Lab Terminology - Package 34

1. Claire West Inquiries

2. LBTEST-CD open change requests

I Claire West email 1: Uric Acid Crystals vs Urate – Email sent 2017-07-04. Phil comment: We should discuss this as a full team.  She should not use “Urate” for crystals, as it is meant to measure the substance (as she notes, it is in a dissolved form in urine, and over time it can crystalize out. The synonym situation is unclear, and if we accept the Japanese labs statement, we should create a new “Urate Crystal” term.

• Urate is synonymous with Uric Acid. It is the substance. (published)
• Acid Urate Crystal was supposed to convey the crystal form of Urate. (published)
• Based on Japanese lab description we think the Japanese lab’s description of ‘Urate’ is actually describing Acid Urate Crystals, hence they should map that to the existing published term C74912/Acid Urate Crystals.
• Team agrees to add a new term of ‘Uric Acid Crystals’ and update definition of C74912 to make clear the difference.

II Claire West email 2: LBTESTCD and LBSTRESU – Email sent 2017-08-22: Different methods may yield different result units. However, structurally these are both non-standard arbitrary concentrations (KIU/L and ARBITRARY U/ML). The lab team does not suggest creating a new term for CMVMC_PLC but instead re-using the existing published CDISC term (CMVIGMAB/Cytomegalovirus IgM Antibody) and differentiating the two via different methods and different units of measure. The different method values ensure a unique record.

III. LBTEST-CD open change requests:

• Row 7 (CDISC-2942) – pulled from P32
• Row 8-16

•  Announcement: Wintrachange in Silver Spring, MD Mar 7-9. Registration open week of Jan 15. After PHuse meeting.
• LBTEST-CD: Rows 17-21 (CDISC-3106)
• METHOD: Rows 7-12 (CDISC-3106)
• SPECTYPE: Row 10-12, 26 (CDISC-3106)

P33 Public review Comments

Open change requests 

Open change requests

• Rows 28-38 (CDISC-3106)

UNIT

• Row 9, 13-20, 22 (CDISC 3106, 3108) 

Open change requests 

• Unit: Row 8, 10-2, 21 – (CDISC-3106)
• LBTEST-CD Row 23-26 and 27-32 – (CDISC-3137)
• 9:30-10:30 - MB team to join lab team from 9:30-10:30 – Nik, Bess, Jon (cc Anna and Phil)

                                                               i.      Discuss new IS modeling strategy. May affect whether things go into MB or IS.

                                                             ii.      Anti-bug antibody produced by a vaccine goes into IS; team is suggesting to use that new proposed variable to house the name of the bug against which the vaccine is created for. Anti-bug antibody produced by a naturally acquired infection goes into MB.

I Discuss F2F Agenda (March 7-9)

Session 1:

1. Review of LOINC-CDISC mapping
1. Review mapping (not too much left)
2. Determine next steps – to send to FDA working group, to send out for public review (Shannon, Lauren, Gary W, etc.)

Session 2:

1. Updates to lab rules document (Erin to prep)
   1. Prep for P34 public Review
   2. Craig has some additions:
   3. Jordan to create drafts for new MB rules
   4. Lab examples – look at what has been done to date, work on Phil’s examples, Phil suggests working on a multiplex panel.

Session 3:

1. MB vs LB
   1. Review terms to be moved out of LB and into MB
   2. Draft communications document
   3. SEND considerations

II SPECTYPE:

• 7-9, 15, 17 (CDISC-3106)
• Row 31 (no code yet)

III Bess at 10 am – LB vs MB - update team on discussion with Shannon

BE/BS Discussion

Outstanding new term requests 

1. Nutrition example
2. New term requests 

1. • Review Nutrition example: https://wiki.cdisc.org/display/TANUTRI/Stool+-+Sample+Collection+and+Characteristics

                                                                  i.      LB or BE/BS?

                                                                ii.      From John Owen: Just for information we originally had these measurements in the BS domain, however, GGG requested these be moved to the LB domain. The rationale was that BS was originally designed to represent information related to tracking of samples as opposed to measurements that may be used in some sort of clinical interpretation (which is what these measurements are required for in the nutrition studies). If BS is decided then it might be good for the lab team to provide a rationale as to why so that I can present this to the GGG.

                                                               iii.      Lab Team discussion: There is the high level principle that an observation should only ever appear in one domain. Since BS domain is supposed to be used for things other than genomics, we agree that BS may be a lab-domain subtype which means that biospecimen-characteristics type lab tests should be moved out of LB and into BS. Team discusses whether it matters whether these tests are used for bio-analytics: The difference here is measuring a stool sample to determine whether you have enough material to do a test vs measuring a stool sample to determine how much material was consumed/absorbed vs defecated. The lab team thinks that this is an important distinction and the latter case is really more appropriate for LB and not BS. However this type of thinking does conform to the principle of one assessment/domain (not modelling based on context).

                                                              iv.      Action for John Owen - LBCAT value doesn’t seem right. Either don’t put it in the dataset at all or change the existing value. Team thinks it doesn’t add value at all so the column should just be removed. The specimen type is already specified in LBSPEC. If the team wants to keep the column, we suggest to change the value to something like ‘Specimen Assessment’ or ‘Specimen Measurement’ or ‘Specimen Observation’.

                                                                v.      Alternate suggestions for how to make values distinct across LB and BS.

   1. Create distinct TEST codes/names
      1. BSTEST: Collected Specimen Weight; Collected Specimen Color;  vs LBTEST: Analyzed Specimen Weight; Analyzed Specimen Color;
      2. Other examples: Color, pH, Consistency, Appearance, LOINC codes with class of SPEC, etc.
      3. Use CAT or SCAT to make the distinction
      4. Retire BS domain – if all of this is context driven, why even have it?

                                                              vi.      If the BS domain was not a domain, the lab team completely agrees that these could belong in Lab. However the presence of the BS domain and the fact that we are trying to stay away from context-driven modeling (ie having the same test appear in two different domains) is making it difficult for the lab team to make the call on whether for this example and context the biospecimen findings belong in BS vs LB. The lab team was unable to come to a decision that these test should be in BS instead.

                                                             vii.      For John Owen - The lab team agrees with the GGG’s decision to model these tests in LB and think that for this case there is a valid reason for context-dependent modeling and would expect that depending on circumstance, to see specimen weight as a test in both LB and BS.

2. METHOD codelist - rows 19-24,  39-42, 48, 54
3. Jordan to review some terms that we think might be clinical classifications.

• Specimen weight issue above – from John Owen email 2018-02-28: Some laboratory tests are adjusted for body weight and some companies had been putting ‘body weight’ into the LB domain so that all parameters for the calculation were in the same domain. Therefore, the lab team felt that it was very important to approve a concept of ‘specimen weight’ to make clear that this test of ‘weight’ in lab was NOT to be used for any body weight measurements. The team also felt that the other specimen-type measurements (like volume, color, clarity, etc.) present in LB would only ever be used to describe the specimen so there wasn’t ever going to be that same confusion. Therefore the team defaulted to using non-specimen specific concepts (concepts where ‘specimen’ is NOT pre-coordinated into the term itself).
• Nate’s SPCCND issue above – from John Owen email 2018-02-28: The lab team does not agree with the commenter’s issue. Fresh vs dried is an attribute of the specimen and the team does not agree it is at all part of the test. The SPCCND variable is available to be used as part of the unique key though we know that Pinnacle 21 does not account for SPCCND as part of the unique key to determine potential duplicates so this would be something that you’d need to explain as an exception in your SDRG. The lab team will not re/consider creating two tests that pre-coordinate the specimen condition into the test value.
• Lab Team rules document – brief overview and let them know we’ll continue discussion for P35.
• METHOD: Row 15 – (CDISC-3106) Erin to research synonym; Row 19, 30, 31, etc.  – Jordan update on these scales.
• Changes to Existing: Rows 8 - See also row 37 in LBTEST-CD.; Rows 10-11 – needed for P34!; Rows 14-38 – needed for P34!