Of these, 1,260 received prophylactic enoxaparin, a low-molecular-weight heparin, while 1,081 received unfractionated heparin, often abbreviated UFH. Crucially, the investigators restricted enrollment to patients without severe renal impairment, defined as an estimated glomerular filtration rate of at least 30 mL/min/1.73 m². The primary endpoints were stark and clinically meaningful: 30-day ICU mortality and 90-day in-hospital mortality.

After adjusting for confounding factors with multivariate Cox proportional hazards regression, the hazard ratio for enoxaparin versus unfractionated heparin was 0.94 for 30-day ICU mortality, with a 95 percent confidence interval of 0.67 to 1.33 and a P value of 0.735. For 90-day in-hospital mortality, the hazard ratio was 0.85, with a confidence interval of 0.66 to 1.11 and a P value of 0.242. In plain terms, patients on enoxaparin showed a numerically lower risk of dying at both time points, but the differences were well within the range expected from random variation alone.

Survival patterns were first visualized with Kaplan-Meier estimates, the standard method for tracking how mortality accumulates over follow-up time in each group. The Cox models then quantified the association while accounting for baseline differences between patients, such as illness severity and demographics. These weighted analyses supported the same conclusion.

Subgroup analyses added another layer of reassurance. The researchers stratified patients by age, sex, and body mass index, testing whether the comparative effect of the two heparins differed across these strata. All interaction P values exceeded 0.05, indicating no significant heterogeneity.

In other words, no examined patient subgroup appeared to derive a clear survival advantage from either agent. Yet whether those pharmacologic advantages translate into better survival in septic patients had remained unknown, and the new study addresses that gap directly—while carefully noting that it does not establish superiority for either drug. The authors are candid about the limits of their evidence.

As a retrospective cohort study, it relies on existing database records rather than random assignment, so unmeasured confounding cannot be excluded. Restricting the cohort to patients without severe renal dysfunction, while statistically sensible, means the findings do not speak to the sickest kidney patients, in whom dosing and drug clearance differ substantially. The authors conclude that larger, prospective studies are warranted to confirm the observed associations, and their formal conclusion states only that enoxaparin was associated with numerically lower but not statistically significant hazard ratios.

Subject of Research: Comparative mortality outcomes of enoxaparin versus unfractionated heparin thromboprophylaxis in sepsis patients

Article Title: Comparative effectiveness of enoxaparin versus unfractionated heparin for 30- and 90-day mortality in sepsis patients without severe renal dysfunction: a retrospective cohort study

Article References: Xue, Z., Guo, R., Lou, X., Tian, Y., Tong, J., & Zhao, W. (2026). Comparative effectiveness of enoxaparin versus unfractionated heparin for 30- and 90-day mortality in sepsis patients without severe renal dysfunction: a retrospective cohort study. Frontiers in Medicine, 13, Article 1886512. https://doi.org/10.3389/fmed.2026.1886512

Image Credits: AI Generated

DOI: 10.3389/fmed.2026.1886512

Keywords: sepsis, enoxaparin, unfractionated heparin, thromboprophylaxis, mortality, retrospective cohort study, eICU database, intensive care, Cox regression, renal function, Comparative, effectiveness