Introduction

This vignette illustrates how EpiStandard can be used to standardise results from the IncidencePrevalence package, a DARWIN-EU package for estimating incidence and prevalence from data mapped to the OMOP Common Data Model.

Set-up

We first load the relevant packages and then create a mock OMOP-formatted dataset containing a population and an outcome of interest.

library(EpiStandard)
library(IncidencePrevalence)
library(omopgenerics)
library(dplyr)

cdm <- mockIncidencePrevalence(
  sampleSize = 10000,
  outPre = 0.25
)

Estimate Incidence Rates

We use the IncidencePrevalence package to generate a denominator cohort with age and sex stratifications, and then calculate overall outcome incidence and incidence by calendar year for each stratum combination. For more information on how to use this package, refer to its website.

Notice that the results are stored as a summarised_result object.

cdm <- generateDenominatorCohortSet(
  cdm = cdm,
  name = "denominator",
  cohortDateRange = as.Date(c("2008-01-01", "2020-01-01")),
  ageGroup = list(c(0, 19), c(20, 64), c(65, 150), c(0, 150)),
  sex = c("Male", "Female"),
  daysPriorObservation = 0
)

inc <- estimateIncidence(
  cdm = cdm,
  denominatorTable = "denominator",
  outcomeTable = "outcome",
  interval = c("years", "overall"),
  outcomeWashout = 0,
  repeatedEvents = FALSE
)

inc |> glimpse()
#> Rows: 1,049
#> Columns: 13
#> $ result_id        <int> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,…
#> $ cdm_name         <chr> "mock", "mock", "mock", "mock", "mock", "mock", "mock…
#> $ group_name       <chr> "denominator_cohort_name &&& outcome_cohort_name", "d…
#> $ group_level      <chr> "denominator_cohort_8 &&& cohort_1", "denominator_coh…
#> $ strata_name      <chr> "overall", "overall", "overall", "overall", "overall"…
#> $ strata_level     <chr> "overall", "overall", "overall", "overall", "overall"…
#> $ variable_name    <chr> "Denominator", "Outcome", "Denominator", "Denominator…
#> $ variable_level   <chr> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, N…
#> $ estimate_name    <chr> "denominator_count", "outcome_count", "person_days", …
#> $ estimate_type    <chr> "integer", "integer", "numeric", "numeric", "numeric"…
#> $ estimate_value   <chr> "277", "14", "85765", "234.812", "5962.217", "3259.59…
#> $ additional_name  <chr> "incidence_start_date &&& incidence_end_date &&& anal…
#> $ additional_level <chr> "2008-01-01 &&& 2008-12-31 &&& years", "2008-01-01 &&…

To use EpiStandard, we first filter the incidence results to the age groups of interest for standardisation and then convert the results into a format that facilitates further manipulation.

Standardise Incidence Rates

incidenceTidy <- inc |>
  filterSettings(denominator_age_group %in% c("0 to 19", "20 to 64", "65 to 150")) |>
  asIncidenceResult()

incidenceTidy |> glimpse()
#> Rows: 69
#> Columns: 26
#> $ cdm_name                             <chr> "mock", "mock", "mock", "mock", "…
#> $ denominator_cohort_name              <chr> "denominator_cohort_2", "denomina…
#> $ outcome_cohort_name                  <chr> "cohort_1", "cohort_1", "cohort_1…
#> $ incidence_start_date                 <date> 2008-01-01, 2009-01-01, 2010-01-…
#> $ incidence_end_date                   <date> 2008-12-31, 2009-12-31, 2010-12-…
#> $ analysis_interval                    <chr> "years", "years", "years", "years…
#> $ analysis_censor_cohort_name          <chr> "None", "None", "None", "None", "…
#> $ analysis_complete_database_intervals <chr> "TRUE", "TRUE", "TRUE", "TRUE", "…
#> $ analysis_outcome_washout             <chr> "0", "0", "0", "0", "0", "0", "0"…
#> $ analysis_repeated_events             <chr> "FALSE", "FALSE", "FALSE", "FALSE…
#> $ denominator_age_group                <chr> "0 to 19", "0 to 19", "0 to 19", …
#> $ denominator_days_prior_observation   <chr> "0", "0", "0", "0", "0", "0", "0"…
#> $ denominator_end_date                 <date> 2020-01-01, 2020-01-01, 2020-01-…
#> $ denominator_requirements_at_entry    <chr> "FALSE", "FALSE", "FALSE", "FALSE…
#> $ denominator_sex                      <chr> "Female", "Female", "Female", "Fe…
#> $ denominator_start_date               <date> 2008-01-01, 2008-01-01, 2008-01-…
#> $ denominator_target_cohort_name       <chr> "None", "None", "None", "None", "…
#> $ denominator_time_at_risk             <chr> "0 to Inf", "0 to Inf", "0 to Inf…
#> $ denominator_count                    <int> 38, 44, 38, 23, 16, 14, 8, 7, 4, …
#> $ outcome_count                        <int> 0, 1, 1, 0, 1, 0, 0, 0, 0, 3, 3, …
#> $ person_days                          <dbl> 12529, 12754, 10404, 7081, 5510, …
#> $ person_years                         <dbl> 34.303, 34.919, 28.485, 19.387, 1…
#> $ incidence_100000_pys                 <dbl> 0.000, 2863.770, 3510.620, 0.000,…
#> $ incidence_100000_pys_95CI_lower      <dbl> 0.000, 72.504, 88.881, 0.000, 167…
#> $ incidence_100000_pys_95CI_upper      <dbl> 10753.810, 15955.908, 19559.921, …
#> $ result_type                          <chr> "tidy_incidence", "tidy_incidence…

Now that the incidence results are in the correct format, we prepare the reference population. In this example, we use the European Standard Population (esp2013). As this population uses 5-year age bands, we merge these groups to match those used to estimate incidence.

standardPop <- mergeAgeGroups(
  standardPopulation("Europe"),
  newGroups = c("0 to 19", "20 to 64", "65 to 150")
) |>
  rename("denominator_age_group" = "age_group")
standardPop |> glimpse()
#> Rows: 3
#> Columns: 2
#> $ denominator_age_group <chr> "0 to 19", "20 to 64", "65 to 150"
#> $ pop                   <int> 21500, 59000, 19500

Finally, we use the directlyStandardiseRates() function to standardise the incidence results to the reference population. We specify the columns containing the event counts, person-years, and population weights, as well as the column identifying age groups. Since we want to standardise within each outcome, sex, and calendar-time stratum, we use the strata argument as follows:

standardInc <- directlyStandardiseRates(
  data = incidenceTidy,
  refdata = standardPop,
  event = "outcome_count",
  denominator = "person_years",
  age = "denominator_age_group",
  pop = "pop",
  strata = c("incidence_start_date", "denominator_sex", "analysis_interval", "outcome_cohort_name")
)
standardInc |> glimpse()
#> Rows: 22
#> Columns: 12
#> $ incidence_start_date         <date> 2009-01-01, 2010-01-01, 2012-01-01, 2008…
#> $ denominator_sex              <chr> "Female", "Female", "Female", "Female", "…
#> $ analysis_interval            <chr> "years", "years", "years", "overall", "ye…
#> $ outcome_cohort_name          <chr> "cohort_1", "cohort_1", "cohort_1", "coho…
#> $ outcome_count                <int> 10, 14, 4, 51, 13, 14, 9, 13, 4, 5, 71, 1…
#> $ person_years                 <dbl> 233.057, 209.706, 132.954, 1298.001, 272.…
#> $ crude_rate                   <dbl> 4290.7958, 6676.0131, 3008.5594, 3929.118…
#> $ crude_rate_95CI_lower        <dbl> 1631.3819, 3178.9678, 60.2254, 2850.7726,…
#> $ crude_rate_95CI_upper        <dbl> 6950.210, 10173.058, 5956.893, 5007.465, …
#> $ standardised_rate            <dbl> 4276.532, 6868.422, 3078.205, 3860.587, 5…
#> $ standardised_rate_95CI_lower <dbl> 1527.5585, 3158.1231, -362.0191, 2744.249…
#> $ standardised_rate_95CI_upper <dbl> 7025.506, 10578.720, 6518.429, 4976.924, …

Standardise OMOP results

Introduction

Set-up

Estimate Incidence Rates

Standardise Incidence Rates