Renal Dysfunction and Serum Sodium-Based Risk Stratification for In-Hospital Mortality in Liver Cirrhosis.

Background and

Objectives

Kidney dysfunction and hyponatremia are prevalent in decompensated liver cirrhosis and are associated with suboptimal prognoses. Most prognostic tools used in cirrhosis are hepatic-centric and necessitate the utilization of multiple laboratory and clinical variables. This study endeavored to delineate and assess kidney function- and serum sodium-based prognostic models for the prediction of in-hospital mortality in patients with liver cirrhosis. Materials and

Methods

This retrospective single-center cohort study comprised 547 hospitalized patients with liver cirrhosis and comprehensive data pertaining to serum urea, creatinine, sodium, age, sex, and in-hospital outcome. In-hospital mortality was the primary endpoint. Kidney function was evaluated via the assessment of serum urea, serum creatinine, and estimated glomerular filtration rate (eGFR), using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) 2021 equation. Simple categorical and point-based scores were constructed. Model discrimination, calibration, and operating characteristics were compared with MELD, Child-Pugh, and ALBI scores, as well as with selected combined models.

Results

Of the 547 patients, 147 individuals (26.9%) succumbed during the hospitalization period. In the full model, lower eGFR, elevated urea levels, and diminished serum sodium concentrations were independently associated with in-hospital mortality, whereas age and sex did not demonstrate statistical significance. The full model appeared to exhibit moderate discrimination (AUC 0.701, 95% CI 0.652-0.750). A biochemical model based on urea, creatinine and sodium appeared to yield a similar AUC (0.696), and a renal-electrolyte model encompassing eGFR, urea, and sodium seemed to demonstrate an AUC of 0.694. A simple creatinine-sodium score may have attained an AUC of 0.681 and appeared to effectuate the stratification of mortality from 16.4% in the low-risk group to 53.1% in the high-risk group. Adding renal-electrolyte variables or the simple score to MELD did not appear to confer substantial enhancement to performance.

Conclusions

Kidney dysfunction and hyponatremia at admission have been identified as independent predictors of in-hospital mortality in liver cirrhosis. A simple creatinine-sodium score may afford practical bedside risk stratification and may complement MELD-based assessment in routine clinical care.