Data Sharing Practices in U.S. Cancer Registry: Results of NAACCR Survey

In May 2026, NAACCR conducted a survey of U.S. registries to better understand the variation in data release practices. Forty-one registries responded to the survey (76% response rate) and data was analyzed by Betsy Kohler, in her role as consultant to NAACCR. The survey included questions about the primary data release to studies approved by the registry or their IRB. We also sought information on registry policies around secondary data sharing (SDS), which occurs when a primary researcher shares data with another researcher to advance science. Sometimes the shared data may contain information obtained from the cancer registry. The NIH Data Management and Sharing Policy requires that NIH-funded studies participate in secondary data sharing to maximize use of the data for scientific advancement.

Key Results Related to Primary Data Sharing:

• Most registries (84%) allow the release of full dates while 16% allow release of month and year only.
• When asked about the smallest geographic area for which data could be released, about 82% of registries allow release of data at the census tract or census block group level (45% and 37% respectively), while about 13% can release data at the county level only.
• Release of registry data varies across the following data items:
  • Summary Stage (100%)
  • AJCC-TNM Stage (82%)
  • Extent of Disease (76%)
  • Genetic Markers (79%)
  • Surgery and Biological Response Modifiers (both 95%)
  • Radiation, chemotherapy, and hormone therapy (all 97%)
• When releasing data, the data element must be shown to be integral to the study. The registry community has more confidence in some of the above data items than others, and discussions with the researcher may be required to understand the quality and utility of these data items. Some registries note that they will release but notify the researcher that the data may be incomplete and/or not validated through routine quality control practices.
• Twelve registries (32%) indicated that a separate application to the state vital statistics office is necessary to obtain mortality data. Among these registries, separate approval is required for vital status (55%), date of death (73%), cause of death (100%), cause of death recode (82%), and survival data items (46%).
• Eight of the responding registries (20%) indicated that they do not release information obtained from the National Death Index (NDI), even though this is an allowable use (see NDI Fact Sheet here: naaccr.org/data-security-confidentiality-issues/).

Key Results Related to Secondary Data Sharing (SDS):

• 52% of registries have a written policy regarding SDS and 59% reported participating in SDS despite the presence or absence of a written policy.
• SDS policies are determined by a variety of individuals and bodies, including the registry’s privacy/legal office or governance board (42%), the registry director or their department (21%), as defined in regulation/legislation (16%) or through adoption of the NIH DMS policy (11%).
• Registries reported that the following conditions needed to be in place to allow SDS:
  • SDS must be described in primary researcher’s application (79%)
  • SDS must be covered in Data Use Agreement (73%)
  • Shared dataset must be de-identified (68%)
  • SDS must be reported to the Registry (60%)
  • Registry must be acknowledged in publications (42%)
  • Data must be shared through a secure data repository or enclave (42%)
  • Registry IRB must approve SDS protocol and primary protocol (42%)
  • No local or state level data included in shared dataset (24%)
  • Ages must be grouped in shared dataset (21%)
  • Shared data may not be downloaded (18%)

Key Recommendations:

• To extract the full benefit of cancer registry data, it is important for registries to share information with researchers, as well as with federal agencies and between states, to advance our knowledge of this group of diseases.
• Registries are encouraged to share data obtained from the National Death Index in accordance with allowable release. See NAACCR NDI Fact Sheet naaccr.org/data-security-confidentiality-issues/.
• Registries should develop written policies regarding secondary data release that align with the NIH Data Management and Sharing Policy.
• Registries with restrictive data sharing policies should advocate for strategies that encourage responsible access to cancer data. Registries with successful data sharing policies could serve as models providing key language and procedures for registries struggling with these issues.
• NAACCR and its partners could develop model data sharing policies that could be used in multiple registries, and a bank of existing data sharing policies could be assembled for use in developing a universal policy throughout the U.S.