The Genomic Implication profile is an abstract profile for therapeutic and diagnostic implications and molecular consequences. Implications are relevant information which are derived from the presence of a haplotype, genotype, variant or molecular biomarker. This portion of the implementation guide relies on the content in the General Genomic Reporting and Variant Reporting portions of this implementation guide.

At the heart of many genomic reports are findings (variants, haplotypes, genotypes, biomarkers) and annotations derived from those findings (diagnostic implications, therapeutic implications, predicted molecular consequences), as shown in the following figure:

This profile defines a common set of elements for other Implications:

• derivedFrom is a required field that provides an explicit reference from an implication to the variant(s) / haplotype(s) / genotype(s) / molecular biomarker(s) from which the implication is derived.
• workflow-relatedArtifact (base extension) supports conveying references to citations, supporting documentation and other information relevant to the asserted implication, such as documentation which applies to the full implication Observation
• workflow-relatedArtifactComponent (component level extension) supports conveying references to citations, supporting documentation and other information specific to the component's code and value, such as documentation supporting a specific clinical significance or evidence level value
• evidence-level (component) indicates the strength of the evidence behind the asserted implication
• clinical-significance (component) indicates the clinical impact of an implication

evidence-level versus clinical-significance

evidence-level: The amount of observed support for the association between an implication and a variant / haplotype / genotype / biomarker. There are dozens if not hundreds of evidence-level value sets, generally providing an ordinal range of codes from low evidence (e.g. 'predicted', 'case report', '1-star') to high evidence (e.g. 'professional society guideline', 'expert panel consensus', '1A').

clinical-significance: The clinical impact of an implication on a person's health. There are dozens if not hundreds of clinical-significance value sets, generally providing an ordinal range of codes from low significance (e.g. 'benign') to high significance (e.g. 'pathogenic', 'oncogenic','predictive of drug response').

While evidence and significance are orthogonal concepts, some codes conflate the two (e.g. 'expert panel consensus of strong efficacy', 'strong evidence refutes the clinical utility of this drug', 'case report evidence of pathogenicity'). In general, using discrete evidence and significance codes allow for a more precise characterization of an implication. Where an implementation is using a conflated code, we recommend communicating that code in the evidence-level field.

Examples of other code sets for evidence-level and clinical-significance:

• ACMG_Level of evidence
• Oncokb levels
• US FDA levels
• AMP/ASCO/CAP levels
• Classification on ClinVar submitted records
• Bioportal clinical variant interpretation
• CIViC Evidence
• CIViC Significance

Differentiating AND vs OR

There are several cases where it is necessary to differentiate 'AND' conditions (e.g. both drug X AND drug Y, in combination, are indicated in the presence of a variant) vs. 'OR' conditions (e.g., either drug X OR drug Y are indicated in the presence of a variant). This situation is not unique to genomic implications, and arises elsewhere within FHIR (e.g. FHIR Search) and outside of FHIR (e.g., ClinVar submission API condition set). To be consistent with other precedents:

Where implications have fields with cardinality > 1, the inclusion of multiple values within a field SHALL indicate an 'AND' condition. An 'OR' condition SHALL be represented by multiple observation instances.

Implication fields affected by this guidance include:

Molecular consequences: derivedFrom, evidence-level, feature-consequence.

Therapeutic implications: derivedFrom, evidence-level, therapeutic-implication, phenotypic-treatment-context, medication-assessed, therapy-assessed.

Diagnostic implications: derivedFrom, evidence-level, predicted-phenotype, genomic-risk-assessment.

Here are examples of these rules with DiagnosticImplication (but the principles would apply to any Implication):

Use Case	Description	Visual
Variant X is pathogenic for Disease Y AND Disease Z (in combination)	One DiagnosticImplication with two predicted-phenotype instances for Disease Y and Disease Z that is derivedFrom one Variant for VariantX
Variant X might be pathogenic for Disease Y alone, or Disease Z alone	Two DiagnosticImplications, one with predicted-phenotype for DiseaseY, second with predicted-phenotype for DiseaseZ, and both are derivedFrom the same Variant for VariantX
Variant X AND Variant Y (in combination) are pathogenic for Disease Z	One DiagnosticImplications with predicted-phenotype for DiseaseZ that is derivedFrom the two Variants (VariantX and VariantY)
Variant X OR Variant Y might be pathogenic for Disease Z	Two DiagnosticImplications, each with predicted-phenotype for DiseaseZ, one that is derivedFrom a Variant for VariantX, the other that is derivedFrom a Variant for VariantY