Association between Red Cell Distribution Width (RDW)-related inflammatory biomarkers and prognosis in ICU Cirrhosis...

Cirrhosis signifies a critical phase in liver disease with high mortality, particularly in patients admitted to intensive care unit (ICU). Red cell distribution width (RDW) demonstrates prognostic value in various conditions, with evidence linking it to inflammation and adverse outcomes in liver disease. However, the predictive utility of RDW and its derived indices in patients with cirrhosis admitted to ICU remains undetermined. This study investigated the associations between RDW-related biomarkers and prognosis in critically ill cirrhosis patients.

A retrospective analysis used data from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database, including 1,871 adults with cirrhosis at first ICU admission. The analyzed biomarkers were RDW, RDW-to-albumin ratio (RAR), RDW-to-platelet ratio (RPR), and hemoglobin-to-RDW ratio (HRR). The primary outcome was 30-day all-cause mortality; secondary outcomes were 90-day, and 365-day mortality. Adjusted multivariate Cox regression models determined biomarker-mortality associations. Receiver operating characteristic (ROC) curve analyses evaluated predictive performance of the biomarkers.

The cohort had a mean age of 58.8 ± 12.3 years and 64.8% of the patients were male. The mortality rates were 31.8% (30-day), 39.3% (90-day), and 48.0% (365-day). Mortality was consistently higher in high versus low RDW groups. After multivariable adjustment, all biomarkers were independently associated with 30-day, 90-day, and 365-day mortality, respectively. When expressed per SD, RDW (HR 1.41, 95% CI 1.30-1.52), RAR (HR 1.17, 95% CI 1.11-1.24), RPR (HR 1.19, 95% CI 1.11-1.28), and HRR (HR 0.52, 95% CI 0.44-0.61) were each significantly associated with 30-day mortality. ROC analysis revealed RDW had the highest predictive value (AUC 63.3%), followed by HRR (AUC 61.7%), RAR (AUC 59.6%), and RPR (AUC 55.6%). RDW demonstrated the highest discriminative ability for 30-day mortality, with incremental predictive value when combined with established scores (absolute AUC increase 0.68-1.58%).

RDW-related biomarkers demonstrate statistically significant but modest associations with mortality in ICU cirrhosis patients. RDW and HRR showed the strongest predictive capabilities among the four indices, though standalone discrimination was limited. These accessible, cost-effective biomarkers may serve as adjunctive prognostic markers when combined with established scoring systems. However, further prospective validation is required before clinical implementation, and dynamic changes in these biomarkers warrant investigation in future studies.