SEER-Medicare: Linkages prior to 2020

The SEER-Medicare linkage was first completed in 1991 and periodically updated thereafter, typically every other year. For each of the linkages, 95 percent of persons age 65 and older in the SEER files were matched to the Medicare enrollment file.

Prior to the 2020 SEER-Medicare linkage, the process for matching persons appearing in the SEER data with their Medicare records is described in: Warren JL, Klabunde CN, Schrag D, Bach PB, Riley GF. Overview of the SEER-Medicare Data: Content, Research Applications, and Generalizability to the United States Elderly Population. Med Care 2002 Aug;40(8 Suppl):3-18.

The SEER data released as part of SEER-Medicare prior to 2020 were in a customized file known as the Patient Entitlement and Diagnosis Summary File (PEDSF). PEDSF contained one record per person for individuals in the SEER database who were matched with Medicare enrollment records. Basic SEER diagnostic information was available for up to 10 diagnosed cancer cases for each person. The PEDSF included information on each person's month and year of birth, date of death (if any), sex, race, and state of residence. It also contained variables from other sources that were added to assist investigators in their analyses. These included Medicare eligibility, reason for Medicare entitlement, and health maintenance organization (HMO) enrollment by month for the period 1991 – 2017.

A Summarized Denominator (SUMDENOM) file was generated for the "non-cancer" cases. This file was used to identify controls for cases or to develop population-based estimates of health care utilization, e.g. use of cancer tests such as PSA and mammography in the entire population. The persons in the SUMDENOM file were identified from a random 5% sample of Medicare beneficiaries residing in the SEER areas. Some of the people in the 5% sample had cancer and appeared in the PEDSF (SEER) data. To prevent double counting, these people with cancer were removed from the SUMDENOM file.

This file allowed investigators easier access to and use of the CMS' yearly denominator data. The file includes one summary record per person with his/her date of birth, date of death (if any), sex, race, state of residence, enrollment in Part A and/or Part B, and enrollment in an HMO by month and year.

Investigators that have an approved application and purchased data prior to the release of the 2020 SEER-Medicare linkage will be allowed to keep these files until the DUA expires. The data retention as stated in the DUA is five years. At the end of five years, all SEER-Medicare data must be destroyed. If the investigator would like to continue work on the project, they must submit a new application and be approved for the files from the 2022 SEER-Medicare linkage.

Below are the Code Documents and SAS input Statements for the data released in 2018. If there are questions about the previously released data, users are encouraged to review the SEER-Medicare Training materials.