Date: June 26, 2020

Time: 12pm ET

Attendees

Planned Agenda Topics

Welcome & Announcements – Zeshan Rajput, MITRE

Zeshan welcomed new members to our Community of Practice. He discussed how CodeX takes a use case driven approach toward the successful implementation of mCODE, which is where the Community of Practice (CoP) comes into play. The CoP is a community that allows organizations to discuss their experiences, with a focus on the implementation of mCODE.

CodeX Updates – Steve Bratt, MITRE

Steve touched on MITRE’s engagement with Point-of-Care Partners to provide insight on how to set up a thriving community. He also spoke about how CodeX is currently focused on cancer efforts and is diverse in its implementations and use cases.

Integrated Trial Matching for Cancer Patients and Providers – Mark Fleury ACS-CAN & Zach Lister, MITRE

Mark began discussing the overarching problem: if you are a patient diagnosed with cancer, and the hospital doesn’t have any clinical trials available, then you are typically told to find clinical trials on your own. On average, only 23% of patients will be able to enroll in on-site clinical trials, thus most trials take a very specific sub-section of the population, since these trial matching services are tough to find and use. The Integrated Trial Matching for Cancer Patients and Providers Project thus proposes, at the request of either the patient or provider, the ability to export a finite set of variables in a de-identified manner using mCODE data elements. This data would then be sent to a third-party matching service, and finally sent back to the EHR for the patient and/or provider to review.

Zach then followed this discussion with a DEMO of the project.

MSK and OHDSI Standardization Challenge – Rimma Belenkaya, Memorial Sloan Kettering Cancer Center belenkar@mskcc.org

Rimma is the head of Memorial Sloan Kettering Cancer Center and co-lead of the OHDSI Oncology Workgroup – a collaborative founded in 2009 that works on supporting a large community of stakeholders to support the observational research of real-world data. OHDSI, at its foundation, is a standardized data representation; the common data model uses clinical data, health system data, drug data, economic data, and a standardized vocabulary known as OMOP. One key difference between OMOP and mCODE is that OMOP’s vocabulary is mostly composed of reused terminology standards – by selecting standards based on comprehensive, high-volume use. The team looks to leverage OMOP – as well as a few other data standards, such as ICD-0 and NAACCR – to create the base-level of the cancer diagnosis language for the model and to work on analytic use cases and research.

The core challenge to address with OHDSI is the problematic way of generating cohorts for clinical studies – only about 20% of the Memorial Sloan Kettering Cancer Center (MSKCC) patient population is on clinical trials. This is very detrimental because it doesn’t create opportunity for researchers to incorporate a larger group of patients. Thus, OHDSI has started working on standardizing and addressing the oncology challenge for representing cancer data to support observational cancer research. One of the team’s goals is to extend the vocabulary, develop ETL for tumor registry and EHR data, and create analytic use cases. Currently, when the team is looking at an individual diagnosis, they can immediately connect to the other diagnostic attributes that apply to this particular diagnosis. The first OMOP Oncology module was released in 2019.

Moving forward, Memorial Sloan Kettering Cancer Center looks to build mCODE into their cancer database which will support cancer research at MSKCC – by standardizing MSKCC’s cancer database using mCODE.

Community Discussion / Questions:

Question: Kaushal Parekh, Roche →  “How do you handle the challenge of some of the more difficult EHR data?”

Answer: Rimma Belenkaya, MSKCC →  “This is why we are using SNOMED as part of the vocabulary. To represent the modules, MSK is using the SNOMED vocabulary and others like ICD-0, NAACCR for this reason.”

Question: Trevor Pugh, Princess Margaret Cancer Center →  “How is genomic data transmitted?”

Answer: Caroline Potteiger, MITRE →  “Tumor marker profiles are what we are using for genomic testing.”

Community of Practice – June Meeting Adjourned