ICD9CM Codes deprecated

Implementers
1

We noticed the following ICD9CM codes are deprecated in the latest vocabulary. Any news on these being added to the vocabulary? Our dataset is focused on oncology and these codes are important to our CDM build.

V158.11, V158.12, V158.1

Thanks,
Urvi

2

@ushah:

There is no such thing as V158. V codes can only have two digits before the dot. Do you mean V15.8? There are:

V15.08
V15.8
V15.8
V15.80
V15.81
V15.82
V15.83
V15.84
V15.85
V15.86
V15.87
V15.88
V15.89

As far as I can tell all of them are nicely mapped to SNOMED.

3

Sorry typo on my part. Looking for V58.11, V58.12 and V58.1

4

One was deprecated in 2005 but the others should be current through 2014: http://icd9diags.jigsaw.io/?code=v58.1

5

Friends:

If you mean by “deprecated” non-billing, so we have all three.

> CONCEPT_CODE	CONCEPT_NAME                                              	CONCEPT_CLASS_ID
> V58.1 	Encounter for antineoplastic chemotherapy and immunotherapy	4-dig nonbill V code
> V58.11	Encounter for antineoplastic chemotherapy               	5-dig billing V code
> V58.12	Encounter for antineoplastic immunotherapy               	5-dig billing V code

But none of them are mapped. Reason is this: It is a drug concept, but it cannot be mapped to anything because “chemotherapy” and “immunotherapy” are not defined drugs at the Ingredient or below level. We need to change the CDM, if we want to use some SNOMED or ATC codes for that and make them standard concepts.

6

Thank you Christian. So in this case, how can we include these in CDM right now ? We would just have records in procedure but mapped to 0? Or should they be in drug?

7

Yes. They have the wrong domain_id. Needs to be fixed.

8

Interesting that you see these are “a drug concept”. I see them as a very specific type of encounter even though they are coming from ICD V codes… Won’t help Urvi’s use case.

9

What’s different? They are coming in to get that drug treatment. Or where am I wrong?

10

In you earlier comment, you said: “It is a drug concept, but it cannot be mapped to anything because “chemotherapy” and “immunotherapy” are not defined drugs at the Ingredient or below level.” If I am reading that text correctly, you wanted to put these concepts into the drug terminology but are unable because there are no drugs at the ingredient or below level. I agree with this statement but I don’t see them as concepts that should go into the drug terminology at all but as potential visit types for. visit_occurrence. I think we are seeing these concepts as belonging to two very different domain_id. You are focusing on the last half of the string description as the key domain-defining idea and I am focusing on the first half of the string description as the key domain-defining idea.

11

@mgkahn:

Why a visit? A visit is Inpation, Outpatient, ER or Long Term Care. That “first half” is taken care of.

12

Typically these are used in research to indicate that chemotherapy was given. See https://healthcaredelivery.cancer.gov/seermedicare/considerations/procedure_codes.html

So, I would classify them as procedures, despite the text that says “encounter”.

13

Nice overview. Will check out whether we got them all right.

Still. I would state that any of these are drug events. Patients are not going to have side effects from the visit or the procedure (other than screwing up the vein). It’s the drug that is the meaningful clinical effect, good or bad.

14

I’d personally put them in the Observation table. You’re just saying that they hand an encounter for treatment, but doesn’t say what the treatment was, what the dose is, etc, so drug domain doesn’t make sense. Maybe procedure makes sense too, but I was thinking procedures geared towards things more specific ‘like amputation of left toe’, but if it’s conveying the idea that the person actually underwent chemo/etc, then procedure makes sense to me.

15

Sorry, Christian, I have limited clinical knowledge in this space, but is “Encounter for antineoplastic chemotherapy and immunotherapy” understood to mean a specific drug ingredient (or clinical form of drug) or is it a more general term indicating treatment for cancer?

16

Hm. I don’t think there is a question that the patient is treated with a drug, and therefore the event belongs into DRUG_EXPOSURE. The problem is not what it is that happens. Only that we don’t know which drug. Hence mapping to concept_id 0. But there was a drug, and it was very real. Chemo and Immunotherapies are the most “real” drugs, they have by far the strongest effects on the body.

17

Wait…wait…mapping to concept_id 0 means you don’t know what it is, but I think from that code you can say that it’s definitely chemotherapy related…so we should be able to do better than concept_id = 0.

I understand the desire to represent drug exposures into the drug_exposure table, but I also think that they should be specific drugs you are exposed to, not a blanket term that could be any combination of chemotherapy drugs that could be used (such as those found in this list: http://www.cancerresearchuk.org/about-cancer/type/all/treatment/chemotherapy/chemotherapy-drugs-for-all-and-their-side-effects)

In this case, however, I feel like the chemotherapy is a procedure that involves 1 or more drugs which would have to be provided in some other level of detail with the medical record…Making up a completely hypothetical code: if there was a ICD9 code for ‘physical therapy’ which may involve the application of anti-inflamatory drugs, I’d argue that you’d have a physical therapy procedure, and an associated drug exposure of anti-inflamatory agent if one was so used.

I’d look at this chemotherapy case the same way…

18

@Chris_Knoll:

Well, drug exposure is defined in the CDM (first paragraph of the table definition). As much as I have empathy for how you “feel like the chemotherapy is a procedure”, the feeling is wrong. It’s a drug. The fact that we don’t know which one doesn’t change that.

And you are right: We do know more than nothing. We know it’s a chemo. So, we have had the debate whether or not to allow drug classes in the drug_concept_id field. The argument against is that we, and the CMS who invented this code, don’t know exactly which drugs exactly fall under this definition.

19

Thanks for pointing that out, I thought this part was interesting:

Drug Exposure is inferred from clinical events associated with orders, prescriptions written, pharmacy dispensings, procedural administrations, and other patient-reported information, for example:
…

Drugs administered as part of a Procedure, such as chemotherapy or vaccines.

Even the documentation is describing chemotherapy as a procedure.

20

That’s correct. You need a procedure to get parenteral drugs into the body: An infusion (or injection for some of them). So, strictly, it is both a procedure and a drug.

But: The infusion part of it we usually don’t record. It’s just too small an event. Just like we don’t record them for all other injections and infusions, or phlebotomies for Measurements. We could. But it would still require a record in DRUG_EXPOSURE.