This page is part of the HL7 FHIR Implementation Guide: minimal Common Oncology Data Elements (mCODE) Release 1 - US Realm | STU1 (v4.0.0-ballot: STU4 Ballot 1) based on FHIR (HL7® FHIR® Standard) R4. The current version which supersedes this version is 3.0.0. For a full list of available versions, see the Directory of published versions
The mCODE Disease Characterization group includes data elements specific to the diagnosis and staging of cancer. This includes:
The cancer diagnosis combines the type, site, and certain characteristics of the cancer. Depending on the EHR and provider organization, different code systems may be used, such as:
Because the use of these code systems vary in different institutions, mCODE supports all three. Two elements and one extension of the FHIR Condition Resource are involved with coding the cancer diagnosis: Condition.code
, Condition.bodySite
, and the HistologyMorphologyBehavior extension. How these attributes are used, depending on the code system, is captured in the table below:
Encoding | Code | Histology Morphology Behavior Extension | Body Site |
---|---|---|---|
SNOMED Encoded | Any SNOMED CT code in the Primary Cancer Disorder Value Set | Any SNOMED CT code in the Histology Morphology Behavior Value Set | Any descendant of 123037004 “Body structure” |
ICD-10-CM Encoded | Any ICD-10-CM primary code (precoodinated) | omit | optional; must be consistent with primary code if provided but may contain more detail |
ICD-O-3 Encoded | SNOMED CT code 55342001 “Neoplastic disease (disorder)”, 363346000 “Malignant neoplastic disease (disorder)”, or 20376005 “Benign neoplastic disease (disorder)” |
Full ICD-O-3 morphology code* | Any ICD-O-3 Topology Code |
*The ICD-O-3 morphology code should have the form HHHH/B G
where HHHH is the 4-digit histology code, B is the 1-digit behavior code (i.e., /1, /2, or /3 suffix for primary cancers, and /6 suffix for secondary cancers), and G is the grade or equivalent in leukemias and lymphomas (1 digit). For more information, see ICD-O Third Edition, Table 8.
Implementers should reference the PrimaryCancerCondition and SecondaryCancerCondition profiles for further details on the use of these terminologies and associated value sets.
Histologic behavior, type, and grade can also be reported in the HistologicBehaviorAndType and HistologicGrade profiles. TheTumorMorphology profile references those profiles in a report.
On initial diagnosis, the Condition.clinicalStatus
element will be active
. Subsequent changes to the disease status should be recorded by updating the clinicalStatus
element. The permitted values are active, recurrence, relapse, inactive, remission, resolved. Recurrence and relapse are often used interchangeably in the context of cancer. The resource’s history can be accessed to see the history of the status value.
Note that there is another resource profile, the CancerDiseaseStatus, that is used to record the patient’s condition on an encounter-by-encounter basis, and uses values such as improved, stable, worsened, as well as full and partial remission. When the value of CancerDiseaseStatus indicates remission, the Condition.clinicalStatus
should be updated to reflect that finding.
Body locations in FHIR are typically represented using a single code. However, a single code is often insufficient to describe where a tumor is located, where a surgery is targeted, or where a radiation treatment is focused. When a single code is insufficient, FHIR recommends using a BodyStructure resource. This is appropriate when the BodyStructure is tracked over time, for example, in the case of Tumor. But generally, it is better to describe a body location without using an additional resource.
mCODE has adopted an approach that allows the user to add additional code or codes to further define the body site, without the need to create an independent resource. This takes the form of the LateralityQualifier and BodyLocationQualifier extensions. These extensions can be used to specify laterality, directionality, and plane.
In mCODE, staging and risk assessment information has three components:
The stage or risk assessment value (3) is always reported. To interpret the meaning of the stage or risk assessment value, the staging or risk assessment system or staging or risk assessment method must be known. Depending on the code used, the stage or risk assessment type (1) can identify the staging or risk assessment system, the kind of stage or risk assessment reported, and factors such as the timing (e.g., at diagnosis or posttherapy) and the type of evidence (e.g., clincal or pathologic). If the stage or risk assessment type does not imply the staging or risk assessment system, the staging or risk assessment system is reported separately (2).
A degree of redundancy may exist between these three elements. For example:
The data sender must assure that the values in these three fields are self-consistent.
Staging or risk assessment information should be provided as Observation resource(s) conforming to the CancerStage or CancerRiskAssessment profile or a constrained version of those profiles. CancerStage and CancerRiskAssessment are parent profiles that should be used only if a more specific profile corresponding to a particular staging system or risk assessment is unavailable. If a patient has been staged or assessed more than once, there will be multiple CancerStage or RiskAssessment observations.
In the CancerStage and CancerRiskAssessment profile and its descendants, the following elements are used to describe a stage or classification:
Stage/Risk Assessment Information | FHIR Element | Description |
---|---|---|
Type | Observation.code |
LOINC or SNOMED term that describes the specific type of stage or risk assessment being reported, for example, a TNM stage group or International Federation of Gynecology and Obstetrics (FIGO) ovarian tumor stage. In terms of the SNOMED CT hierarchy, these are terms of type Observable Entity. |
System | Observation.method |
The staging or risk assessment system, method, or protocol used, for example, AJCC Version 8 or the International Neuroblastoma Staging System. In the SNOMED CT hierarchy, these are terms in the Staging and Scales hierarchy, specifically, terms descending from Tumor Staging that represent staging systems. Observation.method is not required if the staging system is implicit in Observation.code . |
Value | Observation.valueCodeableConcept |
Contains the actual stage or category determined for the cancer. In terms of SNOMED CT, these are terms from the Qualifier Value and Finding hierarchies (some staging values appear, perhaps erroneously, in the Tumor Staging hierarchy). |
Cancer | Observation.focus |
A reference to the cancer condition being staged or assessed. |
Prognostic Factors | Observation.derivedFrom |
A reference to Observations contributing to the stage or risk assessment. |
A reference to the CancerStage observation should be given in the PrimaryCancerCondition’s Condition.stage.assessment
element. If staging has been repeated for a patient, the reference in PrimaryCancerCondition should point to the most recent staging information.
TNM staging is used for many types of solid-tumor cancers. The TNMStageGroup profile is a specialization of CancerStage dedicated to AJCC TNM staging. This profile contains the stage group in Observation.valueCodeableConcept
and provides optional references in Observation.hasMember
to additional resources representing the T, N, and M categories. The Observation.code
element value in TNMStageGroup is used to distinguish the type of staging, e.g., clinical or pathologic. For other types staging (e.g., retreatment (r) or autopsy (a)), a code indicating “other” staging type is used.
Clinical applications vary in their representation of T, N, and M staging category values, falling into one of two naming conventions:
mCODE strongly recommends that the implementers align with AJCC’s convention of representing the staging category value including the classification prefix. This code convention is aligned with the AJCC’s digital data and clearly distinguishes the staging classification as clinical (c), pathologic (p), neoadjuvant (yc or yp), restage (r), or autopsy (a) without having to retrieve further context from the model. The selected prefix SHALL be consistent with the Observation.code
value.
Several widely-used terminologies in the cancer domain, including ICD-O-3 and AJCC staging, are proprietary and cannot be reproduced in this guide. SNOMED-CT has reached an agreement with AJCC to create SNOMED codes that correspond to AJCC stages. mCODE uses these SNOMED codes where applicable, but licensed sites may continue to use AJCC codes and still be in conformance with mCODE. Under the Fair Use doctrine, examples illustrating mCODE’s representation of cancer diagnoses may use the more familiar AJCC staging values for the purposes of implementation guidance to FHIR developers.
Based on discussions with clinical experts, several common non-TNM staging systems and risk assessments were identified and added as profiles in mCODE. Not all cancer types are staged or assessed with a TNM-based staging system, including hematologic cancers like leukemias, multiple myeloma, lymphomas, and some solid tumors, for example, gynecologic tumors are staged using the FIGO (International Federation of Gynecology and Obstetrics) staging system.
Prognostic factors related to the cancer stage group or risk assessment can be specified with the Observation.derivedFrom
element. For example, a hemoglobin lab result which was evaluated in the staging of chronic lymphocytic leukemia (CLL) using the Binet staging system can be referenced under Observation.derivedFrom
element. This example of Binet staging illustrates how this could be represented.
Terminology. SNOMED CT does not offer codes for every staging system. When SNOMED codes are unavailable, mCODE falls back on codes from the NCI Thesaurus (NCIT). From an implementation perspective, managing two code systems is difficult when there is no consistency of which code system is used for what purpose. We are actively working with SNOMED International to create more consistent semantic approach to coding stages, by adding concepts required by non-TNM staging systems. The current approach of mixing SNOMED and NCIT depending on the cancer type should be regarded as temporary.
To keep staging information together with cancer diagnosis, staging values and types can be included in the PrimaryCancerCondition. A stage reported in PrimaryCancerCondition is interpreted as the stage at the time of diagnosis.
In PrimaryCancerCondition, only the stage type and stage value can be reported, as follows:
Stage Information | FHIR Element | Description |
---|---|---|
Stage Type | Condition.stage.type |
See “Reporting Staging Information” above. |
Stage Value | Condition.stage.summary |
See “Reporting Staging Information” above. |
Representing stage information in this way, i.e., without an explicit staging system, is valid only if the staging system is implicit in the stage type.
Tumor markers are key prognostic factors in calculating cancer staging, identifying treatment options, and monitoring progression of disease. For example, an abnormal increase in prostate-specific antigen (PSA) levels is a prognostic factor for prostate cancer. Other tumor markers include estrogen receptor (ER) status, progesterone receptor (PR) status, carcinoembryonic antigen (CEA) levels, among others. mCODE distinguishes tumor marker tests from sequencing-based genomic tests measured at the DNA, RNA, or chromosomal level. The latter are addressed in the Genomics section.
mCODE includes single FHIR profile, TumorMarkerTest, for all labs involving serum and tissue-based tumor markers. This is less than ideal, since without specifying units of measure or answer sets on a per-test basis, reporting could vary. However, given the large number of tumor marker tests, creating individual profiles was judged impractical.