SUPPORT2 ICU Patient Outcomes Study

SUPPORT2 is a clinical dataset from the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments. It contains patient-level data from a multi-center study on seriously ill hospitalized adults, focusing on prognoses, treatments, and outcomes. The dataset is hosted by the UCI Machine Learning Repository.

Use Cases

• Predict patient survival duration or mortality risk using features like age, diagnosis, and physiological measurements.
• Analyze the relationship between treatment preferences, such as DNR (Do Not Resuscitate) orders, and actual patient outcomes.
• Model hospital length of stay or resource utilization based on admission diagnoses and patient comorbidities.
• Identify risk factors for specific outcomes by analyzing clinical variables like APACHE score, organ failure, and functional status.

Strengths

• Derived from a major, well-documented multi-center clinical trial (SUPPORT study).
• Contains a wide range of clinical and prognostic variables relevant to ICU outcomes.

Limitations

• Specific row count, column count, and temporal coverage are unknown from the input.
• Data may be temporally stale, reflecting medical practices from the original study period.

Provenance

Source

UCI Machine Learning Repository.

Collection Method

Collected as part of the SUPPORT clinical study on seriously ill hospitalized adults.

Geography

Multi-center study, likely within the United States.