Technical Confirmation Process

From QIBA Wiki
Revision as of 23:03, 6 July 2021 by Kevino (talk | contribs)
Jump to navigation Jump to search

The Technical Confirmation Stage confirms it is practical for sites to perform the technical requirements to conform to the Profile.

Users in the field (clinical oncologists, imaging scientists, pharma, CRO’s, imaging vendors, software providers, etc.) attempt to conform to the Profile and provide feedback on the practicality of the requirements (and the usability of the document).

Whereas the preceding Public Comment Stage collected feedback based on people reading the Profile document itself, this stage seeks feedback on how well it works in a pilot implementation. As such, comments are solicited specifically from those that have used it and only those that have used it.

Considerations:

  • Profile authors have written requirements they believe are within the range of current reasonable practice in clinical use.
  • Sites are asked to consider, as they generate the biomarker according to the Profile, whether they would find the requirements practical in routine work.
  • Sites do not need to commit to using the Profile for all routine work going forward (although it would certainly be positive feedback if they chose to).
  • Sites are strongly encouraged to try to meet all the requirements
  • If the site is not comfortable using the procedures for patient scans, they can scan a phantom and note in their feedback that they used a phantom, and what details of the procedure made them uncomfortable.
  • If sites are uncomfortable meeting certain requirements with patients, that would make it hard to deploy the Profile so the Biomarker Committee may need to consider revising that requirement or better explaining/justifying it for sites.
  • Try to follow all requirements, even if scanning a phantom.
  • E.g. if the Profile has required breathing instructions, giving them to the phantom may help a site realize that the wording is incompatible with their preferred practice, or the text is poorly written, etc.
  • If the site is unable to conform to any requirements, comments on the nature of their difficulty or their reason for not conforming are very helpful to the Biomarker Committee.


The following Procedures list what is done (by whom).

Planning & Recruitment

  • Identify sites/people that will participate in the field test (Field Test Leader)
    • At least 3 sites are needed to get reasonable confirmation of results
    • Often Committee members may have personal contacts they can explore
      • "External" sites should be recruited to bring "fresh eyes" to better assess the clarity of the profile and bring different assumptions about routine practice for this biomarker.
      • Sites where Biomarker Committee members work may be easier to recruit since there is an existing champion, but they should make a point of engaging other fresh eyes (techs, physicists, radiologists, etc) and possibly stepping back when it comes time to perform the actual field test.
      • It can be helpful to have a mix of large academic centers, large hospitals, and smaller hospitals or imaging centers.
      • International feedback is desirable due to variations in practice and language; and QIBA Profiles are intended to be internationally applicable.
        • <TODO: Add links to contacts for Europe, Japan, US, etc.>
      • Core Labs can be effective field testers since they have experience with a variety of site practices and measurement needs
    • It may also be useful to call for volunteer sites on the QIBA mailing lists (RSNA Staff)
    • If a profile covers multiple organs/cases, the goal is to test all of them.
      • If each site tests all organs/cases, test coverage is simplest; however, it is acceptable for sites to test subsets as long as everything gets covered.
  • Prepare field test materials/documents (RSNA Staff/Biomarker Committee)
    • Cover Letter
      • Many times a meeting is arranged between members of the field test site and members of the Biomarker Committee to describe the purpose of the field test and answer any questions the site may have.
        • Consider recording the questions/answers since they may point to useful clarifications in the Profile
        • If a meeting is not possible, the information may be conveyed in a Cover Letter.
    • Site Details
    • Profile Requirements
      • Generally, the Profile Checklists are the core document.
      • The full Profile should also be made available as some sites will find it useful to have the context provided in the Overview and the explanations/clarifications in the Discussion sections.
    • Feedback Forms
      • One effective approach is to add two columns to the conformance checklists, to indicate for each requirement whether the site conformed to the requirement, and whether they found the requirement practical for routine use.
      • Additional General comments can be collected on a form similar to Public Comment.

Collecting & Processing Feedback

  • Provide field test materials (see above) to each site (RSNA Staff)
  • Email feedback to Field Test Leader (Commenters)
  • Collate all Feedback comments into a Feedback Resolution spreadsheet (RSNA Staff/Field Test Leader)
    • See the Instructions sheet in the spreadsheet for details.
  • Assess test coverage (Biomarker Committee)
    • Were all requirements tested. If coverage is incomplete <including organ coverage>, consider dropping the requirement or document the rationale for finding it technically confirmed in the absence of testing. <some sites may have done less than scheduled/intended><might not have needed coils on hand><BC notes lack of coverage and rationale for how to proceed><are there substantive practical issues, if not OK for Stage 3. Are there substantive performance issues, comes up in Stage 4>
  • Review comments, making changes to the document as needed (Biomarker Committee)
  • Record comment resolutions in spreadsheet


(Cmte can decide if the full procedure has been "adequately" tested). Sites not doing requirement X is key feedback.

  • (Doing clinical patients exposes nuances,
  • <Committee may be faced with a response where none of the sites were comfortable executing the profile on patients and responded based on reading. In this case, the committee has done two public comments and no technical confirmation and has sites saying they are not comfortable with the profile. Should the committee advance the profile in this case? That would be pretty negative feedback>
  • <Profiles are generally written not to be unusual procedures><NOTE Didn't conform isn't a killer, it's the information we are trying to get. Might be better procedures, or might be superfluous>
  • <"Not routine, but can" also useful feedback - can do it but we really have to make the case for the value>


  • Vote on next step for the document (Biomarker Committee), either: <clarify with new wording>
    • refer the document to the Modality Committee for Technically Confirmed publication
    • conduct another round of field test on revised document
    • place the document into inactive status


  • Post Feedback Resolution spreadsheet, revised document to the Wiki (Field Test Leader/RSNA Staff)
  • If profile requirements didn't change significantly, test sites may be able to claim conformance.

<When the TC 2020 version comes out, and a site tried the Consensus 2020 but only missed one requirement. Then the BC relaxed that requirement in the TC2020 (or 2021 explain ) draft, it would be nice if there was "change tracking" so they could see that now they would pass. Could have a "change tracked" document between each draft (but just the preceding, not all the combos, you have to step through the changes of history) Practically, we can point them to the newest, and maybe provide one "generation" of change track>

Retrieved from "https://qibawiki.rsna.org/index.php?title=Technical_Confirmation_Process&oldid=16998"