Under the Hood of the NAACCR Call For Data Edits Metafile

During the annual Call for Data (CFD), NAACCR receives data from central cancer registries for evaluation, certification, and publication. To support the CFD and to assist the central registries in preparing the data, NAACCR releases a special Call for Data edit metafile, separate from the standard edit metafile that central registries use throughout the year to validate incoming and consolidated data. This article describes features of the CFD metafile, focusing on the differences between the CFD and the standard NAACCR metafiles.

TIMELINE

Both CFD and standard metafiles are built on edits and edit sets that are carried forward from year to year, with modifications and additions responding to continuous changes in data item definitions and instructions for the standard metafile, plus research requirements for the CFD metafile. A major version of the standard metafile is released every year, following publication of a new version of the NAACCR Data Dictionary, with one or two subsequent minor releases for updates. The major and minor releases are used to edit every incoming and consolidated data record for the diagnosis year, as well as data records from previous years in any review status. The CFD metafile is released once a year, with possible subsequent corrections, for use on a one-time basis on diagnosis years identified in CFD specifications.

Some registries may choose to use the CFD metafile as a quality control process over the succeeding year, with limitations. The CFD metafile does not contain the full complement of edits required by any standard setter, and many CFD edits do not check all the criteria in similar standard edits.

Years for CFD Data

Three signal years are identified in the Call for Data specifications, with implications for the construction of the CFD metafile.

• Submission year – current calendar year
  • Determines the version of the NAACCR Data Dictionary and the standard edits reviewed for possible updating of CFD edits
• Certification year – two years prior to the calendar year; upper end of date range for data requested for publication
  • Critical data items identified for certification year
  • Data items, including critical items, identified for publication years
  • Historical data items included
  • Data items new for submission year may be included in publication items
  • Edits selected for all data items
  • Edits on critical data items collected into separate edit set
• Evaluation year – one year prior to the calendar year
  • Data items, including critical items, identified for evaluation year
  • Many data items common to publication and evaluation
  • New data items included
  • Historical data items not included
  • Edits selected for all data items
  • Same edits for data items common to publication and evaluation
  • Edits on publication and evaluation data items gathered into one edit set

For the 2025 Call for Data, for example, the submission year is 2025, the certification year is 2023, the publication date range is 1995-2023, and the evaluation year is 2024. Central registries may create separate data files for certification/publication and evaluation years (for 1995-2023 and 2024 in the example. The metafile does not distinguish between publication and evaluation (2023 and 2024) edits. Start and stop dates in Individual edits determine the cases by diagnosis date to which they apply.

EDIT SOURCES

Modifications and additions to edits in the standard metafile follow standard setter requirements as published in the NAACCR Data Dictionary; program code manuals for the Surveillance, Epidemiology, and End Results (SEER) program and the Commission on Cancer (CoC); staging protocols for the American Joint Committee on Cancer (AJCC); SEER data collection systems for Extent of Disease (EOD), Pediatrics (PDCS), Solid Tumor Rules, Hematopoietic Database, and Cancer PathCHART (CPC); and other manuals for Summary Stage, Grade, and Site-Specific Data Items (SSDIs).

The edits in the CFD metafile start with the standard edits. The selection of standard edits follows the selection of data items included in the CFD specifications, created by the NAACCR Program Manager of Data Use and Research. As items are added to or removed from the specifications, relevant edits may also be added or removed. As edits are modified or changed in the standard metafile, the same edits in the CFD metafile may also be changed, depending on criteria. Standard edits may also be brought into the CFD metafile as submission edits or later changed to submission edits.

Acceptance Criteria

Given the three-year timespan between the CFD year and the certification year, standard edits modified for the NAACCR Data Dictionary year, for example 2025, must be evaluated for their impact on 2023 and 2024 data. The general principle is that updates to CFD edits should not result in more restrictive editing than was available at the time the case was initially created. For edit modifications that will impact the results of previous editing, the following criteria will be considered:

• Number of cases that will be affected
• If changes can be made globally by logical data conversions across cases, or if they must be made by manually updating individual cases
• Years of data affected, cases most frequently accessed within the last five years, less frequently within an additional five years, or primarily for trend studies over ten years
• Assessment of importance of the change for data quality and use
• Concurrence of standard setters collecting and using involved data item(s)

Submission Edits

Submission edits are standard edits that have been modified to meet reporting criteria stated on the NAACCR Annual Call for Data Items. In most cases the criterion is “as available,” meaning that the data item is allowed to be blank. The second common criterion is a date or date range for which the item is required. The Administrative Note for any submission edit in the CFD metafile identifies why the standard edit was changed into a submission edit. For recent years the Administrative Note also identifies the initial standard edit by edit tag and the CFD year for the submission edit. The submission edit will usually have the same name as the standard edit with “Subm” added to the edit name.

EDIT SETS

Edit sets are built for the standard metafile based on standard setter requirements and special purpose use. The metafile includes edit sets for Canadian Council of Cancer Registries (CCCR), the National Program of Cancer Registries (NPCR), the CoC, and SEER. Special purpose edit sets include quality check edits, SSDI edits, Inter-Registry Data Exchange edits, and National Death Index edits.

The entire CFD metafile is a special purpose metafile. Edit sets are divided into two groups by standard setter identified as United States, three edit sets, and Canada, one edit set. The three special purpose edit sets for the United States include one with edits required for certification and two with edits required for publication/evaluation. The second of these two edit sets, NAACCR CINA Edits-US Excluding NPCR CFD Edits is unique in referencing an outside CFD metafile.

NAACCR CINA Edits-US Excluding NPCR CFD Edits

NPCR publishes a Call for Data and releases a CFD metafile over basically the same time-frame and for the same data years as NAACCR. NAACCR and NPCR have released a common metafile in the past, but in recent years have used a cooperative approach to two separate metafiles. The NPCR CFD metafile generally includes more standard edits and fewer submission edits. The two metafiles do have some older submission edits in common.

NPCR has published specifications and metafiles earlier than NAACCR, allowing NAACCR to develop this edit set as an alternative to NAACCR CINA Edits-US. This edit set excludes all standard edits that are included in the NPCR CFD metafile, common submission edits, and all edits with stricter requirements for the same data items in the NPCR CFD metafile. For registries that report to both NAACCR and NPCR, this edit set eliminates redundant edits and shortens runtime and edit processing.