README.md

CodelistGenerator

Introduction

CodelistGenerator is used to create a candidate set of codes for helping to define patient cohorts in data mapped to the OMOP common data model. A little like the process for a systematic review, the idea is that for a specified search strategy, CodelistGenerator will identify a set of concepts that may be relevant, with these then being screened to remove any irrelevant codes.

Installation

You can install the development version of CodelistGenerator like so:

install.packages("remotes")
remotes::install_github("darwin-eu/CodelistGenerator")

Connecting to the OMOP CDM vocabularies

# example with postgres database connection details
server_dbi<-Sys.getenv("server")
user<-Sys.getenv("user")
password<- Sys.getenv("password")
port<-Sys.getenv("port")
host<-Sys.getenv("host")

db <- DBI::dbConnect(RPostgres::Postgres(),
                dbname = server_dbi,
                port = port,
                host = host,
                user = user,
                password = password)

# name of vocabulary schema
vocabulary_database_schema<-Sys.getenv("vocabulary_schema")

Example search

Every codelist is specific to a version of the OMOP CDM vocabularies, so we can first check the version.

dplyr::tbl(db, dplyr::sql(paste0(
    "SELECT * FROM ",
    vocabulary_database_schema,
    ".vocabulary"
    ))) %>%
    dplyr::rename_with(tolower) %>%
    dplyr::filter(.data$vocabulary_id == "None") %>%
    dplyr::select("vocabulary_version") %>%
    dplyr::collect() %>%
    dplyr::pull()
#> [1] "v5.0 13-JUL-21"

We can then search for asthma like so

asthma_1<-get_candidate_codes(keywords="asthma",
                    domains = "Condition",
                    db=db,
                    vocabulary_database_schema = vocabulary_database_schema)
kable(head(asthma_1, 10))

concept_id	concept_name	domain_id	vocabulary_id
761844	Inhaled steroid-dependent asthma	Condition	SNOMED
764677	Persistent asthma	Condition	SNOMED
764949	Persistent asthma, well controlled	Condition	SNOMED
3661412	Thunderstorm asthma	Condition	SNOMED
4015819	Asthma disturbs sleep weekly	Condition	SNOMED
4015947	Asthma causing night waking	Condition	SNOMED
4017025	Asthma disturbing sleep	Condition	SNOMED
4017026	Asthma not limiting activities	Condition	SNOMED
4017182	Asthma disturbs sleep frequently	Condition	SNOMED
4017183	Asthma not disturbing sleep	Condition	SNOMED

Perhaps we want to exclude certain concepts as part of the search strategy, in which case this can be added like so

asthma_2<-get_candidate_codes(keywords="asthma",
                    domains = "Condition",
                    exclude = "Poisoning by antiasthmatic",
                    db=db,
                    vocabulary_database_schema = vocabulary_database_schema)
kable(head(asthma_2, 10))

concept_id	concept_name	domain_id	vocabulary_id
761844	Inhaled steroid-dependent asthma	Condition	SNOMED
764677	Persistent asthma	Condition	SNOMED
764949	Persistent asthma, well controlled	Condition	SNOMED
3661412	Thunderstorm asthma	Condition	SNOMED
4015819	Asthma disturbs sleep weekly	Condition	SNOMED
4015947	Asthma causing night waking	Condition	SNOMED
4017025	Asthma disturbing sleep	Condition	SNOMED
4017026	Asthma not limiting activities	Condition	SNOMED
4017182	Asthma disturbs sleep frequently	Condition	SNOMED
4017183	Asthma not disturbing sleep	Condition	SNOMED

We can then also see source codes these are mapped from, for example

asthma_icd_mappings<-show_mappings(candidate_codelist=asthma_2,
                     source_vocabularies="ICD10CM",
                    db=db,
                    vocabulary_database_schema =  vocabulary_database_schema)
kable(head(asthma_icd_mappings %>% 
       select(standard_concept_name,
              standard_vocabulary_id,
              source_concept_name,
              source_vocabulary_id),
     10))

standard_concept_name	standard_vocabulary_id	source_concept_name	source_vocabulary_id
Eosinophilic asthma	SNOMED	Pulmonary eosinophilia, not elsewhere classified	ICD10CM
Eosinophilic asthma	SNOMED	Eosinophilic asthma	ICD10CM
Eosinophilic asthma	SNOMED	Other pulmonary eosinophilia, not elsewhere classified	ICD10CM
Eosinophilic asthma	SNOMED	Pulmonary eosinophilia, not elsewhere classified	ICD10CM
Cryptogenic pulmonary eosinophilia	SNOMED	Chronic eosinophilic pneumonia	ICD10CM
Simple pulmonary eosinophilia	SNOMED	Acute eosinophilic pneumonia	ICD10CM
Asthma	SNOMED	Asthma	ICD10CM
Asthma	SNOMED	Other and unspecified asthma	ICD10CM
Asthma	SNOMED	Unspecified asthma	ICD10CM
Asthma	SNOMED	Other asthma	ICD10CM