Cancer RADAR - Getting Started

Before starting

This document provides a walkthrough of section 4.6 Step 3 Data processing from the cancer RADAR protocol. For additional information, you can visit the cancerradarr package front page : https://cancerradar.gitlab.io/cancerradarr/

Installing cancerradarr Package

The cancerradarr package is hosted on gitlab.com. You can install the package using the install_gitlab function from the remotes package.

# Check if the remotes package is installed and install it if not.
if(!('remotes' %in% rownames(installed.packages()))) install.packages('remotes', dependencies = TRUE)

# Install the cancerradarr package
remotes::install_gitlab("cancerradar/cancerradarr", dependencies = TRUE)

Creating the Cancer RADAR Input File

The first step is to create a standardized input file that cancer registries can complete.

# Load the cancerradarr package
library(cancerradarr)

# Create the working directory
# Please add here the working directory from which you want to work
# Note that the Microsoft environment uses the backslash (\) as a path 
# separator. For R to interpret the path correctly, forward slashes (/) 
# need to be replaced with backslashes (\).
# The use of normalizePath should insure this path is correctly written
workdir <- normalizePath('~/Documents/241107_cancerRADAR_getting_started')

dir.create(workdir, showWarnings = FALSE, recursive = TRUE)

# Create the Cancer RADAR input file
create_registry_input_file(file.path(workdir, 'cancerRADAR_input.xlsx'))
openxlsx::openXL(file.path(workdir, 'cancerRADAR_input.xlsx'))

The file cancerRADAR_input.xlsx has been created and can be filled by cancer registries.

Let’s have a quick look at this file:

# Open the input file in Excel
openxlsx::openXL(file.path(workdir, 'cancerRADAR_input.xlsx'))

The README sheet provides a synthetic version of the spreadsheet content and some brief guidance on how to complete the data. More detailed information can be found in the Cancer RADAR protocol.

Note: the input data excel spreadsheet could also be downloaded here.

Entering Data in the Input File

  1. Open the empty input file and enter the data in the different sheets (for now, only incidence data is requested).
  2. You will see different empty sheets. The first sheet provides an explanation of the data that needs to be entered. In the second sheet you can indicate for which country and time period the data is entered.
  3. Save the file under a different name, providing the period and version date (e.g., cancerRADAR_input_completed_xx_xx_vsxxxx2024.xlsx).

For a given period of time, for each cancer type, sex, and 5-year age group, cancer registries can collect the number of cancer cases stratified by country of birth and fill it in the created input file. If available, the population at risk (as person-years) can also be completed in the same input file.

For the purpose of this exercise, we will use an input file with dummy data that comes with the cancerradarr package.

# Get an example of a filled input file
file.copy(
  file.path(path.package('cancerradarr'), 'extdata/ex_cancerRADAR_input_filled.xlsx'),
  file.path(workdir, 'ex_cancerRADAR_input_filled.xlsx'),
  overwrite = TRUE
)

# Open the example filled input file in Excel
openxlsx::openXL(file.path(workdir, 'ex_cancerRADAR_input_filled.xlsx'))

Computing Cancer RADAR Summary Statistics

The create_canradar_summary_file function will produce a ready-to-share Excel spreadsheet with all the relevant summary statistics.

Here is an example with the example input file we just obtained. The filename.in should be adapted to point to the input file filled with your local cancer registry data when completed. If you do not want to share the data by country of birth, you can set the argument include.by.cob.stat = FALSE.

# Compute summary statistics including country of birth aggregation
create_canradar_summary_file(
  filename.in = file.path(workdir, 'ex_cancerRADAR_input_filled.xlsx'),
  filename.out = file.path(workdir, 'ex_cancerRADAR_output_with_cob_01.xlsx'),
  ncan.min = 5,
  include.by.cob.stat = TRUE
)

# Or without country of birth aggregation
create_canradar_summary_file(
  filename.in = file.path(workdir, 'ex_cancerRADAR_input_filled.xlsx'),
  filename.out = file.path(workdir, 'ex_cancerRADAR_output_without_cob_01.xlsx'),
  ncan.min = 5,
  include.by.cob.stat = FALSE
)

This function might take a few minutes to complete.

During the processing, it ensures that no cells with less than 5 cancer cases are generated. In this way the output file will be in a ready-to-share format.

You should now have an output file (Excel spreadsheet) containing all the Cancer RADAR summary statistics as an output. The README tab of the spreadsheet contains the data dictionary of the output file.

A separate data dictionary is also available and can be opened with the following command:

cancerradarr::open_canradar_dictionary()

Visualizing the Summary Statistics

To help cancer registries perform data quality checks and gain an initial overview of the cancer registry data, an automatic report can be generated using the create_static_report function.

# Create a report of the entered data
create_static_report(file.path(workdir, 'ex_cancerRADAR_output_with_cob_01.xlsx'))

# The report can be opened with the following command or by double-clicking on the filename
browseURL(file.path(workdir, 'ex_cancerRADAR_output_with_cob_01.html'))

For those who want to explore the data further, a dynamic report is also available via the run_dynamic_report function.

Note: Since this will open a Shiny app on your computer, the R console will stay busy until close the window or click the stop button in the R console.

# Create a dynamic report of the entered data
run_dynamic_report(file.path(workdir, 'ex_cancerRADAR_output_with_cob_01.xlsx'))

Sending Cancer RADAR Summary Statistics

When you have completed all steps with your local cancer registry data, please send the created file through the RedCap link created for this project. You can email to cancer.radar@iarc.who.int to receive a personal link.