search

cdisc-org / ddf-core-poc

This repository will contain the results from the Proof of Concept project.
MIT License
0 stars 1 forks source link

Define rule: A study with type "INTERVENTIONAL" must have at least one StudyIntervention #81

Open rbakercdisc1 opened 1 year ago

rbakercdisc1 commented 1 year ago

CHK0099: The StudyDesign has a StudyIntervention. This StudyIntervention is a Procedure. The StudyIntervention has AgentAdministration. AgentAdminstration has a name, label, and description. The AgentAdministration has route, frequency, duration, and quantity.

How would this link to Biomedical Concepts? SDTM EC/EX BiomedicalConceptSurrogate? EC if blinded, EX if open-label. Are these pieces covered in the BC? It seems the USDM Entities noted will inform the BiomedicalConceptSurrogate for the study drug (aka AgentAdministration).

DianeWold commented 12 months ago

A StudyDesign has 0 to many StudyInterventions. It is possible to have a study that does not have a StudyIntervention. I think the description of the rule is sort of backward. There can be many Procedures in a StudyDesign, some of which may be StudyInterventions. I'm not certain the model is correct in saying that a StudyIntervention must have and AgentAdministration, unless we define AgentAdministration very broadly, to include, e.g., talk therapy, education, radiation, surgical procedures.

rbakercdisc1 commented 11 months ago

If trialTypes has any interventional, there must be a study design with at least one StudyIntervention of AgentAdministration

if intervention is single group, it must have only one studyintervention....??

***Check scope of the model to make sure it is only interventional study with an agent that is administered. @daveih what is the scope of the model?

DianeWold commented 9 months ago

In USDM v2.6, for StudyDesign a StudyIntervention is optional (0..1) while a StudyIntervention must have at least one AgentAdministration (1..*). A StudyIntervention is a kind of Procedure, but since a StudyIntervention is optional, I don't think it's true that there must be at least one procedure that is a StudyIntervention. In any case, Procedure would not need to be included as a class in this rule.

@BSnoeijerCD and I have discussed whether a StudyIntervention needs to be something that has an AgentAdministration. AgentAdministration may not be relevant for interventions such as surgery , education, or talk therapy. The cardinality for AgentAdministration may change in Phase 4.

In USDM 2.9, StudyIntervention is still optional (0..1), so I think we should close this rule as not planned.

BSnoeijerCD commented 8 months ago

@DianeWold : see jira ticket 519. This will be updated for version 3.0. The rule as stated here then is no longer valid. We can specify as: at least 1 intervention specified and linked to a procedure in case of an interventional study? Even an intervention is not always applicable as we also want to include observational study information in the future.

DianeWold commented 8 months ago

@BSnoeijerCD I agree that if observational studies are in scope, then Procedures will not always include a StudyIntervention. I also agree that some kinds of interventions will not have AgentAdministrations. All that's left, then, is that if Study type (from SDTM codelist Study Type Response) is "INTERVENTIONAL" then there should be at least one StudyIntervention.

BSnoeijerCD commented 8 months ago

@DianeWold : OK. thank you. I will change the short description in the rule overview accordingly.