datasusr provides fast, in-memory reading of DATASUS .dbc files and a complete workflow for discovering, downloading, caching, and reading Brazilian public health data from the DATASUS FTP.

Looking for a broader toolkit? If your workflow goes beyond the DATASUS FTP — e.g. you also need IBGE surveys (VIGITEL, PNS, PNAD-C, POF, Censo), SISAB primary-care indicators, ANS, ANVISA, or out-of-the-box variable dictionaries and value labels — healthbR is the more complete and currently more active package, and is the recommended first choice in many cases. datasusr focuses on being a small, fast, dependency-light reader for raw DBC files plus a catalog and FTP layer. See the Comparison article for the full breakdown.

❕️ Disclaimer
This package is an independent, community-maintained tool that accesses publicly available data files from the DATASUS FTP server (ftp://ftp.datasus.gov.br). It is not affiliated with the Brazilian Ministry of Health, DATASUS, or any government entity. To maintain consistency with R package development standards, all functions use English names (e.g. datasus_fetch(), datasus_sources()). However, because the source data is produced by Brazilian government systems, parameter values use official DATASUS codes in Portuguese (e.g. source = "SIHSUS", uf = "PE"), and column names in the returned tibbles reflect the original DBC/DBF field names (e.g. uf_zi, ano_cmpt, munic_res, val_tot). For reference on the original data layouts and field descriptions, use datasus_docs_url() or see the official DATASUS documentation.

Installation

# Install from GitHub
# install.packages("remotes")
remotes::install_github("StrategicProjects/datasusr")

Quick start

library(datasusr)
# One-step: list, download, and read SIH data for Pernambuco
df <- datasus_fetch(
  source    = "SIHSUS",
  file_type = "RD",
  year      = 2024,
  month     = 1,
  uf        = "PE"
)
df

Step-by-step workflow

For more control, use the individual functions:

library(datasusr)
# 1. Explore the catalog
datasus_sources()
datasus_file_types(source = "SIHSUS")
# 2. List available files on the FTP
files <- datasus_list_files(
  source    = "SIHSUS",
  file_type = "RD",
  year      = 2024,
  month     = 1:3,
  uf        = c("PE", "PB")
)
# 3. Download (with automatic caching)
downloads <- datasus_download(files, use_cache = TRUE)
# 4. Read a DBC file into a tibble
x <- read_datasus_dbc(downloads$local_file[[1]])
# 5. Read with column selection and type control
x <- read_datasus_dbc(
  downloads$local_file[[1]],
  select     = c("uf_zi", "ano_cmpt", "dt_inter", "val_tot"),
  col_types  = c(dt_inter = "date", val_tot = "double"),
  parse_dates = TRUE
)

Cache management

Downloads are cached by default so repeated runs do not hit the DATASUS FTP:

datasus_cache_info()
datasus_cache_list()
# Prune old files
datasus_cache_prune(older_than_days = 90)
# Or clear everything
datasus_cache_clear()

You can configure the cache directory via the DATASUSR_CACHE_DIR environment variable, the datasusr.cache_dir R option, or the cache_dir argument.

Data sources

Main functions

Progress messages

All functions emit cli progress messages by default. Suppress them with verbose = FALSE.

License

MIT