mCODE Project Scope Statement (PSS)

Describe the project; include what is expected to be accomplished/delivered along with specified features and functions. Include whether the deliverable(s) are universal, realm specific or applicable to various realms. Be sure to spell out all acronyms as these are carried forward to the NIB (Notice of Intent to Ballot) for ballot announcements.

This project will develop a common set of standardized data elements relevant to many types of cancer. It is intended to facilitate collection of "real world data" from clinical encounters with higher uniformity and quality, independent of the provider or EHR vendor. Data structures accurately capturing patient characteristics, disease characteristics, treatments, and outcomes that matter to patients and clinicians would rapidly accelerate progress toward using current therapies optimally and potentially, new clinical guidelines. The resulting Minimal Common Oncology Data Elements (mCODE)-based data sets, with data from the clinical care settings, will not require extensive curation, and will be shareable with other mCODE-based data sets without reformatting. A core team of oncologists will guide the selection and prioritization of content, and the results (subject to HL7 approval) will be delivered as FHIR profiles in an Implementation Guide (IG). The results will be US Realm-specific, but the lessons learned, and many of the actual profiles, could have universal applicability. The intent is to base mCODE on a common base logical model determined and approved by CIMI.

This information is required by ANSI for all ballots. Briefly explain the reason behind the need for this project. This may be related to legislative requirements, industry need, or similar justifications.
Tremendous progress could be made if oncology data gathered in clinical encounters could be captured in standard form. Currently, only 3-5% of cancer patients participate in randomized clinical trials (RCTs), and very little medical knowledge is gained from the other 95-97%. A significant part of the problem with real-world data is caused by the variability in the ways that EHRs capture clinical data, with much information going into clinical notes. Attempts to extract structured information clinical notes using natural language processing (NLP) have not been able to deliver high quality, computable data at scale. The only effective method of involves human data curation, which is expensive and unscalable. Although large repositories of cancer data already exist, the usefulness of this data is hindered by the variability and lack of structure in the data.

The interested parties, led by American Society of Clinical Oncologists (ASCO), believes the best starting point to begin to address this problem is to identify a small set of data elements that should be collected for all cancer cases, an effort they have named "mCODE", for Minimal Common Oncology Data Elements. Of course, mCODE will not be effective unless clinicians are willing to routinely apply it during patient care. Data collection is therefore an essential feature of mCODE use cases, for which the design goal is to minimize disruption of normal workflow and maximize incentives for adoption by the clinicians. MITRE and ASCO have teamed up to develop an mCODE-based Fast Healthcare Interoperability Resources (FHIR) implementation and Substitutable Medical Applications and Reusable Technologies (SMART)-on-FHIR application to extract mCODE data in computable formats. Intermountain Healthcare, as a participating CancerLinQ data repository and leader in quality cancer care, will operate test sites for capturing the required mCODE data and for utilizing the tailored reports during clinical care. The incentive driving initial mCODE adoption is its use in therapeutic decision-making and comparative effectiveness analysis, by providing cancer patients and oncologists with a cohort of (anonymized) similar patients, with their treatments an outcomes, in real-time. The ultimate goal is to use real world data to accelerate progress toward using current therapies optimally and potentially developing new clinical guidelines, particularly for patients currently underserved by RCTs, such as elderly and disadvantaged patients, patients with multiple co-morbidities, for rare tumors, and to assess outcomes in patients who receive multiple lines of treatment.

Describe any external schedules or calendars which may not be known outside of the project team that are driving target dates for this project..

None

What common name does your group use to refer to the product(s) produced?  What alternative names, aliases and keywords does your group use to refer to the product(s) that will be produced?  Some examples: C-CDA, LRI, eDOS.  

mCODE (aka Minimal Common Oncology Data Elements)

If your project creates a Post-Release 1 version; indicate the name of the prior product and if it is supplanting, replacing or coexisting with a previous release:

n/a

Enter any dependencies or the name & Project Insight ID of project(s) that this project is dependent upon to achieve its objectives:
 

Projects and their Project Insight IDs can be found via http://www.hl7.org/special/Committees/projman/searchableProjectIndex.cfm?ref=common

HL7 Clinical Genomics Domain Information Model(s) (Project ID 1050)

X

X

X

x

x

x

x

x

If you checked New Product Definition or New Product Family, please define below:

For FHIR IGs and FHIR Profiles, what product version(s) will the profiles apply to?

DSTU 2

x

x

x

 X

U.S.

Enter “U.S.” or name of HL7 affiliate(s) here.  Provide explanation/justification of realm selection. For projects producing deliverables applicable to multiple realms, document those details here. 

For Investigative projects, indicate if the project is planned to be Realm Specific or Universal, if known. Work Groups are encouraged designating project a Universal project initially, and discover which Realms can contribute to the work effort during the discovery phase of the project. Note: This status is subject to change during the investigative process.

x

x

x  

x

 x 

 x 

 x

x

 x

 x

x

x

x
  

Other:  Indicate other stakeholders, vendors or providers not listed above.