Diabetes-130-Hospitals: 10 Years of U.S. Hospital Admissions for Diabetes

The Diabetes-130-Hospitals dataset represents 10 years of clinical care at 130 U.S. hospitals and delivery networks from 1999 to 2008. Each record is a hospital admission for a patient diagnosed with diabetes, featuring demographics, diagnoses, medications, payer information, and readmission status. This version was processed by the Fairlearn team for a 2021 tutorial, binarizing the target to indicate readmission within 30 days.

Use Cases

• Predicting patient readmission risk based on demographic and clinical encounter features.
• Analyzing disparities in healthcare outcomes using demographic and payer information.
• Training fairness-aware models to audit for bias in readmission predictions.
• Studying patterns in diabetic medication usage and diagnoses over a 10-year period.

Strengths

• Contains 10 years of clinical records from 130 U.S. hospitals.
• Includes a specific binary target for readmission within 30 days, processed for fairness analysis.
• Original collection involved multiple researchers and institutions, documented in a 2014 publication.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Last update date is unknown; freshness unverified.

Provenance

Source

Beata Strack, Jonathan P. DeShazo, Chris Gennings, Juan L. Olmo, Sebastian Ventura, Krzysztof J. Cios, and John N. Clore (2014).

Collection Method

Collected from hospital admission records for patients diagnosed with diabetes.

Time Range

1999 to 2008

Geography

130 U.S. hospitals and delivery networks