Date

 8:30-9:30am

Attendees

Goals

  • Discuss moving forward with Option 4 of LB vs MB Proposal

Option 4 -  If it is an organism based study, all identification and characterization goes into MB.  If it is a non-organism study, then all screening tests for microorganisms goes into LB. Put viral load in MB.

 

Pros
Cons
Pros
Cons
Requires the least amount of change for sponsorsModeling rules may be difficult for sponsors to implement
 The same test can be found in two different domains which goes against our modeling principles

Example 1 

mb.xpt

Row
MBTESTCD
MBTEST
MBTSTDTL
MBORRES
MBORRESU
1MTBMycobacterium TuberculosisIdentificationPRESENT 
2MTBMycobacterium TuberculosisQuantification, Exact14700CFU/mL
3AFBAcid Fast BacilliIdentificationPRESENT 
4AFBAcid Fast BacilliQuantification, Categorical 3+ 
5HIV1Human Immunodeficiency Virus 1 IdentificationPRESENT 
6HIV1Human Immunodeficiency Virus 1Viral Load5log10 copies/mL
7IDENTIdentification Human Immunodeficiency Virus 1 
8IDENTIdentification Mycobacterium Tuberculosis 

 

Example 2

mb.xpt

Row
MBTESTCD
MBTEST
MBRESTRG
NHOID
MBORRES
MBORRESU
1TRGEXMTargeted ExaminationMycobacterium Tuberculosis PRESENT 
2QUANTEXTQuantification, Exact Mycobacterium Tuberculosis14700CFU/mL
3TRGEXMTargeted ExaminationAcid Fast Bacilli PRESENT 
4QUANTCATQuantification, Categorical Acid Fast Bacilli 3+ 
5TRGEXMTargeted ExaminationHuman Immunodeficiency Virus 1  PRESENT 
6VRLOADViral Load Human Immunodeficiency Virus 1 5log10 copies/mL
7IDENTIdentification  Human Immunodeficiency Virus 1 
8IDENTIdentification  Mycobacterium Tuberculosis 

Discussion items

TimeItemWhoNotes
1hLB vs MBLab Team + Bess + Barrie N
  • Barrie makes executive decision to move forward with option 4. Others options are off the table now. Lab team members agree to develop this further.

  • Option 4 -  If it is an organism based study, all identification and characterization goes into MB.  If it is a non-organism study, then all screening tests for microorganisms goes into LB. Put viral load in MB.

  • Outstanding question: Where to put viral loads? Is the team still okay with putting viral load into MB? Phil - You tend to do viral load counts when the focus of your study is an organism so this seems like a natural fit to put viral load in MB. Other team members agree.

  • Next Steps:

    1. Bess LeRoy to begin a write-up of this modeling strategy with examples. Erin to create a wiki page to store this proposal.
    2. Modeling in MB - Jordan Li to invite Bess to next couple of Microbiology team meetings to get this started.
      1. Granularity of what goes into TEST – currently bug name goes in there as per published CT. Consider use of MICROORG codelist?
      2. Also need to consider use of TSTDTL, NHOID, RESTRG and targeted examination test.
      3. Team needs to consider modeling in previous TAs and how to model into the future.
      4. Need a viral load example
      5. May need some cross team discussions since the data/terminology shouldn’t look too different across LB and MB. Microbiology team may need to start the examples and then present to the full lab team meeting at a future date.
    3. Janet Reich to create an exception to modeling principles document given the decision to do context-driven modeling (hence test codes may be present in two places). She’ll take this to the SDS subteam that is working on these principles.

Action items

  •