Benefits of Metafile Testing for Central Registries

Testing new metafiles is a great opportunity for central registries to participate in the edit process. For standard metafiles, you get to review new edits, find concept and logic errors, comment on their usefulness and impact they might have on your data, and perhaps even suggest other likely edits. For call for data (CFD) metafiles, you get to estimate pre-submission data clean-up that might be required, and also help streamline submission for every registry by detecting problems before the final posting.

Standard Metafiles

Testing is an integral part of developing and maintaining edits, and the broader exposure they have to cancer data before release, the better. Comments on testing are released with the standard metafile, along with the most recent version of the change spreadsheet, identifying corrections, modifications, updates for new standards, types of new edits to watch out for.

New edit logic and new edits are generally implemented for the new NAACCR version year, but for testing edits may be set to run from 2018 forward. Testers can assess their general impact on data, and also determine if an earlier implementation date would benefit data quality, either their own or registries’ generally, with an acceptable level of work.

Edits are grouped into test edit sets with no agency designation, so testers can easily add or subtract edits to better fit their own data requirements. New and modified edits will also be included in the standard edit sets that testers may use, and they are encouraged to run these edit sets as well as the designated “tests.”

Testers are looking for faults in edit concepts, errors in edit logic, and conflicts between or among different edit results. Have edit problems previously reported by the registry been resolved? Have any edit suggestions previously made by the registry been incorporated? Have new standards been addressed wherever required?

Call for Data Metafiles

Focusing on the NPCR and NAACCR CFD metafiles, they should both contain mature edits available for central registry use in the standard metafile. The two agencies request many of the same data items, but requirements may differ in terms of date, and whether blanks are allowed or not. The NAACCR CFD metafile is released after the NPCR metafile, and includes an edit set that builds on the NPCR edit set in excluding edits where NPCR requirements are the same or stricter for the data items being edited.

In addition to testing the CFD metafiles to assure that they run smoothly, testers have two specific chores:

• Identify edit logic that has not been part of the standard editing process for the data years involved.
• Assure that running the full NAACCR edit set does not turn up any edit errors not identified by running the NPCR edit set and the NAACCR edit set excluding NPCR edits.

As the body of cancer information expands, questions become more probing and answers more precise, edits are wielded as one tool to manage the quality of data collected. A group of registries, ready and able to effectively test the edit metafiles, is needed to support this aspect of cancer data collection. NAACCR most gratefully thanks the many registries that have participated in testing metafiles, and invites others with the time and resources to respond when a call for additional testing volunteers goes out. It can be illuminating for all.