If you are a standard-setter or a person who must adjust an NPCR–EDITS (Exchangeable-edits, Data-dictionary, and Information Translation Standard) metafile, the following details are very important to you. If you use a metafile prepared by someone else, the internals will be less interesting, as the metafile you receive will be an opaque binary file.
Key Components of a Metafile
NPCR–EDITS metafiles contain everything needed to edit a data file except the data. Metafiles provide portability of edits, in that the same edits can be applied to different data formats for different purposes. NPCR–EDITS metafiles are created and modified using EditWriter.
The agencies component contains contact information and agency codes for each standard-setter. Each metafile object, such as field, record layout, edit, edit set, and table, is created and maintained by an administrator. Locally defined objects do not require an agency code. Assigning an agency code and password prevents users from modifying distributed metafiles.
Agencies using cancer registry metafiles include the National Cancer Institute (NCI) Surveillance, Epidemiology and End Results (SEER) Program, the North American Association of Central Cancer Registries (NAACCR), the American College of Surgeons (ACoS), and CDC's National Program of Cancer Registries (NPCR).
The data dictionary contains all data elements, or fields, edited by NPCR–EDITS in the metafile. An administrator defines standard attributes for each field, such as field name, field type, field length, agency, choice table, and description. These attributes may be changed only by the administrator.
Local attributes can be changed by any user. These include a local name, local number, and local title.
- The local name and local title may be initialized to the field name by default, but can be changed where other names are used locally. For instance, the NCI SEER field Primary Site may be implemented as Site or Primary Site in a particular system, but its meaning is the same.
- The local number is used internally and must be unique. It can be changed to coincide with any numbering system for interactive data editing.
The data dictionary may contain more fields than are present in any single data file.
A record layout is a named group and organization of fields from the data dictionary into a particular file format, such as the NAACCR data exchange record. Many record layouts can be defined within the metafile, with each corresponding to a particular dataset or file format.
Edits contain logic for editing fields. Both single- and multiple-field edits can be written on any field or combination of fields in the data dictionary. Descriptions and error correction help may be stored with each edit. The Edit Name must be unique and consist of as many as 50 characters. An example Edit Name is the NCI SEER edit Morphology-type & behavior (SEER morph).
Edit sets are groups of edits used to edit a data file. Multiple edit sets can be used on the same data. For instance, the NAACCR metafile contains edit sets for hospital use as well as central and state use.
Error messages are referenced within an edit by their unique error message numbers. Error messages may contain text, as in "Laterality must be provided for specified paired organs/sites," or text and macros, as in "%V1 is not valid for %F1." The macros "%V1" and "%F1" are replaced with the current field contents and field title (as defined in the data dictionary).
User Lookup Tables
The user lookup tables are reference tables within the metafile for performing lookups and building choice lists. Tables can be imported or created within EditWriter. Tables formatted in DBF, binary (.BIN), and ASCII (.ASC) are supported. For best performance, indexed .DBF files are preferred over .ASC. A binary table is used to check for compatible sites and histologies in the SEER edit Primary Site, Morphology-Type Check (SEER IF25).