This page highlights any significiant changes to the data or request process for each annual release. Every year, there are also a number of minor updates to the data and fields released. Details of those types of changes can be found in the SEER*Stat Databases Documentation.
On this page:
April 2022 Release
Starting with the 1975-2019 SEER Data, released on April 15, 2022, the following changes were made to the data.
The registry groupings have changed with the November 2021 databases due to changes in which registries are included with this data submission. The new groupings are now SEER 8, SEER 12, SEER 17, and SEER 22. Detroit is no longer included in any databases, and Illinois and Texas are included in the new 22 registry database. Refer to Registry Groupings in SEER Data and Statistics for details.
Statistics by Race and Origin
SEER's recommendations for producing statistics by race and origin changed with the November 2021 databases. Geographic considerations to select correct population/denominators are also required when producing rates by certain race/ethnicities, e.g., American Indian/Alaska Native. Refer to Race and Hispanic Ethnicity Changes for more information.
County attributes that are available in different SEER*Stat session types have changed with the November 2021 databases and SEER*Stat version 8.4.0.
- Incidence-based mortality databases — time-dependent county attributes are linked by year of death rather than years of diagnosis.
- Prevalence sessions — county attributes are not available in prevalence calculations.
- MP-SIR — limited to static attributes.
- Case listing — limited to minimal set of time-dependent attributes that are permitted to be included in case listing results.
April 2021 Release
Starting with the 1975-2018 SEER Data, released on April 15, 2021, the following changes were made to the data.
SEER now uses "Behavior code ICD-O-3" rather than the "Behavior recode for analysis" field in SEER*Stat and any published statistics. This has the following impact:
- The default behavior selection in SEER*Stat for all November 2020 SEER databases is “Behavior code ICD-O-3 = Malignant", for prior submissions it is “Behavior recode for analysis = Malignant".
- All cases that were coded as "Behavior recode for analysis = No longer reportable in ICD-O-3" have been excluded from all databases. This includes approximately 10,000 cases diagnosed between 1973-2000, all of which had ICD-0-3 behavior of borderline and almost all of which were ovary cancers.
- The "Behavior recode for analysis" field is still available in the database.
Read more about the Behavior Variable Changes.
Site Recode Variables
Several new or revised site recode variables have been added and are used for SEER's published statistics. They include:
April 2020 Release
SEER has a long history of enabling and supporting data access to researchers, clinicians and patients. We are committed to continuing that mission. However, as SEER expands the data captured in terms of detail and longitudinal information, the risk of re-identifiability of individuals is increased. Further, in recent years software algorithms have been developed that could leverage other data sources combined with the current limited dataset that has been the primary research file available from the SEER program to re-identify individuals (see References). For these reasons, it is imperative that SEER take measures to protect the confidentiality of the patients within the limited dataset held within the SEER program, to further reduce any risk of re-identifiability. This requires a change in how users can access the data.
We will continue to support free and open access to our data, but additional requirements for authentication and authorization will be required depending on the data product access being requested. The SEER program is working hard to continue to support the needs of the research community and others who utilize the SEER data while simultaneously increasing assurance of patient confidentiality.
Two SEER Data Products: Research and Research Plus
Starting with the 1975-2017 SEER Data, released on April 15, 2020, there are two data products available: SEER Research and SEER Research Plus. The Research Plus databases require a more rigorous process for access that include user authentication. For more information, see the Comparison of Data Products.
Other Data Changes in this Release
- SEER no longer excludes Louisiana cancer cases in all Research and Research Plus databases. This is a change from previous releases. See Adjustments for Areas Impacted by Hurricanes Katrina and Rita for more information.
- Data files are no longer available for download in ASCII text or binary file formats. Data can only be accessed through SEER*Stat in client-server mode. If you need to work with individual cancer cases and export SEER*Stat results for use in SAS, R, or other software refer to the SEER*Stat Case Listing Tutorials.
- "Insurance Recode (2007+)" is being re-evaluated and has been removed from the November 2019 databases. This field and new SEER-Medicaid fields will be available in specialized data in the future.