Table 1_The effect of insulin on 28-day mortality in non-diabetic sepsis patients: a retro

11,289 critically ill non-diabetic sepsis patients from the MIMIC-IV database were analyzed to evaluate insulin's effect on 28-day mortality. The study by Yifei Liu, last updated in April 2026, used propensity score matching and Cox regression, finding insulin use was associated with reduced mortality but longer hospital stays.

Use Cases

• Analyze the association between insulin therapy and 28-day mortality based on the described retrospective cohort.
• Model survival outcomes in sepsis patients based on the described hazard ratios and Kaplan-Meier analysis.
• Investigate length of stay differences between treatment groups based on the reported ICU and hospital LOS metrics.

Strengths

• Analysis is based on a substantial cohort of 11,289 patients from the MIMIC-IV database.
• Methodology includes propensity score matching to balance cohorts, with standardized mean differences reported.
• Specific statistical results are provided, including hazard ratios (0.63-0.67) and p-values.

Limitations

• Row count and column-level documentation for the underlying data are unknown.
• The dataset is very small (13.2 KB), suggesting it likely contains summary results or a manuscript, not the full patient-level data.
• As a retrospective observational study, it cannot establish causality.

Provenance

Source

MIMIC-IV database

Collection Method

Retrospective analysis employing propensity score matching and Cox regression.

Time Range

The temporal coverage of the underlying MIMIC-IV data is not specified.

Freshness

Last updated 2026-04-15 05:22:36

Geography

The geographic coverage of the underlying MIMIC-IV data is not specified.