Sepsis Survival Minimal Clinical Records from Norway and South Korea

110,204 patient admission records from a primary cohort in Norway, with a subset of 19,051 admissions meeting Sepsis-3 criteria and an external validation cohort of 137 patients from South Korea. The dataset was created by Chicco and Jurman (2020) for a study on sepsis survival. All categorical variables have been encoded and there are no missing values.

Use Cases

• Predicting sepsis survival outcomes based on minimal clinical features.
• Validating model generalizability using an external cohort from a different geographic region.
• Analyzing the relationship between sepsis preconditions and patient admissions.
• Benchmarking machine learning models on a clinical dataset with encoded categorical variables.

Strengths

• Large primary cohort of 110,204 patient admissions.
• Includes an external validation cohort of 137 patients for testing generalizability.
• No missing values and categorical variables are pre-encoded.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.
• Dataset size and specific file formats are not provided in the metadata.

Provenance

Source

OpenML, with data from a primary cohort in Norway and a validation cohort from South Korea.

Collection Method

Patient admission records, likely from hospital or clinical databases.

Time Range

null

Freshness

Last updated date is unknown.

Geography

Norway and South Korea