Lab Terminology - Package 37

A. CT Discussion from 8:30-9:00 and 9:45-10:30.

B. BE domain review 9:00-9:30 with Takeda: Mike Hamidi, Sam Kitamura, John Smutko.

C. Chris Connolly at 9:30-9:45 – BE/BS domain team call for volunteers.

A.

B. BE domain review 9:00-9:30 with Takeda: Mike Hamidi, Sam Kitamura, John Smutko
i. Review Takeda BE domain example for input: Please find attached an updated spreadsheet with notes from today’s PGx team discussion. Minutes are also here: https://wiki.cdisc.org/display/PGX/2018-09-18+PGx+Team+Meeting.
ii. Discussion topics to cover
a. Where to put parent sample information? In the example it is important to note whether the RNA was from peripheral blood or bone marrow. Was the BE domain intended to cover this?
i. Possible solutions: Use BE or BS domains and link to PF; use –LOC in PF domain; create a NSV in PF domain of ‘original sample’
ii. Mike Hamidi suggestion: For SAMPMATRIX, my recommendation is to align with BE with SUPPBE of specimen. Utilize the associated ACCNO to link BE and PF.

C. Chris Connolly at 9:30 – BE/BS domain team call for volunteers
i. Scope question: would this only include submission date or would it also include operational data? Not sure yet, good scope question. Pharma companies are becoming interested in this but they would not put this info into their submission dataset.
ii. Team will kick off in Q1 2019. Chris would like the lab team to review the team proposal.

Team Notes:

1. BE was intended to capture parent biospecimen events and what you did with them. From collection to derivation of the final product that was to be assayed.
2. BS was intended to capture parent or child biospecimen assessments. Relates to events in BE.

A. Sam asks, is BSSPECCOND a valid variable or should it be SUPPBE.SPECCOND? BE is meant to only document the event and does not have a structure to support assessments. This information shoudld really be in BS.

B. Sam asks, there is no easy way in the PF domain to indicate that the RNA used in the assessment came from either peripeheral blood or bone marrow, say. One should use BE to indicate that you have derived a new specimen (RNA) from the parent spec (BM or PB)
If RNA is never included in the vendor data (say you were only provided BM or PB), i.e., the vendor is not giving you all of the data.

C. Given for this example and use case, this data doesn't have the full information so we might need to present both use cases as examples: ideal data and real world (incomplete) data.

D. BE example: Q from Mike and Sam about the use of suppqual in BE. Lab team suggests that this might be changed to 'Received Condition' if you choose to create SUPPs. Ideally though (and the team's recommendation) this information should be put into a BS dataset.

E. HL7 has a specimen message within FHIR. Phil worked on it. Might be worthwhile to harmonize. CDISC is probably only interested in a subset says Phil.
Data model is in production but the messaging is still being worked on.

Summary:
-ACCNO will be associated to PFREFID to link to BS.
-No RELSPEC needed for this example
-Info in SUPPBE for specimen condition would go to BS, with same association to PF.

• Specimen:
  • SPECTYPE: Rows 11, 12 (All Done)
• Lab
  • Changes to Existing: Rows 281-284 (all Done)
  • MIFTSDTL: Row 7 (All done)
  • METHOD: Rows 10, 15 (denied by team but still has action item), 19, 23 (agree to add, EVS provides DEF), 27 (agree to add, EVS provides DEF), 28, 29, 30

  • LBTEST-CD:

    • Rows 7-18 – CDISC-3290 – Phil to follow-up at Covance

    • Rows 19-22 (CDISC-3409-3426)

    • Rows 23-73 (CDISC-3431)