Page tree
Skip to end of metadata
Go to start of metadata

Date

September 4th, 2020

Time

12:30pm - 1:15pm 

Attendees

MITRE - Caroline PotteigerSalim Semy Dan Potter

ACS CAN - Mark Fleury Devon Adams 

BreastCancerTrials.org - Elly Cohen

Cancer Insights - Tarun Kumar Sharon Alford 

Action items

  • Caroline Potteiger  - send Elly CoP meeting slides. 
  • Caroline Potteiger  - check who the SOU should be sent to. 
  • Salim Semy - contact MITRE legal experts about sending patient data. 

Planned Agenda Topics

  • <Hold for hot topics from project team>
    • Was ACS CAN's CodeX SOU sent?
      • It was sent to HL7 but with the wrong suffix.
      • It was re-sent, but Caroline will check on the right person it should have been sent to. 
  • Past Action Items 08/27
    • Mark has not heard back from Flatiron yet. 
  • Engagement Update
    • Trial Matching - Documents
    • UCSF
      • The technical Team has briefly looked into the UCSF data. However, it's MITRE policy to go through a policy review before any data is accessed. This is in progress. 
      • Once MITRE's privacy team gives the okay, Adam can set up the meeting with the consultation service.
      • Can walk through a data use agreement offline. 
      • UCSF has one de-identified database and one identified database with clinical notes. 
        • Need to understand how much information is in the de-identified database.  
  • Phase 1 Update
    • Salim and Caroline met with MITRE IRB on Wednesday. Need three components to move forward with IRB review: 
      • More robust study protocol, defining specifics around data flow, security, patient recruitment, consent, etc. 
      • Review and approval by MITRE privacy team 
      • Agreement in place between MITRE and Cancer Insights 
    • This phase is focused on patient recruitment and patient initiation. 
    • The end-to-end workflow is the priority. 
    • Right now, the 20 trials for testing include 13 trials for metastatic breast cancer and 8 trials for newly diagnosed breast cancer. 
      • Newly diagnosed patients probably aren't interested in joining clinical trials. 
    • This IRB process may take a while - need to prioritize efforts.
  • Phase 1B Pilot with UT Southwestern 
    • They will be reviewing the protocol internally this week and we are following up next week. 
    • BreastCancerTrials.org willing to do a manual match? 
      • What manual means: 
        • UTSW sends over patient data for the minimal eligibility criteria to BreastCancerTrials.org and runs the data through the system. 
        • Then, a further screen is done based on the full patient record to see which trials are actual matches. 
      • BreastCancerTrials.org lists over 600 trials.
      • We could further scope the cancer sub-type to limit the amount of potential patients and trials.
      • Next week, we will get a better sense from UTSW on if/how they can send full patient records.
        • If it's only one matching service, what agreements need to happen.
        • They could also enter the patient data into BreastCancerTrials.org and do the next screen themselves. 
      • As this planning progresses, a call with UTSW and Elly could be helpful. 
    • An IRB will be needed with any health system. IRBs need a BAA and HIPAA environment.
    • If these legal agreements are taking took long, institutions can do internal searches and just send us the results. 
    • BreastCancerTrials.org doesn't work with health centers - only the patient community. 
  • Public Call 
    • Draft slides found here - Trial Matching Supporting Materials
    • Practice run next week during our team sync. Will anyone be out?  
    • Proposed agenda with talking points: 
      • Welcome - 5 minutes 
        • Caroline to welcome everyone and introduce the team  
      • Introduction to CodeX - Steve, 5 minutes 
        • Steve to give a brief overview of mCODE/CodeX 
      • Introduction to project - Mark, 10 minutes
        • Birth of the project idea, diagram overview, minimal eligibility criteria, multi phased plan 
      • Demonstration - Zach, 5 minute video + 5 minute questions
        • 5 minute video 
        • Leave extra time for questions
      • Details on Phase 1 - Salim and Tarun, 10 minutes
        • Salim to present on Phase 1
        • Tarun to present on Phase 1A
        • Salim to present on Phase 1B 
      • Call to Action - Elly, 5 minutes 
        • Why it was important to join this project
        • How stakeholder groups can add value (slide 21)  
      • Discussion/QA - 20 minutes 
  • How much signal to noise is okay? We need to start thinking about this. 
    • This could vary with disease type. 
    • If noise ratio is too high, can we add one more variable? Can we slowly peel off the criteria to see what happens? 
    • A statistician is giving us feedback next week. 
    • We need to avoid false negatives. However, we need to separate false negatives due to the criteria and false negatives compared to the matching service algorithm. 
  • Since sending patient data is important for the end goal of this project, we need to do an analysis of the law and what organizations are willing to do. 
    • People aren't as concerned about sending de-identified data as they are about sending identified data. 
    • If we have documentation or a white paper describing the legality, it might open up organizations to sending data.
    • ACS CAN has a legal team that could look into this as well. A few opinions will be valuable. 


  • No labels