REDCap-i2b2 Integration: How Does This Work?

[mabaj...@mac.com]

Dear i2b2-install-help Group,

My questions today are specifically about the REDCap Import feature that was added to i2b2 in version 1.7.12. I'd like to get more information on how this works.

TL;DR... Questions 1-7 below.

According to the Release Notes:

• This is a one-way import of data from a REDCap form (or forms) in a REDCap project into an i2b2 project.
• The fields in the REDCap form will show up as items in a new ontology in the user interface/webclient for i2b2.
• The patient data associated with the REDCap project will be incorporated into the OBSERVATION_FACTS table of the i2b2 project.
• When there are updates or new insertions into the REDCap project's form, these updates and new insertions will propagate into the i2b2 project's OBSERVATION_FACTS table.
• Upon import/update from REDCap to i2b2, there is no matching between patients from the REDCap project and any other patients that may exist in the i2b2 project. That is, while the patients who are part of the REDCap study may also have EHR-origin data in the i2b2 database, i2b2 does not recognize any existing patients upon import from REDCap; all the patients imported from REDCap are treated as separate from the patients that may already exist in i2b2.

Besides what is in the Release Notes, I figure:

• It's best to set up a separate CRC database and ONT database for a new, dedicated project in i2b2.
• Then one would need to create that new dedicated project in the PM_PROJECT_DATA table.
  
• One would need to add these database resources to the crc-ds.xml file and the ont-ds.xml files, and update the DB LOOKUP tables in the hive database.
• Then one could add new users to that project, and the AGG user as well.
• When that project is working okay in i2b2, then one would add the project parameters in the PM database as outlined in the Release Notes for 1.7.12, and set up the trigger for the project in REDCap.
  

So here are my Questions:

1. Did I understand the Release Notes correctly? Is my checklist above reasonable?
   
2. Can the REDCap project have more than one Form? If there is more than one form, will the data from all the forms propagate into i2b2?
3. When the form fields are translated into new items in the ontology, into which table are they placed--the Custom Metadata table, or does the i2b2 importing process create a new table for this REDCap project's form(s)? Does i2b2 also create a new entry in the TABLE_ACCESS table to reflect this new ontology?
4. When the data from the REDCap project are incorporated into the OBSERVATION_FACT table, does the importing process create "pseudo" or "phantom" values for the ENCOUNTER_ID, PATIENT_ID, PROVIDER_ID, CONCEPT_CD, etc.? Does the import process then also add these "pseudo" or "phantom" values to the VISIT_DIMENSION, PATIENT_DIMENSION, PROVIDER_DIMENSION, CONCEPT_DIMENSION, etc. tables?
5. If a REDCap project patient's data gets updated in REDCap, will the import process into i2b2 update the appropriate records in the OBSERVATION_FACT table? Moreover, if a patient is deleted from a REDCap project, will the import process delete the appropriate records from the OBSERVATION_FACT table and other associated tables?
6. Are there any other pertinent questions that I should be asking about the import process from REDCap into i2b2?
   
7. Our target system is i2b2 v1.8.1 on a SQL Server database, importing from the latest REDCap. Has anyone encountered any issues with the import mechanism in this particular setup?
   

Thanks in advance for any responses/insights you can share!

Kind regards,

Mark Abajian

Mark Abajian, Independent Software Consultant
P.O. Box 8017, Glendale, CA  91224
Email  maba...@mac.com   |   Mobile  +1 (818) 726–0372

[Vaibhav Talap]

Hello Mark,

Thank you for your detailed email.

We will review your queries thoroughly and get back to you soon with the information that you need.


Best regards,
Persistent Systems Support Team

[mabaj...@mac.com]

Hello, Persistent Systems,

Belated good wishes for the New Year.  :)

It's been a few weeks since I first posted this topic to the group. Please let me know if there has been any progress on my questions regarding the REDCap-i2b2 Integration.

An additional question... 8. If the REDCap form gets updated with additional fields, will the ontology tree in i2b2 get updated, or is the import of the intake instrument from REDCap a one-time-only event?

Thank you for your guidance.

Kind regards,

Mark Abajian

Mark Abajian, Independent Software Consultant
P.O. Box 8017, Glendale, CA  91224
Email  maba...@mac.com   |   Mobile  +1 (818) 726–0372

[Vaibhav Talap]

Hello Mark,

Apologies for delayed response. Here are some insights for previous questions.

1. In order to setup REDcap its required to satisfy the conditions given by them which are as follows

• REDCap is specifically designed for non-profit organizations, such as universities, research institutions, and hospitals,
  engage in academic research or healthcare-related projects.
• To use REDCap, an organization must join the REDCap consortium by signing an end-user license agreement with Vanderbilt University.
  This agreement requires the organization to meet specific criteria, including being a research-based or non-profit institution.
• Since Persistent Systems do not meet foundational requirements for obtaining the REDCap license.
  

Important Note :-
Because it’s not possible to setup redcap at our environment we haven’t done hands-on at our end.
 The answers to the questions are based on information we gathered and some response is generated by AI ( Mentioned as a note below)

2.The Answers to Questions About REDCap Import Feature in i2b2 v1.7.12+ as follows:

1. Did I understand the Release Notes correctly? Is my checklist above reasonable?

Yes, your understanding almost aligns with the Release Notes, and your checklist is reasonable.

( NOTE : Setting up separate CRC and ONT databases, configuring the crc-ds.xml and ont-ds.xml files, and updating the DB_LOOKUP and PM_PROJECT_DATA tables is a good approach for maintaining isolation between projects. Adding the AGG user and setting project parameters in the PM_PROJECT_PARAM table are also essential steps.)


2. Can the REDCap project have more than one Form? If there is more than one form, will the data from all the forms propagate into i2b2?

Yes, a REDCap project can have multiple forms, and the data from all forms can propagate into i2b2.

( NOTE : But only if each form is explicitly associated with the i2b2 ontology during configuration. Each form will generate a distinct branch in the ontology tree in i2b2, representing its structure and data fields. Be aware of the limitation: renaming forms in REDCap will result in duplicate entries in the ontology.)

3. Where are the form fields placed in i2b2?

When the form fields from REDCap are imported into i2b2, they are stored in the Custom Metadata tables and are represented in the ontology browser. Specifically:

• The metadata entries are added to the TABLE_ACCESS table to represent the new ontology hierarchy.
• Field-level details and mappings are stored in tables such as C_METADATAXML and C_FULLNAME fields within the ontology metadata structure.

i2b2 does not create a new table for the REDCap project’s ontology; instead, it integrates the form’s fields into the existing ontology structure.

5. Does the import process handle updates and deletions?

• Updates: If a REDCap project patient’s data is updated, the i2b2 import process updates the corresponding records in the OBSERVATION_FACT table. However, this update happens only when triggered (e.g., via the Data Entry Trigger on form save or admin edits in REDCap).
• Deletions: The import process does not handle deletions. If a patient is deleted in REDCap, the associated records in i2b2 will remain unless manually removed. This is a known limitation of the integration.

Best regards,
Persistent Systems Support Team

[mabaj...@mac.com]

Hello, Persistent Systems Support Team,

Thank you for your detailed reply. I understand about the limitations of the REDCap licensing. I, too, was unable to secure my own copy of REDCap, working independently as I do, and so that's why I decided to reach out to this Google group.

I appreciate your distinguishing between the information you gathered, versus the AI responses in the "Notes". It's important to understand where the information is coming from.

I understand that because you were not able to conduct your own testing, you were not able to answer all my questions. Perhaps the i2b2/tranSMART team will publish additional documentation to provide answers to these and other questions.

Thank you again!

Kind regards,

Mark Abajian

Mark Abajian, Independent Software Consultant
P.O. Box 8017, Glendale, CA  91224
Email  maba...@mac.com   |   Mobile  +1 (818) 726–0372