Misclassified PPI concepts

[mark-velez]

In analyzing survey data for the All of Us Research Program we identified concepts in the PPI vocabulary where the concept_class_id appear to be incorrect. The table below lists each of the concepts, what concept_class_id we expected and the actual concept_class_id we found.

Please advise how this can be addressed and any additional information that might be needed. This is causing our data validation to fail and also makes it difficult to reconstitute survey instruments (e.g. for visualization). I appreciate any help on this.

concept_id	concept_code	expected	actual
702786	DiagnosedHealthCondition_SkeletalMuscularCondition	Question	Topic
43529966	DiagnosedHealthCondition_VisionCondition	Question	Topic
43528515	DiagnosedHealthCondition_CancerCondition	Question	Topic
43528563	DiagnosedHealthCondition_CirculatoryCondition	Question	Topic
43529158	DiagnosedHealthCondition_KidneyCondition	Question	Topic
43528678	DiagnosedHealthCondition_EndocrineCondition	Question	Topic
43528634	DiagnosedHealthCondition_DigestiveCondition	Question	Topic
43529217	DiagnosedHealthCondition_MentalCondition	Question	Topic
43529272	DiagnosedHealthCondition_NervousCondition	Question	Topic
43529767	DiagnosedHealthCondition_RespiratoryCondition	Question	Topic
1585872	AttemptQuitSmoking_CompletelyQuitAgePreferNo	Answer	Question
1586164	Smoking_AverageDailyCigaretteNumberPreferNo	Answer	Question
903063	Smoking_CurrentDailyCigaretteNumberPreferNo	Answer	Question

[cukarthik]

Tagging @mik-ohdsi

[dimshitc]

Hi @mark-velez
I'm looking at All of Us | Public PPI Codebook
and I see that these concepts have exactly the same classes as we put in the OMOP vocabulary.
Can you please fix them there and then we change them in the OMOP vocabulary then.

[mark-velez]

Thanks @dimshitc ! I am discussing this with Brandy and Henry. I will follow-up on this ticket asap.

[mik-ohdsi]

Hi @mark-velez , were you able to find out if the codebook needs to be adjusted or the vocabulary (or both, but in which sequence :-) ) Please let us know so that we can address this properly in this issue. Thanks ~ Mik

[mark-velez]

Hi @mik-ohdsi. My understanding is that we do not need to adjust All of Us | Public PPI Codebook as it is now defunct. The concept classes should just be fixed in the vocabulary itself.

Sorry for the late reply- some of the PPI vocabulary process is being overhauled and I didn't want to give conflicting information. If there are any questions on the process please e-mail Brandy and she will loop me in as needed.

[dimshitc]

@mark-velez thanks for a reply. And how soon this should be fixed?

[mark-velez]

@dimshitc it wouldn't be my place to say; I understand the survey team has other issues on deck which may be more urgent. I have asked them to follow-up with you on how to prioritize this issue.

[mik-ohdsi]

@dimshitc - is this issue fixed by now?

[dimshitc]

All concepts got the right concept_classes, except these two:
1585872 | AttemptQuitSmoking_CompletelyQuitAgePreferNo | Answer | Question
1586164 | Smoking_AverageDailyCigaretteNumberPreferNo | Answer | Question

They are deprecated, and we didn't process them.
I'm not sure whether the PPI team uses deprecated concepts.
@mark-velez do you have any deprecated concepts in your data?

[TinyRickC137]

The concept classes of AttemptQuitSmoking_CompletelyQuitAgePreferNo and Smoking_AverageDailyCigaretteNumberPreferNo have been finally changed during our latest release.