Abstract
Liver disease is a leading cause of death worldwide. Nucleated red blood cells (NRBCs) are associated with high mortality and poor outcomes in patients with severe illnesses. However, the relationship between NRBCs and severe liver disease remains unclear. Potential confounding effects were managed using propensity score matching. The association between NRBCs and clinical outcomes in patients with liver disease was clarified using Cox proportional hazards regression analysis and smoothing splines. Differences in NRBCs between 30-day survivors and non-survivors within the pre-matched cohort during the first 30 days after ICU admission were assessed using generalised additive mixed models. Compared to the 30-day survivors, the 30-day non-survivors had significantly higher NRBC counts. Higher NRBC counts were significantly correlated with an augmented risk of 30-day, 90-day and in-hospital mortality, with concurrently decreased hospitalisation durations. Inpatients with liver disease, progressive increases in the 30-day mortality risk were associated with increased NRBC counts. The association between NRBCs and enhanced 30-day mortality rates was consistent across stages and etiologies. Moreover, 30-day non-survivors experienced average daily increases in NRBC counts of 0.31% compared with 30-day survivors. Elevated NRBC counts correlated with increased 30-, 90-day and in-hospital mortality in patients with liver disease.