Data Sheet 1_Dynamic TyG trajectories cumulative TyG burden are associated with in-hospita

4,760 ICU admissions from three databases (NSICU, MIMIC-IV, eICU) were retrospectively analyzed. The data includes serial triglyceride–glucose (TyG) measurements, derived trajectories, cumulative exposure metrics, and clinical variables for predicting in-hospital mortality in acute brain injury patients. The dataset, authored by Juan Wang and last updated in March 2026, supports the development of interpretable machine-learning models.

Use Cases

• Training interpretable machine-learning mortality models based on TyG trajectory patterns and cumulative burden metrics.
• Analyzing time-stratified associations between metabolic trajectories and in-hospital mortality after acute brain injury.
• Validating prognostic models that integrate dynamic TyG exposure with conventional severity indicators like SOFA scores.
• Investigating the prognostic relevance of threshold-based mean area under the curve (TBM) for metabolic burden.

Strengths

• Data from 4,760 ICU admissions across three distinct databases, suggesting a multi-center validation approach.
• Includes derived features like TyG trajectories and cumulative TBM metrics, which are central to the analysis.
• Model performance was internally and externally validated, as described in the study.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is a 2.9 MB PDF, which may require extraction of underlying data.

Provenance

Source

Retrospective analysis of three ICU databases: NSICU, MIMIC-IV, and eICU.

Collection Method

Clinical data extraction and derivation of TyG trajectories and cumulative exposure metrics.

Time Range

Temporal coverage is not specified in the provided input.

Freshness

Last updated 2026-03-18 07:38:48; freshness should be verified.

Geography

Spatial coverage is not specified in the provided input.