2021-08-25
- riki.merrick
Agenda
Update on strategic plan document (Riki)
Not discussed - so please feel free to review: https://aphlinformatics.atlassian.net/wiki/download/attachments/1167425547/LIVD FILE EXPANSION COMMITTEE TEXT for Strategic Plan.docx?api=v2
Review diagrams
A picture is worth a 1000 words - trying to give an overview of SHIELD
high level overview of SHIELD (Scott)
Comments: Need to add arrow fromONC to Lab - share requirements - to professional societies might be helpful here, too
under domain experts include professional societies
each of the sides should be considered going through iterative feedback loops
once we have a sustainable product then we get a higher level feedback loop - and at this point we need to get the regulatory agencies involved
Need to think about how the data is created - harmonized results for lab tests would be the goal here
at gets comminucated
data aggregation needs to address harmonization as well, and even when there is only a single code for aggregation, we will still come up with incorrect conclusions - we stil need more than just one code as the basis
Must be sure the data provenance is covered in the process
test data life cycle (Riki)
https://aphlinformatics.atlassian.net/wiki/download/attachments/1167425547/SHIELD_Flows.pptx?api=v2
Comments:
from the FDA box, do you mean that LOINC has to be mandate as is UDI and then sent to the GUDID (as is UDO)? - answer - it may not be a requiremetn for FDA submission, but FDA assignig a UDI should be part of the process, so it can be submitted to LIDR - whetehr or not the LIDR file content is part of the FDA review or not also is open for debate, but it should be produced at the same time, so that when the test is availabel on the market, so is the LIDR content.
How can we make this go faster?
Can define the most important test that need to be interoperable in parallel to creating and building LIDR
work at the same time on harmonization of these high impact tests, if possible / Need to be sure we can calibrate those results -can that be done in parallel?
Should we consider the differnet states of results and limit exchange to final?
in the LIS to EHR-s transactions should support ALL statuses to ensure timely communicaiton for patient care
int he submission to the RWE data hub consider ONLY final and corrected, but need to make sure that includes enough metadata for data provenance
Not discussed on the call, but Scott shared his version (which has the same look and feel as the overview, so we may merge the content of the two, depending on the final outcome of the overview diagram): https://aphlinformatics.atlassian.net/wiki/download/attachments/1167425547/SHIELD_LabLevel_Process.pdf?api=v2
LIVD File Normalization to Third Normal Form (Jerry)
Jerry has been working on transforming the working dB into the third normal form, for distribution and as input into the requirements gathering
We have a separate spreadsheet the we don't publish - this includes all the specific primers and controls for each test case, it also includes identifiers for the package insert, the EUA etc. - the normalized form dB will be based on that more detailed
Will it be the final distribution for the LIVD file?
We don't know at this time
This may be used as input for building LIDR, or it may be used for the minimal viable product - this is TBD
Update on conversation with Jack Bloom (Ana)
Need entity that is responsible for distirbution of all the important data elements in a stnadardized format that all IVDs for the same type of test so results can be comingled to have true equivalency
Comments:
AACC has been involed in harmonizing lab results
ISO standards are built around harmonizing lab results
Has been driven by national clinical society guidelines
Need involvement of manufacturers in this effort
optimizing reuslt from submission by differnt IVD manufacturer studies
Should work on this in parallel and not forget about this
Recording link
https://drive.google.com/file/d/1rSgVZVoyusJNv86IoZ6ASkS5is41LX4w/view?usp=sharing