PatientProfiles contains functions for adding characteristics to OMOP CDM tables containing patient level data (e.g. condition occurrence, drug exposure, and so on) and OMOP CDM cohort tables. The characteristics that can be added include an individual´s sex, age, and days of prior observation Time varying characteristics, such as age, can be estimated relative to any date in the corresponding table. In addition, PatientProfiles also provides functionality for identifying intersections between a cohort table and OMOP CDM tables containing patient level data or other cohort tables.
You can install the latest version of PatientProfiles like so:
install.packages("PatientProfiles")
citation("PatientProfiles")
#>
#> To cite package 'PatientProfiles' in publications use:
#>
#> Catala M, Guo Y, Du M, Lopez-Guell K, Burn E, Mercade-Besora N
#> (????). _PatientProfiles: Identify Characteristics of Patients in the
#> OMOP Common Data Model_. R package version 1.2.1,
#> <https://darwin-eu.github.io/PatientProfiles/>.
#>
#> A BibTeX entry for LaTeX users is
#>
#> @Manual{,
#> title = {PatientProfiles: Identify Characteristics of Patients in the OMOP Common Data Model},
#> author = {Marti Catala and Yuchen Guo and Mike Du and Kim Lopez-Guell and Edward Burn and Nuria Mercade-Besora},
#> note = {R package version 1.2.1},
#> url = {https://darwin-eu.github.io/PatientProfiles/},
#> }
The PatientProfiles package is designed to work with data in the OMOP CDM format, so our first step is to create a reference to the data using the CDMConnector package.
library(CDMConnector)
library(PatientProfiles)
library(dplyr)
Creating a connection to a Postgres database would for example look like:
<- DBI::dbConnect(
con ::Postgres(),
RPostgresdbname = Sys.getenv("CDM5_POSTGRESQL_DBNAME"),
host = Sys.getenv("CDM5_POSTGRESQL_HOST"),
user = Sys.getenv("CDM5_POSTGRESQL_USER"),
password = Sys.getenv("CDM5_POSTGRESQL_PASSWORD")
)
<- cdm_from_con(
cdm
con,cdm_schema = Sys.getenv("CDM5_POSTGRESQL_CDM_SCHEMA"),
write_schema = Sys.getenv("CDM5_POSTGRESQL_RESULT_SCHEMA")
)
To see how you would create a reference to your database please consult the CDMConnector package documentation. For this example though we’ll work with simulated data, and we’ll generate an example cdm reference like so:
<- mockPatientProfiles(numberIndividuals = 1000) cdm
Say we wanted to get individuals´sex and age at condition start date
for records in the condition occurrence table. We can use the
addAge
and addSex
functions to do this:
$condition_occurrence |>
cdmglimpse()
#> Rows: ??
#> Columns: 6
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ person_id <int> 778, 530, 421, 858, 144, 579, 646, 185, 174,…
#> $ condition_start_date <date> 1912-06-09, 1990-01-17, 1946-11-29, 1953-11…
#> $ condition_end_date <date> 1922-03-03, 1991-11-16, 1949-05-22, 1965-10…
#> $ condition_occurrence_id <int> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 1…
#> $ condition_concept_id <int> 2, 5, 5, 5, 7, 7, 10, 4, 10, 5, 6, 7, 2, 4, …
#> $ condition_type_concept_id <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
$condition_occurrence <- cdm$condition_occurrence |>
cdmaddAge(indexDate = "condition_start_date") |>
addSex()
$condition_occurrence |>
cdmglimpse()
#> Rows: ??
#> Columns: 8
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ person_id <int> 778, 530, 858, 144, 579, 185, 174, 426, 9, 5…
#> $ condition_start_date <date> 1912-06-09, 1990-01-17, 1953-11-10, 1966-05…
#> $ condition_end_date <date> 1922-03-03, 1991-11-16, 1965-10-24, 1966-07…
#> $ condition_occurrence_id <int> 1, 2, 4, 5, 6, 8, 9, 10, 11, 13, 14, 15, 17,…
#> $ condition_concept_id <int> 2, 5, 5, 7, 7, 4, 10, 5, 6, 2, 4, 1, 6, 5, 6…
#> $ condition_type_concept_id <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
#> $ age <int> 5, 31, 22, 16, 50, 0, 16, 28, 20, 10, 34, 7,…
#> $ sex <chr> "Male", "Male", "Male", "Female", "Male", "M…
We could, for example, then limit our data to only males aged between 18 and 65
$condition_occurrence |>
cdmfilter(age >= 18 & age <= 65) |>
filter(sex == "Male")
#> # Source: SQL [?? x 8]
#> # Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> person_id condition_start_date condition_end_date condition_occurrence_id
#> <int> <date> <date> <int>
#> 1 530 1990-01-17 1991-11-16 2
#> 2 858 1953-11-10 1965-10-24 4
#> 3 579 1996-12-07 1998-01-13 6
#> 4 426 1991-03-14 1997-04-27 10
#> 5 9 1966-12-27 1969-06-02 11
#> 6 32 2000-06-03 2008-07-17 17
#> 7 9 1973-12-31 1987-05-25 23
#> 8 649 2000-08-22 2008-02-22 33
#> 9 284 1931-11-30 1939-01-08 49
#> 10 707 1977-02-11 1978-10-10 54
#> # ℹ more rows
#> # ℹ 4 more variables: condition_concept_id <int>,
#> # condition_type_concept_id <int>, age <int>, sex <chr>
As with other tables in the OMOP CDM, we can work in a similar way with cohort tables. For example, say we have the below cohort table
$cohort1 |>
cdmglimpse()
#> Rows: ??
#> Columns: 4
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ cohort_definition_id <int> 1, 1, 2, 1, 2, 3, 1, 3, 2, 3, 1, 2, 1, 2, 1, 2, 2…
#> $ subject_id <int> 93, 783, 571, 218, 462, 103, 15, 404, 602, 333, 1…
#> $ cohort_start_date <date> 1970-07-10, 1943-10-13, 1933-05-15, 1934-06-18, …
#> $ cohort_end_date <date> 1984-02-26, 1947-11-11, 1933-06-11, 1949-09-17, …
We can add age, age groups, sex, and days of prior observation to a cohort like so
$cohort1 <- cdm$cohort1 |>
cdmaddAge(
indexDate = "cohort_start_date",
ageGroup = list(c(0, 18), c(19, 65), c(66, 100))
|>
) addSex() |>
addPriorObservation()
$cohort1 |>
cdmglimpse()
#> Rows: ??
#> Columns: 8
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ cohort_definition_id <int> 1, 2, 2, 2, 1, 3, 2, 2, 3, 1, 1, 2, 1, 2, 2, 2, 3…
#> $ subject_id <int> 2, 4, 5, 7, 8, 9, 11, 13, 14, 15, 17, 18, 22, 24,…
#> $ cohort_start_date <date> 1931-10-24, 1975-04-04, 1978-03-13, 1914-07-03, …
#> $ cohort_end_date <date> 1953-08-30, 1987-07-31, 1988-01-13, 1925-06-15, …
#> $ age <int> 10, 0, 3, 0, 12, 5, 21, 14, 8, 16, 0, 10, 1, 0, 8…
#> $ age_group <chr> "0 to 18", "0 to 18", "0 to 18", "0 to 18", "0 to…
#> $ sex <chr> "Male", "Male", "Female", "Female", "Male", "Male…
#> $ prior_observation <int> 3948, 93, 1167, 183, 4705, 2100, 7942, 5238, 3203…
We could use this information to subset the cohort. For example limiting to those with at least 365 days of prior observation available before their cohort start date like so
$cohort1 |>
cdmfilter(prior_observation >= 365)
#> # Source: SQL [?? x 8]
#> # Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> cohort_definition_id subject_id cohort_start_date cohort_end_date age
#> <int> <int> <date> <date> <int>
#> 1 1 2 1931-10-24 1953-08-30 10
#> 2 2 5 1978-03-13 1988-01-13 3
#> 3 1 8 1933-11-19 1935-05-11 12
#> 4 3 9 1951-10-02 1966-05-19 5
#> 5 2 11 1987-09-30 1991-02-25 21
#> 6 2 13 1934-05-05 1937-09-21 14
#> 7 3 14 1988-10-08 1990-12-31 8
#> 8 1 15 1920-05-19 1924-07-11 16
#> 9 2 18 1973-11-17 1978-07-31 10
#> 10 1 22 1938-09-28 1956-07-31 1
#> # ℹ more rows
#> # ℹ 3 more variables: age_group <chr>, sex <chr>, prior_observation <int>
We can use addCohortIntersectFlag
to add a flag for the
presence (or not) of a cohort in a certain window.
$cohort1 |>
cdmglimpse()
#> Rows: ??
#> Columns: 4
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ cohort_definition_id <int> 3, 1, 2, 2, 3, 2, 2, 2, 1, 3
#> $ subject_id <int> 10, 4, 1, 5, 3, 7, 9, 6, 2, 8
#> $ cohort_start_date <date> 1974-02-03, 1918-10-07, 1967-07-21, 1951-05-19, 1…
#> $ cohort_end_date <date> 1974-12-26, 1920-03-27, 1967-12-09, 1953-09-30, 1…
$cohort1 <- cdm$cohort1 |>
cdmaddCohortIntersectFlag(
targetCohortTable = "cohort2",
window = c(-Inf, -1)
)
$cohort1 |>
cdmglimpse()
#> Rows: ??
#> Columns: 7
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ cohort_definition_id <int> 3, 2, 2, 2, 2, 1, 3, 2, 1, 3
#> $ subject_id <int> 10, 1, 5, 7, 9, 4, 3, 6, 2, 8
#> $ cohort_start_date <date> 1974-02-03, 1967-07-21, 1951-05-19, 1934-06-19, 1…
#> $ cohort_end_date <date> 1974-12-26, 1967-12-09, 1953-09-30, 1938-06-10, 1…
#> $ cohort_1_minf_to_m1 <dbl> 0, 1, 0, 0, 1, 0, 0, 0, 0, 0
#> $ cohort_2_minf_to_m1 <dbl> 1, 0, 0, 0, 0, 0, 0, 0, 0, 0
#> $ cohort_3_minf_to_m1 <dbl> 0, 0, 1, 1, 0, 0, 0, 0, 0, 0
If we wanted the number of appearances, we could instead use the
addCohortIntersectCount
function
$cohort1 |>
cdmglimpse()
#> Rows: ??
#> Columns: 4
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ cohort_definition_id <int> 2, 1, 3, 2, 1, 1, 2, 3, 2, 1
#> $ subject_id <int> 4, 1, 7, 2, 3, 5, 10, 6, 9, 8
#> $ cohort_start_date <date> 1961-04-27, 1997-02-08, 1956-03-09, 1980-07-04, 1…
#> $ cohort_end_date <date> 1970-08-24, 2011-04-10, 1956-12-19, 1989-02-23, 1…
$cohort1 <- cdm$cohort1 |>
cdmaddCohortIntersectCount(
targetCohortTable = "cohort2",
targetCohortId = 1,
window = list("short_term" = c(1, 30), "mid_term" = c(31, 180))
)
$cohort1 |>
cdmglimpse()
#> Rows: ??
#> Columns: 6
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ cohort_definition_id <int> 2, 2, 1, 3, 2, 1, 1, 2, 3, 1
#> $ subject_id <int> 9, 4, 1, 7, 2, 3, 5, 10, 6, 8
#> $ cohort_start_date <date> 1922-08-27, 1961-04-27, 1997-02-08, 1956-03-09, 1…
#> $ cohort_end_date <date> 1928-06-22, 1970-08-24, 2011-04-10, 1956-12-19, 1…
#> $ cohort_1_mid_term <dbl> 1, 0, 0, 0, 0, 0, 0, 0, 0, 0
#> $ cohort_1_short_term <dbl> 1, 0, 0, 0, 0, 0, 0, 0, 0, 0
Say we wanted the date at which an individual was in another cohort
then we can use the addCohortIntersectDate
function. As
there might be multiple records for the other cohort, we can also choose
the first or the last appearance in that cohort.
First occurrence:
$cohort1 |>
cdmglimpse()
#> Rows: ??
#> Columns: 4
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ cohort_definition_id <int> 3, 3, 2, 1, 1, 1, 3, 1, 3, 2
#> $ subject_id <int> 9, 4, 6, 8, 1, 5, 10, 7, 2, 3
#> $ cohort_start_date <date> 1972-05-17, 1936-05-20, 1954-03-09, 1935-10-23, 2…
#> $ cohort_end_date <date> 1988-05-11, 1945-08-14, 1966-11-09, 1937-04-23, 2…
$cohort1 <- cdm$cohort1 |>
cdmaddCohortIntersectDate(
targetCohortTable = "cohort2",
targetCohortId = 1,
order = "first",
window = c(-Inf, Inf)
)
$cohort1 |>
cdmglimpse()
#> Rows: ??
#> Columns: 5
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ cohort_definition_id <int> 3, 2, 1, 2, 3, 1, 1, 3, 1, 3
#> $ subject_id <int> 9, 6, 8, 3, 4, 1, 5, 10, 7, 2
#> $ cohort_start_date <date> 1972-05-17, 1954-03-09, 1935-10-23, 1938-12-02, 1…
#> $ cohort_end_date <date> 1988-05-11, 1966-11-09, 1937-04-23, 1941-11-19, 1…
#> $ cohort_1_minf_to_inf <date> 1980-05-05, 1967-03-31, 1916-11-14, 1919-01-09, …
Last occurrence:
$cohort1 |>
cdmglimpse()
#> Rows: ??
#> Columns: 4
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ cohort_definition_id <int> 2, 3, 3, 2, 3, 2, 2, 1, 2, 3
#> $ subject_id <int> 4, 10, 7, 1, 2, 9, 6, 3, 8, 5
#> $ cohort_start_date <date> 1921-12-06, 1989-08-06, 1992-06-11, 1981-10-14, 1…
#> $ cohort_end_date <date> 1928-08-03, 1995-01-24, 1997-09-11, 1982-11-12, 1…
$cohort1 <- cdm$cohort1 |>
cdmaddCohortIntersectDate(
targetCohortTable = "cohort2",
targetCohortId = 1,
order = "last",
window = c(-Inf, Inf)
)
$cohort1 |>
cdmglimpse()
#> Rows: ??
#> Columns: 5
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ cohort_definition_id <int> 2, 2, 1, 2, 3, 3, 2, 3, 2, 3
#> $ subject_id <int> 9, 6, 3, 4, 10, 7, 1, 2, 8, 5
#> $ cohort_start_date <date> 1969-11-14, 1938-04-30, 1983-05-27, 1921-12-06, 1…
#> $ cohort_end_date <date> 1973-09-22, 1939-02-01, 1993-11-26, 1928-08-03, 1…
#> $ cohort_1_minf_to_inf <date> 1971-02-23, 1942-11-21, 1971-04-30, NA, NA, NA, …
Instead of returning a date, we could return the days to the
intersection by using addCohortIntersectDays
$cohort1 |>
cdmglimpse()
#> Rows: ??
#> Columns: 4
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ cohort_definition_id <int> 3, 2, 2, 1, 1, 3, 1, 1, 2, 1
#> $ subject_id <int> 1, 9, 5, 7, 4, 3, 2, 10, 6, 8
#> $ cohort_start_date <date> 1966-12-06, 1955-09-03, 1963-11-24, 1995-11-30, 1…
#> $ cohort_end_date <date> 1990-01-06, 1977-12-24, 1978-05-27, 1996-07-27, 2…
$cohort1 <- cdm$cohort1 |>
cdmaddCohortIntersectDays(
targetCohortTable = "cohort2",
targetCohortId = 1,
order = "last",
window = c(-Inf, Inf)
)
$cohort1 |>
cdmglimpse()
#> Rows: ??
#> Columns: 5
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ cohort_definition_id <int> 2, 1, 1, 3, 2, 1, 1, 3, 1, 2
#> $ subject_id <int> 5, 2, 8, 1, 9, 7, 4, 3, 10, 6
#> $ cohort_start_date <date> 1963-11-24, 1913-08-23, 1982-09-08, 1966-12-06, 1…
#> $ cohort_end_date <date> 1978-05-27, 1914-08-12, 1991-06-28, 1990-01-06, 1…
#> $ cohort_1_minf_to_inf <dbl> 8809, 4412, -2029, NA, NA, NA, NA, NA, NA, NA
If we want to combine multiple cohort intersects we can concatenate
the operations using the pipe
operator:
$cohort1 |>
cdmglimpse()
#> Rows: ??
#> Columns: 4
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ cohort_definition_id <int> 2, 1, 2, 3, 2, 3, 3, 2, 2, 3
#> $ subject_id <int> 2, 4, 3, 7, 10, 1, 6, 8, 9, 5
#> $ cohort_start_date <date> 1945-02-11, 1955-06-28, 1957-11-19, 1986-03-17, 1…
#> $ cohort_end_date <date> 1980-06-10, 1962-08-23, 1958-05-22, 1992-08-12, 1…
$cohort1 <- cdm$cohort1 |>
cdmaddCohortIntersectDate(
targetCohortTable = "cohort2",
targetCohortId = 1,
order = "last",
window = c(-Inf, Inf)
|>
) addCohortIntersectCount(
targetCohortTable = "cohort2",
targetCohortId = 1,
window = c(-Inf, Inf)
)
$cohort1 |>
cdmglimpse()
#> Rows: ??
#> Columns: 5
#> Database: DuckDB v1.1.1 [martics@Windows 10 x64:R 4.2.1/:memory:]
#> $ cohort_definition_id <int> 2, 3, 2, 1, 2, 3, 2, 3, 3, 2
#> $ subject_id <int> 8, 5, 2, 4, 3, 7, 10, 1, 6, 9
#> $ cohort_start_date <date> 1972-02-17, 1944-11-17, 1945-02-11, 1955-06-28, 1…
#> $ cohort_end_date <date> 1987-10-18, 1947-12-20, 1980-06-10, 1962-08-23, 1…
#> $ cohort_1_minf_to_inf <dbl> 1, 1, 0, 0, 0, 0, 0, 0, 0, 0
mockDisconnect(cdm)