simplify {Rdiagnosislist}

Retrieves closest single ancestor within a given set of SNOMED CT concepts

Description

Returns a vector of SNOMED CT concept IDs for an ancestor of each concept that is within a second list. If multiple ancestors are included in the second list, the concept is not simplified (i.e. the original concept ID is returned). This functionality can be used to translate concepts into simpler forms for display, e.g. 'Heart failure' instead of 'Heart failure with reduced ejection fraction'.

Usage

simplify(
  conceptIds,
  ancestorIds,
  SNOMED = getSNOMED(),
  tables = c("RELATIONSHIP", "STATEDRELATIONSHIP")
)

Arguments

conceptIds

character or integer64 vector of SNOMED concept IDs for concepts for which an ancestor is sought

ancestorIds

character or integer64 vector of SNOMED concept IDs for possible ancestors

SNOMED

environment containing a SNOMED dictionary

tables

character vector of relationship tables to use

Details

This function is intended for use with active SNOMED CT concepts only.

Value

a data.table with the following columns: originalId (integer64) = original conceptId, ancestorId (integer64) = closest single ancestor, or original concept ID if no ancestor is included among ancestorIds

Examples

SNOMED <- sampleSNOMED()

original_terms <- c('Systolic heart failure', 'Is a',
  'Heart failure with reduced ejection fraction',
  'Acute kidney injury due to circulatory failure (disorder)')
# Note in this example 'Is a' has no parents in ancestors,
# and acute kidney failure has two parents in ancestors
# so neither of the parents will be chosen.
# Also test out inclusion of duplicate concepts.

ancestors <- simplify(c(as.SNOMEDconcept(original_terms),
  as.SNOMEDconcept(original_terms)[3:4]),
  as.SNOMEDconcept(c('Heart failure', 'Acute heart failure',
  'Cardiorenal syndrome (disorder)')))
print(cbind(original_terms, description(ancestors$ancestorId)$term))