CHG Index and In-Hospital Mortality in Heart Failure Patients

A dual-cohort retrospective study of 1,687 hospitalized heart failure patients from the MIMIC-IV database (2008–2019) and an independent hospital cohort (2023–2025). The dataset, authored by Zefan Huang and shared on figshare in 2026, examines the Cholesterol, High-Density Lipoprotein, and Glucose index as a predictor of in-hospital mortality.

Use Cases

• Validate the CHG index as a prognostic biomarker based on its association with in-hospital mortality.
• Train mortality prediction models for heart failure patients based on metabolic-nutritional indicators.
• Compare risk stratification performance between the CHG index and other metabolic indicators mentioned in the description.
• Conduct survival analysis on hospitalized patient cohorts based on the Kaplan-Meier and Cox regression methods described.

Strengths

• Includes external validation on an independent cohort of 296 patients, strengthening generalizability.
• Reports specific hazard ratios (HR=1.768, 95% CI 1.337–2.338) and mortality rates (8.91% to 17.87%) from the training cohort.
• Uses a dual-cohort retrospective design with a total of 1,687 patient records.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the primary table is unknown, which may limit suitability assessment.
• Data may reflect temporal and institutional bias inherent to the specific hospital cohorts used.

Provenance

Source

MIMIC-IV database and an independent hospital-based cohort.

Collection Method

Dual-cohort retrospective design.

Time Range

Training cohort: 2008–2019; Validation cohort: July 2023 to December 2025.

Freshness

Last updated 2026-05-13 07:41:42; freshness should be verified.