OHDSI Home | Forums | Wiki | Github

NULL concept_name in Vocabulary version v5.0 10-NOV-15

Vocabulary Users
#1

Loaded vocabulary version v5.0 10-NOV-15 on 11/23/2015 had the following errors.

Following load errors in vocab 5:
16,807 concepts have NULL concept_name, but in DDL concept name is NOT NULL in ddl

16,807 concept_synonyms have NULL concept_synonym_name, but concept_synonym_name is NOT NULL in ddl

select vocabulary_id, count(*) from concept where concept_name IS NULL group by vocabulary_id;
vocabulary_id count
GPI 16,367
Read 158
ICD10CM 282

I also noticed that a large number of NDC concept have been deprecated. Is this expected?

vocabulary_id count valid deprecated updated
NDC 676,286 427,675 248,611 0

Query:
select vocabulary_id, count(*), sum( case when Invalid_reason IS NULL then 1 else 0 end ) as valid
, sum( case when Invalid_reason = ‘D’ then 1 else 0 end ) as deprecated
, sum( case when Invalid_reason = ‘U’ then 1 else 0 end ) as Updated
FROM vocabulary_5_0_NOV_15.concept
WHERE vocabulary_id = ‘NDC’
group by 1;

Has anyone seen similar results?

#2

@DTorok:

Thanks. It’s very good to know somebody is watching and we don’t make mistakes.

GPI: We don’t have a source currently that gives us the concept_names. So, we put a space in (not nothing). We are working on this problem. However, the mapping is correct, so you can use those records just fine.
Read: They are all deprecated and wrong entries. We should probably delete or rename them.
ICD10CM: This is not ready. Don’t use it.

For NDC, we are re-working the entire list as we speak. We put in the actual time intervals in which they were marketed. So, an NDC with the dates 1-Jan-2014 and 31-Dec-2014 was only valid in 2014. For you that is almost irrelevant, unless you want to study these NDC movements. The relationships Maps to point to the correct RxNorm. We are re-factoring these too. Quality of NDC will be high (if such a thing is even possible).

#3

@Christian_Reich Is this still a work in progress? We have seen this issue in our most recent download of the vocabulary (June 2016) both in the concept and concept_synonym tables.

#4

When you say this issue, what item in the above thread are you referring
to. The prior thread discusses two items: 1) Null values for concept names
and 2) a large increase in the number of deprecated NDC’s from the previous
release?

#5

@burrowse:

Both

select * from concept where concept_name is null
select * from concept_synonym where concept_synonym_name is null

return an empty list. What are you seeing?

#6

I think it depends on how you load the data, if it is not null its a single or two space(s). Here are a few examples:

concept_class_id concept_code concept_id concept_level concept_name domain_id invalid_reason standard_concept valid_end_date valid_start_date vocabulary_id
GPI 43993002301208 45406247 Drug D 2015-05-05 1970-01-01 GPI
GPI 86309902214210 45407671 Drug D 2015-05-05 1970-01-01 GPI
GPI 86309902402020 45407672 Drug D 2015-05-05 1970-01-01 GPI

In exploring the codes, they are all Deprecated concepts so they should not be used in any event for coding. We just wanted to bring it to your attention so you are aware.

t