FHIR Chat · GF#19520 · terminology

Stream: terminology

Topic: GF#19520


view this post on Zulip Grahame Grieve (Dec 06 2018 at 23:38):

GF#19520 says:

Verify (will check with NLM) that the existing text

@Ted Klein @Rob Hausam Has this "check with LM" happened?

view this post on Zulip Rob Hausam (Dec 06 2018 at 23:40):

It hasn't to my knowledge, unless @Robert McClure or @Carol Macumber possibly may have done it?

view this post on Zulip Grahame Grieve (Dec 07 2018 at 06:16):

@Robert McClure @Rob Hausam @Ted Klein the resolution of this task says

Verify (will check with NLM) that the existing text

Have we checked with NLM?

view this post on Zulip Rob Hausam (Dec 07 2018 at 11:09):

I suppose maybe should have been more tentative about whether we would do that. :( If we don't hear from @Robert McClure maybe I can try to reach out to someone on the RxNorm team today.

view this post on Zulip Carol Macumber (Dec 07 2018 at 17:27):

I've asked our informaticist that works directly with the RxNorm tables regularly, i'll report back when i have an answer

view this post on Zulip Carol Macumber (Dec 07 2018 at 18:36):

The query as written can bring back multiple rows (one for each Term Type available) . As an example for SCDs, this is what we get back from rxcui = 1375932 (format TTY|STR).

  • PSN|acetaminophen 650 MG / diphenhydrAMINE HCl 25 MG / phenylephrine HCl 10 MG in 20 mL Oral Solution
  • SCD|Acetaminophen 32.5 MG/ML / Diphenhydramine Hydrochloride 1.25 MG/ML / Phenylephrine Hydrochloride 0.5 MG/ML Oral Solution
  • TMSY|Acetaminophen 32.5 MG/ML / diphenhydrAMINE Hydrochloride 1.25 MG/ML / Phenylephrine Hydrochloride 0.5 MG/ML Oral Solution
  • TMSY|APAP 32.5 MG/ML / diphenhydrAMINE Hydrochloride 1.25 MG/ML / Phenylephrine Hydrochloride 0.5 MG/ML Oral Solution
  • TMSY|APAP 650 MG / diphenhydrAMINE hydrochloride 25 MG / phenylephrine hydrochloride 10 MG per 20 ML Oral Solution
  • TMSY|APAP 325 MG / diphenhydrAMINE hydrochloride 12.5 MG / phenylephrine hydrochloride 5 MG per 10 ML Oral Solution
  • TMSY|acetaminophen 325 MG / diphenhydrAMINE HCl 12.5 MG / phenylephrine HCl 5 MG per 10 ML Oral Solution
  • TMSY|acetaminophen 650 MG / diphenhydrAMINE HCl 25 MG / phenylephrine HCl 10 MG per 20 ML Oral Solution

If we want to provide more guidance to implementers on which they might choose, I looked at the same rxcui in RxNav using the NLM's REST API (https://rxnav.nlm.nih.gov/REST/rxcui/1375932). We get the following name returned "Acetaminophen 32.5 MG/ML / Diphenhydramine Hydrochloride 1.25 MG/ML / Phenylephrine Hydrochloride 0.5 MG/ML Oral Solution"

Thus, RxNav is using, as their "display name" the string with TTY = SCD (coincidentally the same as a Tall Man Lettering String Synonym, TTY=TMSY). Same pattern holds true for another example , rxcui = 1046630 (https://rxnav.nlm.nih.gov/REST/rxcui/1046630) where TTY = SBD.

view this post on Zulip Grahame Grieve (Dec 07 2018 at 20:30):

so the question is about TTY? SCD or SBD?

view this post on Zulip Rob Hausam (Dec 07 2018 at 20:44):

I agree it's a good idea to follow the RxNav precedent. In this example I don't see a TMSY string that completely matches the SCD string (there's a case difference with 'diphenhydrAMINE'?). It sounds like we can make the query more specific and (hopefully) match the RxNav choice. In any particular instance you should have only SCD or SBD, I believe, so then the query could be:

Select STR from rxnconso where RXCUI = :code and SAB = 'RXNORM' and TTY in ('SCD', 'SBD')

I think that would do it? The only question I have with that is whether it leaves out any instances - if so the query might need to be a bit more complex with a "fallback" case of some type. I think that might happen with obsolete drugs as they don't follow the same patterns always as current ones, but I would have to check to be sure - and I don't know if we need to worry about the obsolete drugs in this case.

view this post on Zulip Grahame Grieve (Dec 07 2018 at 21:02):

I found plenty of codes without SCD or SBD TTYs

view this post on Zulip Rob Hausam (Dec 07 2018 at 21:03):

I'm not surprised. The question is can we determine a pattern for those to get a "preferred" display string.

view this post on Zulip Rob Hausam (Dec 07 2018 at 21:08):

It would be easiest to stay with the query as originally specified and know that multiple strings normally are expected to be returned, all of which could be considered to be "correct" (i.e. usable as desired in a valid instance).

view this post on Zulip Grahame Grieve (Dec 07 2018 at 23:27):

well, that would be clear, but is it correct?

view this post on Zulip Rob Hausam (Dec 07 2018 at 23:30):

The bottom line is that I'm not sure if they've actually decided what is "correct". I've worked with RxNorm quite a bit in the past, and I'm pretty certain that there is no designated "preferred" term (or set of terms). And as far as I am aware this isn't incorrect, so that might be as good as we can get. That's my take at the moment.

view this post on Zulip Rob Hausam (Dec 07 2018 at 23:31):

If you want I can read through the documentation on this aspect again, as I haven't done that recently.

view this post on Zulip Grahame Grieve (Dec 08 2018 at 00:52):

better to send an email to Olivier + Patick/Steve and the team asking them to confirm that there is no particular preferred display for a CUI, I think

view this post on Zulip Rob Hausam (Dec 08 2018 at 02:15):

Sure, would be good to confirm.

view this post on Zulip Robert McClure (Dec 09 2018 at 16:18):

A couple of clarifications. It should be true that any active concept that is a SBD is also linked to a SCD, but it will not have a SCD name as part of it's record, you'd have to follow the concept model. Also to get all concepts that represent general (some say generic) concepts for things ordered or dispensed will also need to include GPCK ( so you might also want BPCK) because of things like a "z-pack." A GPCK will also be linked to the component SCDs. SO I think those two TTYs need to be added.

As for why there are many concepts in rxnorm that do not have tty SBD or SCD, assuming we're not tripping on understanding the basic concept model, perhaps it's because of retired concepts, concepts that are from NDFRT, and other concepts in the files that are not current, active rxnorm concepts. I'd understand better if @Grahame Grieve you gave an example of the concepts that don't have SBD or SCD TTY.

Ah, as for the GF ticket and what string to display, I'd say that needs to work for any TTY, not just the blessed ones we just cited because we certainly also need ingredient types (IN, BN, PIN, MIN) and likely any other TTY for other situations, so we need a display selection query that is not focused on a particular TTY for the concept. I'd ask Olivier how he's doing it. Or ask Patrick.

view this post on Zulip Rob Hausam (Dec 09 2018 at 16:24):

Agree with that to be complete. Let's ask them what they recommend.

view this post on Zulip Grahame Grieve (Dec 09 2018 at 20:12):

looking at the rxnorm codes used in FHIR examples:

view this post on Zulip Grahame Grieve (Dec 09 2018 at 20:12):

1049623  PSN, TMSY, SBD, SY (Roxicodone 5 MG Oral Tablet)
1114879  TMSY, SCDC  (Dopamine Hydrochloride 40 MG/ML)
1160593  SCDG  (cashew nut allergenic extract Injectable Product)
12574  IN  (gemcitabine)
1313112  SY, SCD  (Phenytoin 25 MG/ML Oral Suspension)
1594660  SBDC  (alemtuzumab 10 MG/ML [Lemtrada])
161  IN  (Acetaminophen)
17767  IN, TMSY  (amLODIPine)
200371  PSN, SY, SCD  (Citalopram 20 MG Oral Tablet)
204520  SY, SY, SY, SCD  (Potassium Chloride 2 MEQ/ML Injectable Solution)
206819  PSN, TMSY, SY, SBD  (Lorazepam 2 MG/ML Injectable Solution [Ativan])
213293  PSN, SY, SBD  (capecitabine 500 MG Oral Tablet [Xeloda])
221167  PIN  (TAZOBACTAM SODIUM)
226827  PSN, SY, SBD  (Azithromycin 500 MG Oral Tablet [Zithromax])
2556  IN  (Citalopram)
285018  PSN, SBD, SY  (Lantus 100 UNT/ML Injectable Solution)
308047  PSN, TMSY, SCD  (Alprazolam 0.25 MG Oral Tablet)
313002  SY, SY, SY, SCD  (Sodium Chloride 0.154 MEQ/ML Injectable Solution)
315266  SCDC  (Acetaminophen 500 MG)
316365  SCDC  (Nitroglycerin 0.4 MG)
329528  TMSY, SCDC  (Amlodipine 5 MG)
358793  SCDC  (mometasone furoate 0.05 MG/ACTUAT)
40048  IN, TMSY  (CARBOplatin)
4917  IN  (Nitroglycerin)
562721  SCDF  (Amylases / CELLULASE / Endopeptidases / Hyoscyamine / Lipase / phenyltoloxamine Oral Capsule)
6918  IN  (Metoprolol)
746763  PSN, SY, SBD  (200 ACTUAT Albuterol 0.09 MG/ACTUAT Metered Dose Inhaler [Proventil])
763179  PSN  ,TMSY, SY, GPCK  ({48 (Prednisone 5 MG Oral Tablet) } Pack)
7980  IN  (Penicillin G)
82063  PIN, TMSY  (oxyCODONE Hydrochloride)
856907  PSN, TMSY, TMSY, SY, SBD, TMSY, SY  (Vicodin 5/500 (hydrocodone bitartrate / APAP) Oral Tablet)
866426  SCDC  (Metoprolol Tartrate 25 MG)
866435  SCDC  (Metoprolol Tartrate 50 MG)
884308  PSN, SY, SCD  (Nystatin 100 UNT/MG Topical Ointment)
884375  SY, PSN, SBD  (Nystatin 100 UNT/MG Topical Ointment [Nystex])
901813  TMSY, SCDC  (Diphenhydramine Hydrochloride 25 MG)

view this post on Zulip Carol Macumber (Dec 10 2018 at 18:53):

Rob H, did you send Olivier or Patrick an email? Let me know if there's anything I can do.

view this post on Zulip Rob Hausam (Dec 10 2018 at 19:49):

Done, and copied you and the rest of us.

view this post on Zulip Robert McClure (Dec 12 2018 at 01:58):

@Grahame Grieve What is this supposed to be? Clearly the strings in the () are not an example of all the TTYs you have there, so what are you showing?

view this post on Zulip Grahame Grieve (Dec 12 2018 at 02:14):

It's a list of every RxNorm code used in the spec, and for each code, what type of descriptions they have, and then one (random) description for the thing. I didn't bother thinking about which - just to give you a sense of what it was


Last updated: Apr 12 2022 at 19:14 UTC