Getting started with tatooheene

In tatooheene: Technology Appraisal Toolbox for Health Economic Evaluations in the Netherlands

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
library(tatooheene)
library(dplyr)

Overview

tatooheene stands for Technology Appraisal Toolbox for Health Economic Evaluations in the Netherlands and helps you apply the Dutch costing manual according to the Dutch economic evaluation (EE) guideline, consistently in R. It provides:

• Programmatic access to reference values from the Dutch costing manual (e.g., unit costs, informal care, travel assumptions).
• Standard functions to calculate for example productivity loss via the friction cost method, purchasing power adjustments (PPP/OECD-style converters), and many more.

Functions are designed to slot into R-based cost(-effectiveness) workflows and are tidyverse-friendly.

Installation

# CRAN
install.packages("tatooheene")

Loading

# Load the package
library(tatooheene)

# We recommend using the tidyverse style
library(tidyverse)

Using tatooheene: a short tour

Inflation-correct historical costs (CBS CPI)

Suppose your costing study reported 2019 EUR costs. Bring them to 2024 EUR using CPI indices.

# Look up CPI indices with function nl_price_index()
idx19_23 <- nl_price_index(
  start_year = 2019,
  end_year = 2023,
  output = "factor"  # see ?nl_price_index for available outputs
)

idx19_23

# Example tibble
costs <- tibble::tibble(item = c("GP consult", "MRI"), cost_2019 = c(34.5, 285))

# Adjust costs to 2024 EUR using the CPI index
costs |>
  mutate(cost_2024 = cost_2019 * idx19_23)

Productivity losses via the friction cost method

Estimate productivity costs from sick leave spells using the Dutch friction cost approach.

First, we need to know the actual friction period for the analysis year 2023. This we can retrieve with the friction_period() function (see ?friction_period for details).

# Get the friction period for 2023
v_5yr_mean_friction <- tatooheene::friction_period(
  year = 2023, # Year of interest
  units = "days", # We are interested in the number of days
  avg = "5yr", # Use the 5-year average friction period as stated in the costing manual,
  output = "value" # Return a single value
  )

Based on this, we now know that the friction period in 2023 is r v_5yr_mean_friction days. We can now calculate the productivity costs for sick leave spells.

# Example sick-leave spells (days) and reference prices (EUR/day)
df_spells <- tibble::tibble(
  id = 1:3,
  sick_days = c(140, 122, 30)) 

# Get reference prices for productivity losses (paid work)
# These are in table df_rp_prod

# Get reference price for paid work in 2023
p_ref_prod_paid_2023 <- df_ref_prices |> 
  filter(short_var == "prodloss_paid_hour") %>% 
  pull(`2023`)

# Calculate productivity costs using the friction cost method
df_spells %>% 
  mutate(
    # Apply the friction cost method
    sick_days_friction = ifelse(
    sick_days > v_5yr_mean_friction, v_5yr_mean_friction, sick_days),
    # Calculate productivity costs
    prod_cost = sick_days_friction * p_ref_prod_paid_2023)

More examples of productivity costs can be found in the friction_period vignette.

tatooheene documentation built on Dec. 15, 2025, 5:06 p.m.