Phenotype diagnostics

Introduction: Run PhenotypeDiagnostics

In this vignette, we are going to present how to run PhenotypeDiagnostics(). We are going to use the following packages and mock data:

library(CohortConstructor)
library(PhenotypeR)
library(dplyr)

con <- DBI::dbConnect(duckdb::duckdb(), 
                      CDMConnector::eunomiaDir("synpuf-1k", "5.3"))
cdm <- CDMConnector::cdmFromCon(con = con, 
                                cdmName = "Eunomia Synpuf",
                                cdmSchema   = "main",
                                writeSchema = "main", 
                                achillesSchema = "main")
cdm

Note that we have included achilles tables in our cdm reference, which will be used to speed up some of the analyses.

Create a cohort

First, we are going to use the package CohortConstructor to generate three cohorts of warfarin, acetaminophen and morphine users.

# Create a codelist
codes <- list("warfarin" = c(1310149, 40163554),
              "acetaminophen" = c(1125315, 1127078, 1127433, 40229134, 40231925, 40162522, 19133768),
              "morphine" = c(1110410, 35605858, 40169988))

# Instantiate cohorts with CohortConstructor
cdm$my_cohort <- conceptCohort(cdm = cdm,
                               conceptSet = codes, 
                               exit = "event_end_date",
                               overlap = "merge",
                               name = "my_cohort")

Run PhenotypeDiagnostics

Now we will proceed to run phenotypeDiagnotics(). This function will run the following analyses:

• Database diagnostics: This includes information about the size of the data, the time period covered, the number of people in the data, and other meta-data of the CDM object. See Database diagnostics vignette for more details.
• Codelist diagnostics: This includes information on the concepts included in our cohorts’ codelist. See Codelist diagnostics vignette for further details.
• Cohort diagnostics: This summarises the characteristics of our cohorts, as well as comparing them to age and sex matched controls from the database.. See Cohort diagnostics vignette for further details.
• Population diagnostics: Calculates the frequency of our study cohorts in the database in terms of their incidence rates and prevalence. See Population diagnostics vignette for further details.

We can specify which analysis we want to perform by setting to TRUE or FALSE each one of the corresponding arguments:

result <- phenotypeDiagnostics(
  cohort = cdm$my_cohort, 
  diagnostics = c("databaseDiagnostics", "codelistDiagnostics", 
                  "cohortDiagnostics", "populationDiagnostics"),
  cohortSample = 20000,
  matchedSample = 1000,
  populationSample = 1e+06,
  populationDateRange = as.Date(c(NA, NA))
  )
result |> glimpse()

Notice that we have three additional arguments:

• populationSample: It allows to specify a number of people that randomly will be extracted from the CDM to perform the Population diagnostics analysis. If NULL, all the participants in the CDM will be included. It helps to reduce the computational time. This is particularly useful when outcomes of interest are relatively common, but when they are rarer we may wish to maximise statistical power and calculate estimates for the dataset as a whole in which case we would set this argument to NULL.
• populationDateRange: We can use it to specify the time period when we want to perform our Population diagnostics analysis.
• cohortSample: This argument will subset a random sample of people from our cohort and performs the cohortDiagnostics on this sample (notice that the attrition and the cohort counts will be calculated from the original cohorts, not the sampled ones). If the sample specified is bigger than the number of individuals in the cohort, no sampling will be performed. We recommend to use this option when there are cohorts bigger than 20,000 individuals.
• matchedSample: Similar to populationSample, this arguments subsets a random sample of people from our cohort and performs the matched analysis on this sample. If the sample specified is bigger than the size of the cohort, no sampling will be performed. If we have specified a cohortSample, the sampling will be performed on top of the sampled cohorts. If we do not want to create matched cohorts, we can define matchedSample = 0.

Save the results

To save the results, we can use exportSummarisedResult function from omopgenerics R Package:

exportSummarisedResult(result, path = here::here(), minCellCount = 5)

Visualisation of the results

Once we get our Phenotype diagnostics result, we can use shinyDiagnostics to easily create a shiny app and visualise our results:

shinyDiagnostics(result,
                 directory = tempdir(),
                 minCellCount = 5, 
                 open = TRUE)

Notice that we have specified the minimum number of counts (minCellCount) for suppression to be shown in the shiny app, and also that we want the shiny to be launched in a new R session (open). You can see the shiny app generated for this example in here.See Shiny diagnostics vignette for a full explanation of the shiny app.