Quality Assurance and IV&V
Strategy for ensuring the quality, fit-for-purpose, and usefulness of available knowledge.
Detailed Description
The clinical interoperability strategy will result in two general categories of data: 1) Data about tests as a class, e.g. the IVD name, mapped code, and device that produces it; and 2) Individual testing data, i.e. the actual results being produced for use. A quality assurance strategy should aim to assess the quality of both sorts of data.
“Class” data quality assurance:
Ensuring LIDR resources are generated according to defined practices
Assuring IVDs are uniquely identifiable
Assuring UDIs are generated according to defined practices
Allowing laboratories to assess whether their testing menu is conformant to LIDR resources
Allowing third parties to assess whether a laboratory’s menu is conformant to LIDR resources
“Instance” data quality assurance:
Maintaining a repository of traceability/harmonization information to establish comparability of results both within and between IVDs
Third party verification that individual testing data have mapping and harmonization elements required for interoperability
Third party verification that individual tests do not have unexpected variation by performing laboratory
Third party performance monitoring of devices and IVD kits
Use of aggregated testing data to identify opportunities for harmonization not currently described in harmonization procedures
Use of aggregated testing data to support Real World Evidence (RWE) generation to support post-market determinations of quality
Supporting Tactics
Tactic | Impact |
|---|---|
LIDR/UDI Vendor Authoring Tools | Make it easy for vendors to assign codes quickly and well. |
Tools for Automated lab catalog and data dictionary maintenance | Laboratories/vendors would not need to manually look up necessary LIDR or semantic mappings--searching within the LIS via API calls to necessary repositories would ease burden and increase reliability |
Reference Standard/Harmonization Hub | Contains information on what analytes have a traceable standard or harmonization procedure. |
Laboratory testing QC data repository | By monitoring individual testing data, a QC data hub would potentially 1) identify testing that can safely be treated as similar 2) identify incorrectly mapped tests |
Grouping determination re: standardization/harmonization | Primarily a governance problem: will there be a central location telling people if tests from two IVDs can be trended together, or will this be left to local judgment? |
Real World Evidence data hub | National database of testing data allowing performance monitoring of IVD and device-related issues. With linkages, could be used in evidence generation. |
Supporting Use Cases
Use Case |
|---|
An IVD vendor creates a new IVD and wants to include new information in the LIDR. Using a freely available authoring tool, the vendor is able to generate mappings and necessary IDs according to a pre-specified manner. |
A laboratory wants to update its mapped encodings for laboratory tests so that they are accurate with the new recommendations. The LIS has the ability to query outside systems through an API. A technologist is able to directly search within the LIS for mapping information appropriate to that IVD. |
A laboratory submits some laboratory results (e.g. proficiency testing results) to a trusted repository. The repository is able to query the LIDR and other resources for discrepancies in order to determine if the results are encoded correctly. |
A laboratory submits some laboratory results (e.g. proficiency testing results) to a trusted repository. The repository compares the lab’s results to prior testing from this lab and other laboratories and determines that the most recent reagent lot version of the IVD appears to produce aberrant results. The vendor and other laboratories are alerted. |
A laboratory submits some laboratory results (e.g. proficiency testing results) to a trusted repository. The repository compares the lab’s results to those of other laboratories and determines that a recently published harmonization procedure is working appropriately. |
A vendor markets an IVD for detection of SARS-CoV-2, but wants to determine if the sensitivity is similar to that of other IVDs for this purpose. They approach a pre-existing registry of patients who were all tested under similar circumstances, but the registry does not have detailed vendor-level testing data. Using a privacy-preserving linkage to an IVD data hub containing testing results, the registry is able to augment its patient-level data and provide the vendor with analysis showing that its IVD appears to be less sensitive than similar products when used in saliva testing. |
Key Initiatives & Action Items
Define necessary authoring tools
Define protocol for LIDR and semantic resource knowledge creation and maintenance
Engage pre-existing traceability community and its efforts
Define scope of QA and RWE data hubs, and whether they are in scope at all