Metadata supporting the published article: A Medicare-based Comparative Mortality Analysis of Active Surveillance in Older Women with DCIS

online resource
posted on 09.10.2020, 08:26 by Igor Akushevich, Arseniy Yashkin, Rachel Greenup, Shelley Hwang
Over 97% of individuals diagnosed with preinvasive cancer, or ductal carcinoma in situ (DCIS) will choose to receive guideline concordant care (GCC), which was originally designed to treat invasive cancers and is associated with treatment related morbidity. An alternative to GCC is active surveillance (AS) where therapy is delayed until medically necessary. Differences in mortality risk between the two approaches in women age 65+ were analyzed in this study.

Data access: All the datasets analysed in the current study are in .SAS file format. SEER-MEDICARE data were used to generate all the tables and figures in this article. The Centers for Medicare and Medicaid Services do not allow the redistribution of their data by researchers. SEER-MEDICARE data are distinct from the publicly available SEER database, and can be obtained by researchers, by following the process described on https://healthcaredelivery.cancer.gov/seermedicare/obtain/requests.html (access requirements include Institutional Review Board approval, and the completion of a Data Use Agreement). Please note that the process and range of data availability has changed in 2020 and therefore obtaining some of earlier years of data used in this study may prove a challenge.

Study approval: The Duke University Institutional Review Board approved the protocol used in this study.

Study aims and methodology: Active surveillance (AS) is an alternative to GCC under which no definitive therapy is undertaken at diagnosis. The use of AS is especially pertinent to patient groups characterized by high levels of co-morbidity, such as women age 65+, for whom aggressive treatment may be a sub-optimal choice. The purpose of this study was to evaluate the differences in breast-cancer-specific and all-cause mortality between three groups of women age 65+ diagnosed with DCIS: i) those who initiated GCC within a year of diagnosis and before any evidence of cancer progression was identified; ii) those who delayed providing consent to GCC treatment until presented with evidence of cancer progression within a year of the initial diagnosis of DCIS; iii) those who refused GCC treatment for one year or longer.
Data drawn from the Surveillance, Epidemiology and End Results program linked to administrative health insurance claims records from the Medicare program (SEER-Medicare) were used for this study. To ensure Medicare coverage, the authors restricted the sample pool to females who were age 65+ at the time of their breast cancer diagnosis over the 1992-2011 period (N=381,056). The final sample contained 22,576 female patients diagnosed with DCIS.
For more details on the methodology and statistical analysis, please read the related article.

Description of data and analysis: Upon accessing the SEER-MEDICARE data, the authors converted all raw data files into SAS format using SAS version 9.4. The denominator files were used to identify the diagnosis date, first course of treatment chosen by the patient, date of death (if any), gender, race/ethnicity and age of each individual as well as the length of lookback and follow-up available. The authors then queried the carrier, medpar, and outpatient claims files (for each year of the data) to identify instances of cancer treatment (using Current Procedural Terminology Version 4, Procedural International Classification of Disease version 9 and/or Healthcare Common Procedural Coding System codes as appropriate) and to verify cancer diagnosis and ascertain the presence of selected comorbidities (using International Classification of Disease version 9 Codes). The algorithms used are described in detail in the citations provided in the Methodology section of this manuscript. All analysis was done using SAS 9.4 and following the methods described in the Methodology section of the manuscript.

Software needed to access the SAS files: The SAS statistical software is required to access the .SAS files.

Funding

This research was supported by Patient-Centered Outcomes Research Institute Grants CER-1503-29572

History

Research Data Support

This record was produced by Springer Nature’s Research Data Support service. This service focuses on maximising the findability and accessibility of the data, and does not involve peer review of data.