Acute Physiology Score and Sepsis Association in ICU Non-Surviving IBD Patients

Chenyang Li published a dataset on 2026-06-03 containing results from a retrospective cohort study of 190 adult inflammatory bowel disease patients who died during ICU admission. The data, from the eICU-CRD database, includes analysis of the association between Acute Physiology Score and sepsis using weighted logistic regression and XGBoost. The file is 10.3 KB and is licensed under CC-BY-4.0.

Use Cases

• Validate the association between Acute Physiology Score and sepsis based on the described cohort analysis.
• Train predictive models for sepsis risk in critically ill IBD patients based on physiological scores.
• Conduct subgroup analysis for elderly or male ICU patients based on the reported enhanced predictive performance.
• Investigate correlations between APS and metabolic markers like anion gap, lactate, and bicarbonate.
• Benchmark machine learning models against the described XGBoost and SHAP analysis methodology.

Strengths

• Dataset includes results from a cohort of 190 adult IBD patients who died in the ICU.
• Analysis includes multiple statistical methods: weighted logistic regression, restricted cubic splines, and XGBoost.
• Specific effect sizes are reported, such as an OR of 4.856 for the highest APS quartile.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The study notes a small sample size and its findings are hypothesis-generating, restricted to a specific subgroup.

Provenance

Source

eICU-CRD database

Collection Method

Retrospective cohort study

Freshness

Last updated 2026-06-03 04:48:51; freshness should be verified.