Data and definitions

This page tells what every code set is built from and what the words on a code set page mean.

Sources

FDA prescription drug labels (DailyMed). These give the indications. Labels of discontinued products are included.
FDA NDC Directory and NSDE. These give the NDC codes, their products and their marketing dates.
RxNorm (National Library of Medicine). This links NDC codes to medications.
CMS HCPCS crosswalks. These give the HCPCS codes.
FDA established pharmacologic classes and the WHO ATC classification. These give the medication classes.
UMLS Metathesaurus (National Library of Medicine). This matches the words of a request to conditions.

The notices these sources ask for are on Sources and notices.

What a code set covers

Indications are as the labels state them.
An off-label use is included only if the AI raised it. Each one shows its source.
OTC products are not complete. The data has no OTC labels. An OTC NDC is included only if RxNorm links it to the same medication, strength and form as a prescription label that states the indication.
Some medications have no FDA class or no ATC class. A code set page shows them as "not classified".

Definitions

Medication: An active ingredient, or a combination of active ingredients.
Product: One marketed form of a medication, sold under one name. The name is a brand or a generic name. A medication is included whole, or product by product.
Indication statement: What a label says the medication is for. An included medication cites the statements it rests on. Each statement shows the Indications section of its labels, and a link to each label on DailyMed.
Off-label use: A use that the labels do not state. The page shows the source the AI read, or where to verify it.
Other sources for this use: Sources other than the labels that the AI examined, for example a guideline, a payer policy or a publication. A tag tells what each source did for the decision: "supports the inclusion", "not needed: included on its label", "not used for the inclusion" or "does not support the use".
AI Explanation: Text that the AI wrote when it decided.
label: A label states the use.
label: AI-provided link: A label states a use. The AI tied that use to the request from AI knowledge. Check it.
exposure: A request for a medication, or exposure to it: the product contains it. Every use counts.
class: A request for a class: the medication is in the class, by ATC or FDA, its own class or that of an ingredient. Every use counts.
off-label: documented: The AI read a source for the off-label use. The card names the source.
off-label: AI knowledge: The AI read no source for the off-label use. Check it.
Other indications: Other uses that the medication's labels state. Its codes are used for these uses too. Uses that labels word alike but for case or punctuation are shown once. Click "and N more" to see them all, in alphabetical order.
Not Included: A medication that the AI examined and left out. Each one shows why.
Left out without a decision of the AI's: A medication that the rules leave out, for example an assembled kit that contains an included medication.
Class: A code set page groups medications by FDA class or by ATC class. A medication is shown once, under the class that most of its labels give, or that suits how its products are given. A medication that would be alone in its class is shown under another of its classes, if that class is on the page.
ATC class: WHO gives a substance one ATC class for each way it is used. For example, dexamethasone has classes for the eye, the skin and the whole body. A card shows the class that suits how its products are given. Click "and N more" to see the others.