Skip Navigation Bar

Unified Medical Language System® (UMLS®)

RxNorm Source Inclusion Evaluation Criteria and Background Questions

Version 1.0 - April 24, 2017

To consider a new source for inclusion into RxNorm, the National Library of Medicine (NLM) needs to understand details about your source and why you would like NLM to add your source to RxNorm. Answering the questions below will provide NLM with a better understanding of your source and if it meets the criteria for inclusion. Please provide as much information as possible to assist NLM in making an informed decision regarding the inclusion of your source into RxNorm. NLM reserves the right to change these questions and criteria at any time.

You can download an editable version of the questions below and then submit the completed document to

Inclusion Evaluation Criteria

Value to RxNorm

NLM wants to understand the use cases for your source and its user communities.

  • What are the use cases for your source?
    • Which user communities will the source serve?
  • How will adding your source to RxNorm contribute to interoperability among drug resources?
    • For example, will the addition of your source help link currently available electronic resources, such as patient records or data elements?
  • Do you include national drug code (NDC) data?
    • How is your NDC data maintained?
  • In addition to RxNorm, NLM maintains the Unified Medical Language System (UMLS). The UMLS is a large biomedical thesaurus with nearly 200 different vocabularies and is updated twice a year. All sources included in RxNorm are also included in the UMLS.
    • Considering the content of your source and its update frequency, would your source be a better fit in the UMLS?
    • Why should your source be included in RxNorm instead of directly in the UMLS?

Source Content

NLM wants to understand your source content, including the organizational principles and editorial guidelines used to create your source.

  • Are thesaurus characteristics and principles used to create your source?
    • Is it concept-oriented (i.e., organized by meaning)?
    • Are there preferred terms or synonyms?
    • Do you have codes/identifiers associated with names?
    • Is there a principled hierarchy?
    • Do explicitly defined relationships exist between the terms, such as broader than/narrower than relationships?
    • Are there definitions or other associated information?
    • Are there mappings or connections to other vocabularies?
  • Do you use standard naming conventions for drug information terms (e.g. United States Adopted Names (USAN), International nonproprietary names (INN), the Unified Code for Units of Measure (UCUM))?
    • What standard conventions do you use?
  • Are the meanings of your terms clearly distinct when seen alone, rather than in the associated context? For example, the term “calcium” seen alone can refer to the element or a drug class.
    • Will there be many cases where the ‘face’ meaning of a term is different from the meaning of the same term in other areas of biomedicine?
  • What parts of your source would NLM be permitted to publish in RxNorm?
    • For example, will the drug name and NDC be published in RxNorm?
  • Are your generic, trade, and branded products fully differentiated with respect to identifiers and NDCs?
  • Do you maintain current information about whether drugs are off-the-market?
    • How is off-the-market data represented and with what metadata?

Technical Process and Development

NLM wants to understand the technical aspects of your source.

  • Is the source available in a well-structured, computable electronic format?
    (NLM cannot accept printed publications, nor convert them.)
    • What is the format?
    • What is the data size of your source?
    • How many terms are in your source?
    • Is a sample set of files and documentation available?
    • Is there a technical contact person for questions about the files?
  • Will you provide your source data files to NLM free of charge?
    • If yes, how will you provide your source data files to NLM?
    • Will there be an automated download capability?
  • Do you provide fully specified drug names with ingredient, strength, and dose form in one data field or must the fully specified names be built across fields/tables?
    • For example, RxNorm currently concatenates ingredient, strength and dose form information to create a final drug name for some sources. Will your source require this type of data processing?

Background Questions

NLM wants to understand additional details about your source.

  • Is your source actively maintained?
    (RxNorm will only accept sources that are actively maintained.)
    • What type of ongoing support does your source receive to ensure that it will remain a viable asset to RxNorm?
    • Is your source and its updates officially sponsored?
    • If so, please describe by whom and how your source is sponsored?
    • Who is the responsible party or parties for your source?
  • How often do you update your source content?
    • What is your update model?
  • Who is the source author?
    (Some sources are authored and maintained by different organizations.)
    • Who creates your source content?
    • Who owns the source content?
  • Do you have a mechanism to maintain the integrity of your source?
    • How do you maintain the integrity of your source?
  • Is there a point of contact if there are content related questions?
    • Do they have computer connectivity, equipment, and support?
  • Do you have an online browsing resource that NLM editors can use to access your source data or additional data about the drug information in your source?

Terms of Agreement

Sources included in RxNorm follow the terms of the UMLS Metathesaurus License Agreement. Please examine the UMLS Metathesaurus License Agreement and resolve any questions about copyright ownership.

  • According to Appendix 1 and section 12 of the UMLS Metathesaurus License Agreement (and outlined on the License Category Help page), what have you determined your license restriction level to be (0, 1, 2, 3, or 4)?