Hugging Face's logo

Datasets:
SidneyBissoli
/
sinasc

Tasks:
Tabular Classification
Languages:
Portuguese
Size:
10M<n<100M
Tags:
health
brazil
public-health
parquet
datasus
sinasc
License:
Dataset card Files
xet
Community

The dataset is currently empty. Upload or create new data files. Then, you will be able to explore them in the Dataset Viewer.

SINASC — Live Birth Records (Brazil, 1994–2022)

Individual-level live birth records from Brazil's Live Birth Information System (SINASC), covering 29 years of vital statistics data. Each record corresponds to one Declaration of Live Birth (DNV) and includes maternal, newborn, and delivery characteristics. Converted from legacy .dbc files to Apache Parquet.

Part of the healthbr-data project — open redistribution of Brazilian public health data.

Summary

Item Detail
Official source DATASUS FTP / Ministry of Health
Temporal coverage 1994–2022
Geographic coverage All 27 Brazilian states (by municipality of occurrence)
Granularity Individual: one row per live birth
Volume 85M+ records (783 .dbc files processed)
Format Apache Parquet, partitioned by ano/uf
Data types All fields stored as string (preserves original format)
Update frequency Annual (source publishes ~12–15 months after reference year)
License CC-BY 4.0

Resumo em português

SINASC — Registros de Nascidos Vivos (Brasil, 1994–2022)

Microdados individuais de nascidos vivos do Sistema de Informações sobre Nascidos Vivos (SINASC), cobrindo 29 anos de estatísticas vitais. Cada registro corresponde a uma Declaração de Nascido Vivo (DNV) e inclui características maternas, do recém-nascido e do parto. Convertidos de arquivos .dbc legados para Apache Parquet.

Item Detalhe
Fonte oficial FTP DATASUS / Ministério da Saúde
Cobertura temporal 1994–2022
Cobertura geográfica Todos os 27 estados brasileiros (por município de ocorrência)
Granularidade Individual: uma linha por nascido vivo
Volume 85M+ registros (783 arquivos .dbc processados)
Formato Apache Parquet, particionado por ano/uf
Atualização Anual (fonte publica ~12–15 meses após o ano de referência)

Para documentação completa em português, consulte o repositório do projeto.

Data access

Data is hosted on Cloudflare R2 and accessed via S3-compatible API. The credentials below are read-only and intended for public use.

R (Arrow) 

library(arrow)
library(dplyr)

Sys.setenv(
  AWS_ENDPOINT_URL      = "https://5c499208eebced4e34bd98ffa204f2fb.r2.cloudflarestorage.com",
  AWS_ACCESS_KEY_ID     = "28c72d4b3e1140fa468e367ae472b522",
  AWS_SECRET_ACCESS_KEY = "2937b2106736e2ba64e24e92f2be4e6c312bba3355586e41ce634b14c1482951",
  AWS_DEFAULT_REGION    = "auto"
)

# Open a single year (fastest)
ds <- open_dataset("s3://healthbr-data/sinasc/ano=2022/", format = "parquet")

# Example: live births in São Paulo, 2022, by sex
ds |>
  filter(uf == "SP") |>
  count(SEXO) |>
  collect()

Important: Point to year partitions (ano=YYYY/), not to the dataset root. The root contains README.md and manifest.json, which Arrow cannot read as Parquet files.

Python (PyArrow) 

import pyarrow.dataset as pds
import pyarrow.fs as fs

s3 = fs.S3FileSystem(
    endpoint_override="https://5c499208eebced4e34bd98ffa204f2fb.r2.cloudflarestorage.com",
    access_key="28c72d4b3e1140fa468e367ae472b522",
    secret_key="2937b2106736e2ba64e24e92f2be4e6c312bba3355586e41ce634b14c1482951",
    region="auto"
)

# Single year
dataset = pds.dataset(
    "healthbr-data/sinasc/ano=2022/",
    filesystem=s3,
    format="parquet",
    partitioning="hive"
)
# Example: live births in São Paulo, 2022
table = dataset.to_table(filter=(pds.field("uf") == "SP"))
print(table.to_pandas().head())

Note: These credentials are read-only and safe to use in scripts. The bucket does not allow anonymous S3 access — credentials are required. Point to year partitions, not the dataset root (see note above).

File structure 

s3://healthbr-data/sinasc/
  README.md
  manifest.json
  ano=1994/
    uf=AC/
      part-0.parquet
    uf=AL/
      part-0.parquet
    ...
  ano=1995/
    ...
  ...
  ano=2022/
    uf=AC/
      part-0.parquet
    ...
    uf=TO/
      part-0.parquet

Historical schemas

The Declaration of Live Birth (DNV) form underwent multiple revisions since 1994. The dataset preserves 12 distinct schemas:

Schema Period Columns Key characteristics
1 1994–1995 27 Distinct naming convention; mapped to modern names
2 1996–1998 21 First modern naming convention
3 1999–2000 20
4 2001 23
5 2002–2005 26
6 2006–2009 29
7 2010 55 Major form expansion
8 2011 56
9 2012 56
10 2013 59
11 2014–2017 61
12 2018–2022 61 Same as schema 11 (case normalization only)

The number of columns varies by year and state. Columns not present in a given era will be absent from that partition's Parquet file.

Schema (modern era, 2018–2022)

Key variables in the most recent schema (61 columns):

Variable Description
NUMERODN Declaration of Live Birth ID
CODESTAB Health facility code
CODMUNNASC Municipality of occurrence (IBGE 6-digit)
LOCNASC Place of birth (1=hospital, 2=other health facility, 3=home, 4=other, 5=indigenous village)
IDADEMAE Mother's age (years)
ESCMAE Mother's education level
CODMUNRES Mother's municipality of residence (IBGE 6-digit)
GESTACAO Gestational age (weeks, coded)
GRAVIDEZ Type of pregnancy (1=single, 2=twin, 3=triplet+)
PARTO Type of delivery (1=vaginal, 2=cesarean)
CONSULTAS Number of prenatal visits (coded)
DTNASC Date of birth (DDMMYYYY)
SEXO Sex (1=male, 2=female, 0=undetermined)
PESO Birth weight (grams)
APGAR1 Apgar score at 1 minute
APGAR5 Apgar score at 5 minutes
RACACOR Race/color (1=white, 2=black, 3=asian, 4=brown, 5=indigenous)
IDANOMAL Congenital anomaly detected (1=yes, 2=no, 9=unknown)
CODANOMAL ICD-10 code of congenital anomaly
DTCADASTRO Registration date
DTNASCMAE Mother's date of birth
QTDFILVIVO Mother's number of living children
QTDFILMORT Mother's number of deceased children

For the complete variable list across all eras, see the exploration document.

Schema unification (1994–1995)

Records from 1994–1995 use a different naming convention. The pipeline maps 20 fields to their modern equivalents (e.g., CODIGO → NUMERODN, MUNI_OCOR → CODMUNNASC). Date fields are converted from YYYYMMDD to DDMMYYYY for consistency. Six local fields without national equivalents are preserved as extra columns (CARTORIO, DATA_CART, AREA, BAIRRO_MAE, CRS_MAE, CRS_OCOR). Four internal control fields are discarded (ETNIA, FIL_ABORT, NUMEXPORT, CRITICA — all nearly 100% null).

Source and processing

Original source: 843 .dbc files from the DATASUS FTP server, covering two directories: NOV/DNRES/ (1996–2022, 734 files) and ANT/DNRES/ (1994–1995, 109 files). Of these, 783 were successfully processed (remaining were unavailable on the FTP server at processing time).

Processing: .dbc → R (read.dbc::read.dbc()) → schema unification (1994–1995 name mapping + date format conversion) → all fields cast to character → Parquet (arrow::write_dataset()) → upload to R2 (rclone). No value transformations are applied — field values are published exactly as provided by the Ministry of Health.

Bootstrap: 783 files, 85,033,402 records, 0 errors, 117 minutes on Hetzner CX21.

Known limitations

  1. Government data, not ours. Values are preserved exactly as in the original .dbc files, including any inconsistencies or missing data.
  2. Twelve historical schemas. The number of columns varies by year (21–61). Queries spanning multiple eras must handle missing columns.
  3. All fields are strings. Numeric fields (weight, Apgar scores, age) and dates (DDMMYYYY format) must be parsed by the user.
  4. OpenDATASUS inaccessible. The S3 endpoint returned HTTP 403 at processing time (March 2026). FTP is the only viable source.
  5. Coverage ends at 2022. The most recent year available on the FTP server at processing time. Newer years will be added as they become available.
  6. By municipality of occurrence. Records are classified by where the birth occurred, not where the mother resides. Use CODMUNRES for residence-based analysis.
  7. 60 files unavailable. 60 of the 843 expected files (grid of 27 UFs × all years) were not found on the FTP server. These are mostly from 1994–1995 for states with smaller populations.

Related datasets

Dataset Period Records Link
SI-PNI Microdados (vaccination) 2020–present 736M+ sipni-microdados
SI-PNI COVID (vaccination) 2021–present 608M+ sipni-covid
SI-PNI Agregados — Doses 1994–2019 84M+ sipni-agregados-doses
SI-PNI Agregados — Cobertura 1994–2019 2.8M+ sipni-agregados-cobertura
SI-PNI Dicionários Static 263 rows sipni-dicionarios

Citation 

@misc{healthbrdata,
  author = {Sidney da Silva Bissoli},
  title  = {healthbr-data: Redistribution of Brazilian Public Health Data},
  year   = {2026},
  url    = {https://huggingface.co/datasets/SidneyBissoli/sinasc},
  note   = {Original source: Ministry of Health / DATASUS}
}

Contact

Last updated: 2026-03-08
Downloads last month
5