Add realistic test data (#24)

* Insert dummy data to MEASUREMENT and OBSERVATION for the test db. Create realistic test dataset.

* Clean CSV files.

* Regenerate realistic test data after cleaning dummy CSVs.

* Add script to produce the realistic test data.

* Update dev/test_db/produce_test_data.R

Co-authored-by: Milan Malfait <[email protected]>

* Update README to explain how the reproduce test data.

* Add 'ORDER BY' clause to produce results.

* Update README with test dataset creation process.

* Update README with test dataset creation process.

* Update README with test dataset creation process.

* Make sure DB connections are always closed

* Remove clean up steps

* Remove more cleanup steps

* Use test data in data getters

* Test consistency of test_data files

* Fix indentation

* Improve logging

* Update test data

* Update test data

* Overwrite existing data to ensure consistency

* Add sanity checks for dummy data

Make sure the data in the database matches the local dummy files

* Explicitly specify column types for dummy data to avoid errors later on

* Overwrite database tables for summary results as well

* Update test data (should be consistent now)

* Use `readr` to write data as we're dealing with tibbles

* Update snapshot test data

* Just do a simple `arrange()` across all columns to order data

* Clean up code

* Reduce test data size

* Readd sorting of tables

* Update test data (really the last time now)

* Read test data as tibble

* Add readr as dependency

---------

Co-authored-by: Milan Malfait <[email protected]>

.Rprofile

@@ -16,7 +16,7 @@ source("renv/activate.R")
 # Path to download Eunomia datasets
 Sys.setenv(EUNOMIA_DATA_FOLDER = file.path("dev/test_db/eunomia"))
 # Name of the synthetic dataset to use
-Sys.setenv(TEST_DB_NAME = "GiBleed")
+Sys.setenv(TEST_DB_NAME = "synthea-allergies-10k")
 # OMOP CDM version
 Sys.setenv(TEST_DB_OMOP_VERSION = "5.3")
 # Schema name for data

DESCRIPTION

@@ -16,6 +16,7 @@ Imports:
     glue,
     tidyr,
     withr,
+    readr,
     lubridate,
     dplyr
 Suggests:

R/utils_get_data.R

@@ -5,19 +5,9 @@
 #' @noRd
 get_concepts_table <- function() {
   if (golem::app_dev()) {
-    return(data.frame(
-      concept_id = c(40213251, 133834, 4057420),
-      concept_name = c(
-        "varicella virus vaccine",
-        "Atopic dermatitis",
-        "Catheter ablation of tissue of heart"
-      ),
-      domain_id = c("Drug", "Condition", "Procedure"),
-      vocabulary_id = c("CVX", "SNOMED", "SNOMED"),
-      concept_class_id = c("CVX", "Clinical Finding", "Procedure"),
-      standard_concept = c("S", "S", "S"),
-      concept_code = c("21", "24079001", "18286008")
-    ))
+    return(
+      readr::read_csv(app_sys("test_data", "calypso_concepts.csv"), show_col_types = FALSE)
+    )
   }
 
   con <- connect_to_test_db()
@@ -28,15 +18,7 @@ get_concepts_table <- function() {
 get_monthly_counts <- function() {
   if (golem::app_dev()) {
     return(
-      data.frame(
-        concept_id = c(
-          rep(c(40213251, 133834, 4057420), each = 3)
-        ),
-        date_year = c(2019L, 2020L, 2020L, 2019L, 2020L, 2020L, 2020L, 2019L, 2019L),
-        date_month = c(4L, 3L, 5L, 5L, 8L, 4L, 11L, 6L, 3L),
-        person_count = c(1, 1, 3, 4, 2, 3, 2, 4, 1),
-        records_per_person = c(1, 1, 1, 1, 1, 1, 1, 1, 1)
-      )
+      readr::read_csv(app_sys("test_data", "calypso_monthly_counts.csv"), show_col_types = FALSE)
     )
   }
 
@@ -48,12 +30,7 @@ get_monthly_counts <- function() {
 get_summary_stats <- function() {
   if (golem::app_dev()) {
     return(
-      data.frame(
-        concept_id = rep(c(40213251, 133834, 4057420), each = 2),
-        summary_attribute = rep(c("mean", "sd"), times = 3),
-        value_as_string = rep(NA, 6),
-        value_as_number = c(1.5, 0.5, 2.5, 0.7, 3.5, 0.8)
-      )
+      readr::read_csv(app_sys("test_data", "calypso_summary_stats.csv"), show_col_types = FALSE)
     )
   }
 

README.md

@@ -50,6 +50,17 @@ as it has good support for R package development and Shiny.
 
 The `dev/02_dev.R` script contains a few helper functions to get you started.
 
+Calypso test data can be found in [`inst/test_data`](https://github.com/UCLH-Foundry/omop-data-catalogue/tree/main/inst/data). These data have been generated by using the synthetic dataset '[synthea-allergies-10k](https://darwin-eu.github.io/CDMConnector/reference/eunomiaDir.html)', and adding some [dummy data](https://github.com/UCLH-Foundry/omop-data-catalogue/tree/main/dev/test_db/dummy) for the MEASUREMENT and OBSERVATION tables (to have some records in the 'calypso-summary-stats' table).
+
+If you want to recreate a test dataset, you can run the following R scripts:
+
+```r
+source(here::here("dev/test_db/setup_test_db.R"))
+source(here::here("dev/test_db/insert_dummy_tables.R"))
+source(here::here("dev/omop_analyses/analyse_omop_cdm.R"))
+source(here::here("dev/test_db/produce_test_data.R"))
+```
+
 ### Updating the `renv` lockfile
 
 Make sure to regularly run `renv::status(dev = TRUE)` to check if your local library and the lockfile

dev/omop_analyses/analyse_omop_cdm.R

@@ -1,4 +1,8 @@
-library(tidyverse)
+cli::cli_h1("Generating summarys statistics")
+
+suppressPackageStartupMessages(
+  library(tidyverse)
+)
 
 dir <- Sys.getenv("EUNOMIA_DATA_FOLDER")
 name <- Sys.getenv("TEST_DB_NAME")
@@ -180,8 +184,7 @@ write_results <- function(data, con, table) {
       table = table
     ),
     value = data,
-    append = TRUE,
-    overwrite = FALSE
+    overwrite = TRUE
   )
 }
 
@@ -220,3 +223,5 @@ ids <- unique(c(monthly_counts$concept_id, summary_stats$concept_id))
 # Retrieve concept properties from the list of ids
 get_concepts_table(cdm, ids) |>
   write_results(con, "calypso_concepts")
+
+cli::cli_alert_success("Summary statistics generated successfully")

dev/test_db/dummy/measurement.csv

@@ -0,0 +1,24 @@
+measurement_id,person_id,measurement_concept_id,measurement_date,measurement_datetime,measurement_time,measurement_type_concept_id,operator_concept_id,value_as_number,value_as_concept_id,unit_concept_id,range_low,range_high,provider_id,visit_occurrence_id,visit_detail_id,measurement_source_value,measurement_source_concept_id,unit_source_value,value_source_value
+10000000,103,4354252,2021-06-01,2021-06-01T10:45:00Z,NA,32817,0,115,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000001,12,4354252,2020-08-25,2020-08-25T14:15:00Z,NA,32817,0,122,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000002,866,4354252,2021-11-14,2021-11-14T13:05:00Z,NA,32817,0,125,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000003,12,4354252,2019-08-23,2019-08-23T02:28:00Z,NA,32817,0,131,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000004,51,4354252,2021-02-11,2021-02-11T15:01:00Z,NA,32817,0,111,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000005,3028,4354252,2018-01-18,2018-01-18T12:14:00Z,NA,32817,0,138,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000006,7,4248525,2021-10-15,2021-10-15T10:20:00Z,NA,32817,0,169,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000007,553,4248525,2015-04-19,2015-04-19T16:47:00Z,NA,32817,0,131,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000008,1641,4248525,2019-10-11,2019-10-11T07:00:00Z,NA,32817,0,128,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000009,553,4248525,2020-06-26,2020-06-26T00:00:00Z,NA,32817,0,114,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000010,12,4248525,2019-05-01,2019-05-01T20:55:00Z,NA,32817,0,122,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000011,978,4353843,2002-03-03,2002-03-03T22:00:00Z,NA,32817,0,162,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000012,12,4353843,2021-09-18,2021-09-18T02:00:00Z,NA,32817,0,152,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000013,6459,4353843,2021-12-28,2021-12-28T02:00:00Z,NA,32817,0,118,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000014,995,4353843,2023-04-08,2023-04-08T08:00:00Z,NA,32817,0,99,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000015,110,4353843,2015-09-20,2015-09-20T07:00:00Z,NA,32817,0,117,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000016,8746,4353843,2021-10-01,2021-10-01T09:00:00Z,NA,32817,0,125,0,8876,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000017,978,4108450,2001-06-15,2001-06-15T07:50:00Z,NA,32817,0,0.6666666666666666,0,8523,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000018,8916,4108450,2019-09-13,2019-09-13T08:29:00Z,NA,32817,0,0.6666666666666666,0,8523,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000019,51,3001079,2020-05-15,2020-05-15T22:44:00Z,NA,32817,0,NA,45878588,0,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000020,909,3001079,2018-03-11,2018-03-11T13:30:00Z,NA,32817,0,NA,45878588,0,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000021,553,4128111,2020-12-02,2020-12-02T00:00:00Z,NA,32817,0,NA,1635564,0,NA,NA,NA,NA,NA,NA,NA,NA,NA
+10000022,7,4128111,2020-11-25,2020-11-25T00:00:00Z,NA,32817,0,NA,1633781,0,NA,NA,NA,NA,NA,NA,NA,NA,NA

dev/test_db/dummy/observation.csv

@@ -0,0 +1,13 @@
+observation_id,person_id,observation_concept_id,observation_date,observation_datetime,observation_type_concept_id,value_as_number,value_as_string,value_as_concept_id,qualifier_concept_id,unit_concept_id,provider_id,visit_occurrence_id,visit_detail_id,observation_source_value,observation_source_concept_id,unit_source_value,qualifier_source_value
+10000000,11,45766147,2022-06-24,2022-06-24T09:00:00Z,32817,NA,NA,4086518,NA,0,NA,NA,NA,NA,NA,NA,NA
+10000001,59,4257036,2018-09-02,2018-09-02T09:19:00Z,32817,NA,NA,37208662,NA,0,NA,NA,NA,NA,NA,NA,NA
+10000002,237,4257036,2014-11-19,2014-11-19T17:10:00Z,32817,NA,NA,37208662,NA,0,NA,NA,NA,NA,NA,NA,NA
+10000003,299,4257036,2017-02-17,2017-02-17T11:14:00Z,32817,NA,NA,37208662,NA,0,NA,NA,NA,NA,NA,NA,NA
+10000004,673,4216746,2011-03-22,2011-03-22T16:00:00Z,32817,8,NA,0,NA,44777590,NA,NA,NA,NA,NA,NA,NA
+10000005,11,4353717,2022-12-05,2022-12-05T17:00:00Z,32817,10.6,NA,0,NA,8698,NA,NA,NA,NA,NA,NA,NA
+10000006,1502,4353717,2021-05-13,2021-05-13T11:00:00Z,32817,16.8,NA,0,NA,8698,NA,NA,NA,NA,NA,NA,NA
+10000007,986,4216746,2008-09-28,2008-09-28T20:00:00Z,32817,8,NA,0,NA,44777590,NA,NA,NA,NA,NA,NA,NA
+10000008,299,4353717,2016-10-09,2016-10-09T01:00:00Z,32817,11,NA,0,NA,8698,NA,NA,NA,NA,NA,NA,NA
+10000009,299,4353713,2018-01-12,2018-01-12T15:58:00Z,32817,5,NA,0,NA,44777590,NA,NA,NA,NA,NA,NA,NA
+10000010,6288,4353713,2021-07-19,2021-07-19T15:14:00Z,32817,7,NA,0,NA,44777590,NA,NA,NA,NA,NA,NA,NA
+10000011,3362,4353713,2019-02-08,2019-02-08T12:25:00Z,32817,5,NA,0,NA,44777590,NA,NA,NA,NA,NA,NA,NA

dev/test_db/eunomia/.gitignore

@@ -3,3 +3,7 @@
 
 # duckdb databases
 *.duckdb
+
+# duckdb temp files
+# (in case of failure)
+*.duckdb.wal

dev/test_db/insert_dummy_tables.R

@@ -0,0 +1,79 @@
+# PRODUCED FOR A SPECIFIC DATASET:
+# synthea-allergies-10k
+# (but could work for others)
+
+cli::cli_h1("Inserting dummy tables")
+
+library(readr)
+
+dir <- Sys.getenv("EUNOMIA_DATA_FOLDER")
+name <- Sys.getenv("TEST_DB_NAME")
+version <- Sys.getenv("TEST_DB_OMOP_VERSION")
+
+# Connect to the duckdb test database
+con <- DBI::dbConnect(
+  duckdb::duckdb(dbdir = glue::glue("{dir}/{name}_{version}_1.0.duckdb"))
+)
+
+withr::defer(DBI::dbDisconnect(con))
+
+# Function to write data to a table in the cdm schema
+write_table <- function(data, con, table) {
+  # Insert data into the specified table
+  # (in the cdm schema)
+  DBI::dbWriteTable(
+    conn = con,
+    name = DBI::Id(
+      schema = Sys.getenv("TEST_DB_CDM_SCHEMA"),
+      table = table
+    ),
+    value = data,
+    overwrite = TRUE
+  )
+}
+
+## Load dummy data and write tables to database
+## We explicitly set the column types for columns that are needed later down the pipeline
+dummy_measurements <- read_csv(
+  here::here("dev/test_db/dummy/measurement.csv"),
+  col_types = cols(
+    measurement_id = col_double(),
+    person_id = col_double(),
+    measurement_concept_id = col_double(),
+    measurement_date = col_date(),
+    value_as_number = col_double(),
+    value_as_concept_id = col_double(),
+  )
+)
+write_table(dummy_measurements, con, "measurement")
+
+dummy_observations <- read_csv(here::here(
+  "dev/test_db/dummy/observation.csv"),
+  col_types = cols(
+      observation_id = col_double(),
+      person_id = col_double(),
+      observation_concept_id = col_double(),
+      observation_date = col_date(),
+      value_as_number = col_double(),
+      value_as_string = col_logical(),
+      value_as_concept_id = col_double(),
+    )
+)
+write_table(dummy_observations, con, "observation")
+
+# Sanity check: read the data back and make sure its consistent
+db_measurements <- DBI::dbReadTable(con, "measurement")
+stopifnot(all.equal(db_measurements, as.data.frame(dummy_measurements)))
+
+db_observations <- DBI::dbReadTable(con, "observation")
+stopifnot(all.equal(db_observations, as.data.frame(dummy_observations)))
+
+# Load the CMD object to verify integrity of the schema after insertions
+cdm <- CDMConnector::cdm_from_con(
+  con = con,
+  cdm_schema = Sys.getenv("TEST_DB_CDM_SCHEMA"),
+  write_schema = Sys.getenv("TEST_DB_RESULTS_SCHEMA"),
+  cdm_name = name
+)
+
+cli::cli_alert_success("Dummy tables inserted successfully")

dev/test_db/produce_test_data.R

@@ -0,0 +1,52 @@
+cli::cli_h1("Producing test data")
+
+suppressPackageStartupMessages({
+  library(dplyr)
+})
+
+dir <- Sys.getenv("EUNOMIA_DATA_FOLDER")
+name <- Sys.getenv("TEST_DB_NAME")
+version <- Sys.getenv("TEST_DB_OMOP_VERSION")
+
+# Connect to the duckdb test database
+con <- DBI::dbConnect(
+  duckdb::duckdb(dbdir = glue::glue("{dir}/{name}_{version}_1.0.duckdb"))
+)
+withr::defer(DBI::dbDisconnect(con))
+
+# Function to write results from a table to the test data folder
+read_table <- function(con, table) {
+  schema <- Sys.getenv("TEST_DB_RESULTS_SCHEMA")
+  # Get all rows from the table
+  query <- glue::glue("SELECT * FROM {schema}.{table}")
+  # Run the query and write results
+  con |>
+    DBI::dbGetQuery(query) |>
+    arrange(across(everything()))
+}
+
+# Get the relevant tables and filter
+table_names <- c("calypso_concepts", "calypso_monthly_counts", "calypso_summary_stats")
+tables <- purrr::map(table_names, read_table, con = con)
+names(tables) <- table_names
+
+# Keep only concepts for which we have summary statistics
+keep_concepts <- tables$calypso_summary_stats$concept_id
+tables <- purrr::map(tables, ~ .x[.x$concept_id %in% keep_concepts, ])
+
+# Keep only data from 2019 onwards
+monthly_counts <- tables$calypso_monthly_counts
+filtered_monthly <- monthly_counts[monthly_counts$date_year >= 2019, ]
+tables$calypso_monthly_counts <- filtered_monthly
+
+# Filter the other tables to match the concepts left over after year filtering
+tables <- purrr::map(tables, ~ .x[.x$concept_id %in% filtered_monthly$concept_id, ])
+
+# Write all results to the test data folder
+purrr::iwalk(tables, function(tbl, name) {
+  path <- here::here(glue::glue("inst/test_data/{name}.csv"))
+  cli::cli_alert_info("Writing {name} to {path}")
+  readr::write_csv(tbl, file = path)
+})
+
+cli::cli_alert_success("Test data produced")

dev/test_db/setup_test_db.R

@@ -1,3 +1,4 @@
+cli::cli_h1("Setting up test database")
 
 # Create an duckdb database from Eunomia datasets
 con <- DBI::dbConnect(
@@ -18,3 +19,5 @@ CDMConnector::cdm_from_con(
   write_schema = Sys.getenv("TEST_DB_RESULTS_SCHEMA"),
   cdm_name = Sys.getenv("TEST_DB_NAME")
 )
+
+cli::cli_alert_success("Test database setup successfully")

inst/test_data/calypso_concepts.csv

@@ -0,0 +1,10 @@
+concept_id,concept_name,vocabulary_id,domain_id,concept_class_id,standard_concept,concept_code
+3001079,Blood group antibody screen [Presence] in Serum or Plasma,LOINC,Measurement,Lab Test,S,890-4
+4108450,Inspiration/expiration time ratio,SNOMED,Measurement,Observable Entity,S,250822000
+4128111,T - Tumor stage,SNOMED,Observation,Attribute,S,260878002
+4248525,Lying systolic blood pressure,SNOMED,Measurement,Observable Entity,S,407556006
+4353713,Positive end expiratory pressure,SNOMED,Observation,Observable Entity,S,250854009
+4353717,Ventilator delivered minute volume,SNOMED,Observation,Observable Entity,S,250875001
+4353843,Invasive systolic arterial pressure,SNOMED,Measurement,Observable Entity,S,251071003
+4354252,Non-invasive systolic arterial pressure,SNOMED,Measurement,Observable Entity,S,251070002
+45766147,Appearance,SNOMED,Observation,Observable Entity,S,703248002

inst/test_data/calypso_monthly_counts.csv

@@ -0,0 +1,23 @@
+concept_id,concept_name,date_year,date_month,person_count,records_per_person
+3001079,Blood group antibody screen [Presence] in Serum or Plasma,2020,5,1,1
+4108450,Inspiration/expiration time ratio,2019,9,1,1
+4128111,T - Tumor stage,2020,11,1,1
+4128111,T - Tumor stage,2020,12,1,1
+4248525,Lying systolic blood pressure,2019,5,1,1
+4248525,Lying systolic blood pressure,2019,10,1,1
+4248525,Lying systolic blood pressure,2020,6,1,1
+4248525,Lying systolic blood pressure,2021,10,1,1
+4353713,Positive end expiratory pressure,2019,2,1,1
+4353713,Positive end expiratory pressure,2021,7,1,1
+4353717,Ventilator delivered minute volume,2021,5,1,1
+4353717,Ventilator delivered minute volume,2022,12,1,1
+4353843,Invasive systolic arterial pressure,2021,9,1,1
+4353843,Invasive systolic arterial pressure,2021,10,1,1
+4353843,Invasive systolic arterial pressure,2021,12,1,1
+4353843,Invasive systolic arterial pressure,2023,4,1,1
+4354252,Non-invasive systolic arterial pressure,2019,8,1,1
+4354252,Non-invasive systolic arterial pressure,2020,8,1,1
+4354252,Non-invasive systolic arterial pressure,2021,2,1,1
+4354252,Non-invasive systolic arterial pressure,2021,6,1,1
+4354252,Non-invasive systolic arterial pressure,2021,11,1,1
+45766147,Appearance,2022,6,1,1

inst/test_data/calypso_summary_stats.csv

@@ -0,0 +1,17 @@
+concept_id,concept_name,summary_attribute,value_as_number,value_as_string
+3001079,Blood group antibody screen [Presence] in Serum or Plasma,frequency,2,Not present
+4108450,Inspiration/expiration time ratio,mean,0.6666666666666666,NA
+4108450,Inspiration/expiration time ratio,sd,0,NA
+4128111,T - Tumor stage,frequency,1,NA
+4128111,T - Tumor stage,frequency,1,NA
+4248525,Lying systolic blood pressure,mean,132.8,NA
+4248525,Lying systolic blood pressure,sd,21.25323504786977,NA
+4353713,Positive end expiratory pressure,mean,5.666666666666667,NA
+4353713,Positive end expiratory pressure,sd,1.1547005383792517,NA
+4353717,Ventilator delivered minute volume,mean,12.799999999999999,NA
+4353717,Ventilator delivered minute volume,sd,3.469870314579495,NA
+4353843,Invasive systolic arterial pressure,mean,128.83333333333334,NA
+4353843,Invasive systolic arterial pressure,sd,23.65938855225694,NA
+4354252,Non-invasive systolic arterial pressure,mean,123.66666666666667,NA
+4354252,Non-invasive systolic arterial pressure,sd,9.993331109628395,NA
+45766147,Appearance,frequency,1,Well nourished

tests/testthat/_snaps/utils_get_data/concepts_table.csv

@@ -0,0 +1,10 @@
+concept_id,concept_name,vocabulary_id,domain_id,concept_class_id,standard_concept,concept_code
+3001079,Blood group antibody screen [Presence] in Serum or Plasma,LOINC,Measurement,Lab Test,S,890-4
+4108450,Inspiration/expiration time ratio,SNOMED,Measurement,Observable Entity,S,250822000
+4128111,T - Tumor stage,SNOMED,Observation,Attribute,S,260878002
+4248525,Lying systolic blood pressure,SNOMED,Measurement,Observable Entity,S,407556006
+4353713,Positive end expiratory pressure,SNOMED,Observation,Observable Entity,S,250854009
+4353717,Ventilator delivered minute volume,SNOMED,Observation,Observable Entity,S,250875001
+4353843,Invasive systolic arterial pressure,SNOMED,Measurement,Observable Entity,S,251071003
+4354252,Non-invasive systolic arterial pressure,SNOMED,Measurement,Observable Entity,S,251070002
+45766147,Appearance,SNOMED,Observation,Observable Entity,S,703248002

tests/testthat/_snaps/utils_get_data/monthly_counts.csv

@@ -0,0 +1,23 @@
+concept_id,concept_name,date_year,date_month,person_count,records_per_person
+3001079,Blood group antibody screen [Presence] in Serum or Plasma,2020,5,1,1
+4108450,Inspiration/expiration time ratio,2019,9,1,1
+4128111,T - Tumor stage,2020,11,1,1
+4128111,T - Tumor stage,2020,12,1,1
+4248525,Lying systolic blood pressure,2019,5,1,1
+4248525,Lying systolic blood pressure,2019,10,1,1
+4248525,Lying systolic blood pressure,2020,6,1,1
+4248525,Lying systolic blood pressure,2021,10,1,1
+4353713,Positive end expiratory pressure,2019,2,1,1
+4353713,Positive end expiratory pressure,2021,7,1,1
+4353717,Ventilator delivered minute volume,2021,5,1,1
+4353717,Ventilator delivered minute volume,2022,12,1,1
+4353843,Invasive systolic arterial pressure,2021,9,1,1
+4353843,Invasive systolic arterial pressure,2021,10,1,1
+4353843,Invasive systolic arterial pressure,2021,12,1,1
+4353843,Invasive systolic arterial pressure,2023,4,1,1
+4354252,Non-invasive systolic arterial pressure,2019,8,1,1
+4354252,Non-invasive systolic arterial pressure,2020,8,1,1
+4354252,Non-invasive systolic arterial pressure,2021,2,1,1
+4354252,Non-invasive systolic arterial pressure,2021,6,1,1
+4354252,Non-invasive systolic arterial pressure,2021,11,1,1
+45766147,Appearance,2022,6,1,1