# PhysioNet/CinC Challenge 2019: Early Prediction of Sepsis from Clinical Data

## Facts

- Provider: Matthew Reyna, Chris Josef, Russell Jeter, Supreeth Shashikumar, Benjamin Moody, M. Brandon Westover, Ashish Sharma, Shamim Nemati, Gari D. Clifford
- License: CC-BY-4.0 (https://physionet.org/content/challenge-2019/1.0.0/)
- Stored at: Amazon S3, s3://physionet-open/challenge-2019 (us-east-1)
- Access: public, readable without an account
- DOI: https://doi.org/10.13026/v64v-d857
- Category: Physiology
- Source: https://physionet.org/content/challenge-2019/1.0.0/
- Page: https://scigantic.com/archive/aafa36d8-597b-5be6-a46e-cc64b48e94c9
- Keywords: Physiology, time series, benchmark, AWS Open Data

## About

The 2019 PhysioNet/Computing in Cardiology Challenge asked for causal algorithms that predict sepsis six hours before clinical recognition using routinely collected ICU data, scored by a clinical utility function that rewards early and penalizes late or false predictions. Data come from ICU patients in three separate hospital systems; two were released publicly and the third was held out for scoring. The public training database has 40,336 subjects in two parts, training set A (20,336 subjects) and training set B (20,000 subjects), with one pipe-separated .psv file per subject (for example p00101.psv). Each row is one hour of ICU stay and the columns are 40 time-dependent variables (vital signs such as HR, O2Sat and Temp, laboratory values, and demographics such as age and gender, and time markers HospAdmTime and ICULOS) plus the SepsisLabel defined with Sepsis-3 criteria. Measured 2026-11-13: training/ holds training_setA/ and training_setB/, alongside hackathon/, sources/, the Critical Care Medicine manuscript PDF and utility function diagrams. Load files with pandas.read_csv(sep='|'). A standard benchmark for sepsis early warning, clinical time series and ICU prediction. Released under CC BY 4.0; cite the challenge paper and the standard PhysioNet reference (Goldberger et al., 2000) when using the data.

## Load the data

```python
import boto3
from botocore import UNSIGNED
from botocore.client import Config
s3 = boto3.client('s3', region_name='us-east-1', config=Config(signature_version=UNSIGNED))
s3.list_objects_v2(Bucket='physionet-open', Prefix='challenge-2019/', Delimiter='/')
```

Lists the files with anonymous access. Runs on any machine, with no cloud account.

## Data structure

Example file: `challenge-2019/1.0.0/hackathon/sources/SailOcean.zip`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/aafa36d8-597b-5be6-a46e-cc64b48e94c9
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/aafa36d8-597b-5be6-a46e-cc64b48e94c9
- Over MCP: `get_schema_card` with id `aafa36d8-597b-5be6-a46e-cc64b48e94c9` at https://api.scigantic.com/mcp

## Related topics

- [Benchmark datasets](https://scigantic.com/topics/benchmark)
- [Physiology datasets](https://scigantic.com/topics/physiology)
- [Time series datasets](https://scigantic.com/topics/time-series)
