Critical Care Patient Data Derived from MIMIC-III

A dataset derived from the MIMIC-III critical care database, containing patient health records from intensive care units. It was created by Leo-TX and last updated in March 2026. The specific temporal coverage and row count are not provided.

Use Cases

• Predict patient mortality or length of stay using features like vital signs and lab results.
• Analyze treatment patterns and medication administration records across patient cohorts.
• Identify risk factors for specific conditions by modeling relationships between demographic data and clinical events.
• Benchmark new clinical prediction algorithms against established MIMIC-III derived models.

Strengths

• Based on the established and widely-used MIMIC-III critical care database.
• Contains real-world patient data from intensive care unit settings.

Limitations

• The specific sample size, column count, and data completeness are unknown.
• Inherits the geographic and demographic biases present in the original MIMIC-III source data.
• The exact temporal coverage and update methodology are not specified.

Provenance

Source

Derived from the MIMIC-III (Medical Information Mart for Intensive Care) database.

Collection Method

Presumably a processed or curated subset of the original MIMIC-III data.

Time Range

null

Freshness

Last updated March 2026.

Geography

Region: US (inferred from platform tags).