It is useful to distinguish between code categories (for example, race codes), and the codes assigned to them (01-99). Standard codes are necessary for reporting or exchanging data and are often more convenient to use internally. Standard categories must be retained in the registry structure, however, and conversion to and from standard codes must be built into routine computer operations. Because the application of codes must be similar for codes to indicate the same thing, it is not sufficient to convert nonstandard code categories into standard codes.
There is a great need for consistent data standards. Data elements having the same intent should mean the same thing in every registry that collects them. When they do not, coders may need to code similar but not identical codes twice. Standard code categories, therefore, must meet multiple data goals. To and from standard codes must be built into routine computer operations.
The standards specify not only what codes to apply to "known" categories but also distinguish categories such as "not available in the record," "not appropriate to this case," or "not collected by this registry." The format of the code — mixed or uppercase, left- or right-justified, numeric with leading blanks or numbers with a zero fill, blanks permitted or not — is also standardized.