Have option to include/exclude sequelae

[catalamarti]

Instead of using a exclude by exclude = "due to" and so. It would be nice to have an option that is: includeSequelae = FALSE. According to @cgreich if we ask SNOMED we will have a relation between the concepts and if they are or not sequelae

[cgreich]

select distinct concept_id, concept_name -- and whatever else you want to output from concept_ancestor a1 join concept_relationship on descendant_concept_id=concept_id_1 and relationship_id in ('Due to of', 'Occurs before') join concept_ancestor a2 on a2.ancestor_concept_id=concept_id_2 join concept on concept_id=a2.descendant_concept_id and standard_concept='S' where a1.ancestor_concept_id=201820 -- this is where you put the concept_id in for which you want to find the sequelae and vocabulary_id='SNOMED'

[cgreich]

This will find concepts like "Kimmelstiel-Wilson syndrome", which you would probably think of in the weeks before and after final exam in Med School.

[edward-burn]

This is now possible as below:

library(CodelistGenerator)
library(dplyr)
#> Warning: package 'dplyr' was built under R version 4.2.3
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
db <- DBI::dbConnect(RPostgres::Postgres(),
  dbname = Sys.getenv("DB_SERVER_cdmgold202007_dbi"),
  port = Sys.getenv("DB_PORT"),
  host = Sys.getenv("DB_HOST"),
  user = Sys.getenv("DB_USER"),
  password = Sys.getenv("DB_PASSWORD")
)
cdm <- CDMConnector::cdm_from_con(
  con = db,
  cdm_schema = "public"
)

# Occurs before relationship 
codes1 <- getCandidateCodes(cdm=cdm,
                  keywords=c("Acute chest syndrome"),
                  domains = "Condition",
                  includeDescendants=FALSE)
codes2 <- getCandidateCodes(cdm=cdm,
                            keywords=c("Acute chest syndrome"),
                            domains = "Condition",
                            includeDescendants=FALSE, 
                            includeSequela = TRUE)
compareCodelists(codes1, codes2)
#> # A tibble: 4 × 3
#>   concept_id concept_name                                               codelist
#>        <int> <chr>                                                      <chr>   
#> 1     254062 Acute chest syndrome                                       Both    
#> 2     765385 Acute chest syndrome due to sickle cell-hemoglobin C dise… Both    
#> 3      22281 Sickle cell-hemoglobin SS disease                          Only co…
#> 4     443721 Sickle cell-hemoglobin C disease with crisis               Only co…

# Due to of relationship
codes1 <- getCandidateCodes(cdm=cdm,
                            keywords=c("sickle cell-hemoglobin C disease"),
                            domains = "Condition", 
                            exactMatch = TRUE,
                            includeDescendants=FALSE)
codes2 <- getCandidateCodes(cdm=cdm,
                            keywords=c("sickle cell-hemoglobin C disease"),
                            domains = "Condition",
                            exactMatch = TRUE,
                            includeDescendants=FALSE, 
                            includeSequela = TRUE)
compareCodelists(codes1, codes2)
#> # A tibble: 2 × 3
#>   concept_id concept_name                         codelist       
#>        <int> <chr>                                <chr>          
#> 1      24006 Sickle cell-hemoglobin C disease     Both           
#> 2    4334887 Sickle cell-hemoglobin C retinopathy Only codelist 2

^{Created on 2023-03-29 with reprex v2.0.2}

[edward-burn]

@cgreich @catalamarti I'll close this because I think this is now implemented, but please comment/ reopen this issue if this isn't what you were after