OHDSI Home | Forums | Wiki | Github

Route 'standard' concepts not standard anymore?

It’s a different story.
You run the “boiler” only against the concepts. You’d not consider the route of administration, only the form of a drug product.

Some sources provide the ingredient only + route of administration. You’d not map them to drug forms since they are more/less specific than forms. And a substantial part won’t end up in any form. That’s why routes are being mapped separately to the route_concept_id where we need a supertype “Injectable” placed above the “Intravascular”, “Intradermal”, “Subcutaneous” and some others.

Also, I’m thinking about the “parenteral” and “systemic” administration being a part of this hierarchy.

Looks like the SNOMED hierarchy of routes of administration is a little lopsided. While there is a rich hierarchy for Digestive tract route, there is nothing for systemic or parenteral routes. IV, intramuscular or subcutaneous are direct desendants of the generic Route of administration value. We may want to talk to SNOMED, and in the meantime put in a manual fix.


I would introduce the following categories:
Systemic Parenteral , Systemic Enteral , Systemic Topical (Nicotine patches are probably the systemic topical).

Local topical, local intra-organ

And link SNOMED concepts to them

Injection without notion of systemic or local doesn’t make sense for me as, for example, intravenous and intraarticular injections have totally different clinical meaning.

In contrast to what parenteral? Aren’t all alternatives are already in topical?

In contrast to topical enteral? Does it exist?

Sonds like Transdermal route. Want to put it on the top of the hierarchy?

In order to exclude Transdermal route? Is that the point?
How would you know the distribution of the drug? It depends on the drug rather than on the route.
There is no strict connection between the actual route and the drug’s penetration into the bloodstream.

Many sources show the predominance of the “injectable” route. We don’t want to drop it.
It seems we have one option only. Assume they’re all injections to the bloodstream (or close to it - IM, subcutaneous, etc.) since other is not specified. And map them all to the “injectable systemic” leaving the “intraarticular” (and related) outside this hierarchy.