Association between red cell distribution width to albumin ratio and all-cause mortality in patients with acute pancreatitis admitted to the intensive care unit: a retrospective study based on the MIMIC-IV database

Xuan Chen Yuchen Luo ^* Side Liu ^*

• Department of Gastroenterology, Nanfang Hospital, Southern Medical University, Guangzhou, China

Background: Red blood cell volume distribution width (RDW) and albumin (Alb) have been proved to be predictors of mortality in various diseases, such as acute pancreatitis (AP). However, until now the relationship between RDW to Alb ratio (RAR) and mortality in AP has not been fully elucidated. Therefore, this study aims to evaluate the relationship between RAR and all-cause mortality in AP.Method: Patients with AP in the Critical Care Medical Information Market (MIMIC-IV) database who met criteria were included in this retrospective study. Associated baseline data was obtained, cleaned and analyzed. Kaplan Meier (K-M) survival curve and Cox proportional hazards regression model were utilized to evaluate the relationship between RAR and all-cause mortality. Restricted Cubic Spline (RCS) was used for exploring how hazard ratio (HR) changes as RAR varied. Additionally, Receiver Operating Characteristic (ROC) analysis and subgroup analysis were conducted to assess the predictive value and to explore the significance of RAR in different populations.Results: 499 patients were included in this study. Survival curve showed that patients with RAR > 5.14 had higher mortality rate at 7-day (d), 14-d, 21-d, 28-d, 90-d, 180-d and 1-year (y). The univariate and multivariate Cox models revealed an independent association between high-level RAR and allcause mortality at 28-d, 90-d and 1-y. RCS showed that RAR became a risk factor when exceeding 5.14. RAR only had linear relationship with mortality at 1-y after adjusting for the potential confounders. Subgroup analysis suggested that increased RAR caused higher risk of death in male, non-white people or those patients without respiratory failure (RF). ROC analysis indicated that compared with other parameters such as SOFA score, RAR exhibited higher efficiency in predicting in-hospital and all-cause mortality at 14-d, 21-d, 28-d, 90-d. Combined RAR with BISAP, RARmodified BISAP showed superiority in predicting short-term mortality (28-d). Conclusion: For patients with AP in ICU, RAR has a strong association with short-and long-term prognosis. Especially, RAR is a promising indicator for short-term all-cause mortality in patients with AP. For males, non-white patients and those without RF, elevated RAR may be a more dangerous signal of mortality.