healthbR

Overview

healthbR provides easy access to Brazilian public health data directly from R: the DATASUS microdata systems (SIM mortality, SINASC live births, SIH hospital admissions, SIA outpatient production, SINAN notifiable diseases, CNES facilities, SI-PNI vaccination), the IBGE and Ministry of Health surveys (VIGITEL, PNS, PNAD Continua, POF, Censo) and the regulatory agencies (ANS, ANVISA). The package downloads, caches, and processes data from official sources, returning clean, analysis-ready tibbles following tidyverse conventions.

If you already use microdatasus or read.dbc (or PySUS in Python) for DATASUS, see healthbR, microdatasus and read.dbc -- a measured comparison of the three on the same files: what each returns, how fast, the same tasks side by side, and the Parquet mirror read with no package at all.

Surveys (IBGE / Ministry of Health)

Module	Description	Years
VIGITEL	Surveillance of Risk Factors for Chronic Diseases by Telephone Survey	2006--2024
PNS	National Health Survey (microdata + SIDRA API)	2013, 2019
PNAD Continua	Continuous National Household Sample Survey	2012--2024
POF	Household Budget Survey (food security, consumption, anthropometry)	2002--2018
Censo	Population denominators via SIDRA API	1970--2022

DATASUS (Ministry of Health FTP)

Module	Description	Granularity	Years
SIM	Mortality Information System (deaths)	Annual/UF	1996--2024
SINASC	Live Birth Information System	Annual/UF	1996--2024
SIH	Hospital Information System (admissions)	Monthly/UF	1992--2026
SIA	Outpatient Information System (13 file types)	Monthly/UF	2008--2026
SINAN	Notifiable Diseases Surveillance (31 diseases)	Annual/UF	2007--2026
CNES	National Health Facility Registry (13 file types)	Monthly/UF	2005--2026
SI-PNI	National Immunization Program (aggregates 1994--2019, microdata 2020+; served from the healthbr-data R2 mirror with automatic DATASUS fallback)	Annual or Monthly/UF	1994--2026

DATASUS modules download .dbc files (compressed DBF) and decompress them internally using vendored C code -- no external dependencies required. SIH and SI-PNI read by default from the healthbr-data Parquet mirror (byte-identical to the Ministry's files, with per-file provenance: source URL, MD5, timestamp) and fall back to the FTP.

Primary Care & Regulatory Agencies

Module	Source	Description	Years
SISAB	Ministry of Health REST API	Primary Care coverage indicators (APS, oral health, community agents, PNS)	2007--present
ANS	ANS Open Data Portal	Supplementary health beneficiaries, consumer complaints, financial statements	2007--present
ANVISA	ANVISA Open Data Portal	Product registrations, pharmacovigilance, hemovigilance, technovigilance, SNGPC	snapshot + 2014--present

healthbR, microdatasus and read.dbc

Three R routes read DATASUS microdata, at different layers: read.dbc reads a .dbc file you downloaded, microdatasus downloads and labels, healthbR downloads, caches and types. Measured on 2026-10-14 (healthbR 0.4.0.9000, microdatasus 3.0.0, read.dbc 1.2.0) and detailed in the vignette healthbR, microdatasus and read.dbc:

	healthbR	microdatasus	read.dbc
DATASUS systems	SIM, SINASC, SIH, SIA, SINAN (31 diseases), CNES, SI-PNI	SIM, SINASC, SIH, SIA, SINAN (8 diseases), CNES	any .dbc you download
Beyond DATASUS	VIGITEL, PNS, PNAD Continua, POF, Censo, SISAB, ANS, ANVISA	SIGTAP, CADGER and municipality tables	--
.dbc reader	vendored C (blast)	the same code, adapted from healthbR in 3.0.0	its own C; also writes .dbc
Codes	kept as codes; typed (Date, integer, double); labels via *_dictionary()	replaced by labels by process_*(); all columns character	as the DBF declares them
Source	healthbr-data Parquet mirror with provenance (SIH, SI-PNI), FTP otherwise	DATASUS FTP	--
Lazy / cache	lazy = TRUE (arrow or duckdb), local Parquet cache	none	none
On CRAN	yes	yes	archived 2025-12-14
SIH-SP 2023-01, from nothing	7.7 s (mirror)	18.0 s	9.7 s (with the download)

Same rows and records in all three on every file compared. Each fits a different workflow; the vignette says which.

Installation

From CRAN (0.2.0):

install.packages("healthbR")

The development version, with the SIH mirror and the newest year coverage, from GitHub:

# install.packages("pak")
pak::pak("SidneyBissoli/healthbR")

Quick start

library(healthbR)

# see all available data sources
list_sources()

DATASUS modules

All DATASUS modules follow a consistent API: *_years(), *_info(), *_variables(), *_dictionary(), *_data(), *_cache_status(), *_clear_cache().

# mortality data -- deaths in Acre, 2022
obitos <- sim_data(year = 2022, uf = "AC")

# filter by cause of death (CID-10 prefix)
obitos_cardio <- sim_data(year = 2022, uf = "AC", cause = "I")

# live births in Acre, 2022
nascimentos <- sinasc_data(year = 2022, uf = "AC")

# hospital admissions in Acre, January 2022
internacoes <- sih_data(year = 2022, month = 1, uf = "AC")

# filter by diagnosis (CID-10 prefix)
intern_resp <- sih_data(year = 2022, month = 1, uf = "AC", diagnosis = "J")

# outpatient production in Acre, January 2022
ambulatorial <- sia_data(year = 2022, month = 1, uf = "AC")

# different file type (e.g., high-cost medications)
medicamentos <- sia_data(year = 2022, month = 1, uf = "AC", type = "AM")

Additional DATASUS modules

# disease notifications -- dengue, 2022
dengue <- sinan_data(year = 2022, disease = "DENG")

# health facilities in Acre, January 2023
cnes <- cnes_data(year = 2023, month = 1, uf = "AC")

# vaccination data -- Acre (R2 mirror by default, DATASUS fallback)
vacinas <- sipni_data(year = 2019, uf = "AC")
micro <- sipni_data(year = 2024, uf = "AC", month = 1)

Survey modules

# VIGITEL telephone survey
vigitel <- vigitel_data(year = 2024)

# PNS national health survey
pns <- pns_data(year = 2019)

# PNAD Continua
pnadc <- pnadc_data(year = 2023, quarter = 1)

# POF household budget survey
pof <- pof_data(year = 2018, register = "morador")

# Census population
pop <- censo_populacao(year = 2022, territorial_level = "state")

Primary care & regulatory agencies

# SISAB -- primary care coverage by state, January 2024
sisab <- sisab_data(year = 2024, month = 1)

# ANS -- health plan beneficiaries in Acre, December 2023
ans <- ans_data(year = 2023, month = 12, uf = "AC")

# ANVISA -- registered medicines
med <- anvisa_data(type = "medicines")

Explore variables and dictionaries

# list variables for any module
sim_variables()
sia_variables(search = "sexo")
sinan_diseases(search = "dengue")

# data dictionary with category labels
sim_dictionary("SEXO")
sia_dictionary("PA_RACACOR")

Parallel downloads

When downloading multiple files (e.g., several years, months, or states), you can speed up downloads by enabling parallel processing. Install furrr and future, then set a parallel plan before calling any *_data() function:

install.packages(c("furrr", "future"))

library(future)
plan(multisession, workers = 4)

# downloads 4 states in parallel
df <- sih_data(year = 2022, month = 1:6, uf = c("SP", "RJ", "MG", "BA"))

# reset to sequential when done
plan(sequential)

All 12 download-based modules support parallel downloads: SIM, SINASC, SIH, SIA, SINAN, CNES, SI-PNI, SISAB, ANS, ANVISA, PNS, and PNADC.

Caching

All modules cache downloaded data automatically. Install arrow for optimized Parquet caching:

install.packages("arrow")

Each module provides cache management functions:

# check what is cached
sim_cache_status()
sih_cache_status()
sia_cache_status()

# clear cache for a module
sim_clear_cache()

Data sources

All data is downloaded from official Brazilian government repositories (DATASUS, IBGE, ANS, ANVISA):

• VIGITEL: Ministry of Health
• PNS / PNAD Continua / POF / Censo: IBGE
• SIM / SINASC / SIA / SINAN / CNES: DATASUS FTP
• SIH / SI-PNI: healthbr-data mirror on Cloudflare R2 (byte-identical to the Ministry's files, with provenance metadata), falling back to DATASUS FTP / OpenDataSUS
• SISAB: Ministry of Health REST API
• ANS: ANS Open Data Portal
• ANVISA: ANVISA Open Data Portal

Citation

If you use healthbR in your research, please cite it:

citation("healthbR")

Contributing

Please see CONTRIBUTING.md for guidelines on how to contribute to this project.

Code of Conduct

Please note that the healthbR project is released with a Contributor Code of Conduct. By contributing to this project, you agree to abide by its terms.

License

MIT © Sidney da Silva Pereira Bissoli