Predictive value of red cell distribution width to albumin ratio for acute kidney injury in patients with acute pancreatitis.

Acute pancreatitis (AP) is a necro-inflammatory disease of the pancreas characterized by abnormal activation of pancreatic enzymes, which triggers systemic inflammation and contributes to poor prognosis [ 1 , 2 ]. Acute kidney injury (AKI), a common complication in AP patients characterized by a rapid deterioration of renal function, further exacerbates AP-related adverse outcomes and is associated with high mortality [ 3 ]. The pathophysiology of AP-associated AKI (AP-AKI) is multifactorial, and recent studies indicate that oxidative stress is a key driver of AP-AKI. Excess reactive oxygen species (ROS) production during acute pancreatitis not only exacerbate inflammation and pancreatic tissue damage, but also induce endothelial dysfunction and renal hypoperfusion, resulting in the development of AKI. The inflammatory cascade triggered by AP causes further tissue damage, underlining the complex role of inflammation and oxidative stress in AP [ 4 , 5 ]. Notably, recent research highlights the crosstalk between pancreatic inflammation and renal dysfunction in AP. Pancreatic injury releases pro-inflammatory cytokines (e.g., TNF-α, IL-6) and damage-associated molecular patterns (DAMPs) into the systemic circulation. Then these factors trigger renal tubular apoptosis and impair renal microcirculation [ 6 , 7 ]. Conversely, compromised renal function exacerbates pancreatic injury via accumulated uremic toxins, reduced clearance of inflammatory mediators, and electrolyte disturbances [ 8 ]. This interaction emphasizes the need for biomarkers that reflect both local pancreatic inflammation and systemic organ crosstalk. Red blood cell distribution width (RDW), a routine hematological parameter reflecting erythrocyte size variability, has been increasingly considered as a biomarker of systemic inflammation and oxidative stress in recent years. Elevated RDW is associated with impaired antioxidant capacity and exacerbated inflammatory status, potentially reflecting redox imbalance in AP patients [ 9 , 10 ]. In contrast, serum albumin, a negative acute-phase protein, decrease in inflammatory status due to increased vascular permeability and reduced hepatic synthetic capacity [ 11 – 13 ]. The decrease in albumin levels not only indicates nutrient depletion but also reflects a persistent inflammatory state, both of which are commonly observed in severe AP [ 14 ]. Therefore, the RDW-to-albumin ratio (RAR) integrates two pathophysiological mechanisms: oxidative stress-driven erythrocyte dysregulation and inflammation-associated hypoalbuminemia, serving as a novel composite biomarker for evaluating disease severity in AP patients with AKI. Research on prognostic biomarkers for AP-AKI has increased in recent years.. Numerous biomarkers have been found to effectively predict AKI development in AP patients, such as neutrophil gelatinase-associated lipocalin (NGAL), β2-microglobulin (β2-MG), and cystatin C [ 15 , 16 ]. Despite their high sensitivity and specificity, these novel biomarkers are limited by high costs and the need for dynamic monitoring. In contrast, RAR can be rapidly calculated based on routine hematological and biochemical parameters obtained at admission, making it both convenient and affordable. However, the predictive effect of RAR on AP-AKI remains unconfirmed. In view of the role of systemic inflammation and oxidative stress in multiple organ failure [ 17 , 18 ], we speculated that RAR might serve as an important and easily accessible biomarker for AKI prediction and risk stratification. Thus, this study aims to explore the relationships between RAR and the incidence of AKI in AP patients, thereby providing new insights for early risk identification.

The data utilized in this study were derived from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database, a publicly available critical care repository jointly developed by Massachusetts Institute of Technology researchers and clinical investigators at Beth Israel Deaconess Medical Center [ 19 ]. This database includes information on all patients admitted to the Beth Israel Deaconess Medical Center during the years from 2008 to 2019. Since all data were publicly available and de-identified, the study was exempt from ethical approval and informed consent requirements. The lead author of the study, Zhao Jin, completed the required credentialing process and obtained full access privileges to the MIMIC-IV database (User ID: 67544010). A total of 1151 patients diagnosed with AP were extracted according to the ICD codes 9 and 10. The exclusion criteria were as follows: age < 18 years; Length of hospital stay < 24h; patients with renal disease; and incomplete medical data. After applying the exclusion criteria, 600 patients were finally included in the study ( Fig 1 ). All variables were extracted from the MIMIC-IV database using PostgresSQL software. The demographic characteristics included sex, age and race. Laboratory variables within the first 24h of ICU admission included hemoglobin (Hb), red blood cell (Rbc), red blood cell distribution width (Rdw), hematocrit (Hct), white blood cell (Wbc), platelet count (Plt), albumin (Alb), prothrombin time (PT), international normalized ratio (INR), creatinine (Cr), blood urea nitrogen (Bun), aspartate aminotransferase (Ast) and alanine aminotransferase (Alt). The severity scores on admission included simplified acute physiology score II (SAPSII) and sequential organ failure assessment (SOFA) score. Comorbidities included sepsis, hypertension, diabetes, myocardial infarction (MI), chronic obstructive pulmonary disease (COPD), invasive ventilation, cardiopulmonary resuscitation (CPR) and the application of renal continue replacement therapy (CRRT). The defined endpoint event was the incidence of AKI resulted from AP. Continuous variables are presented as means with standard deviations, and categorical variables as counts with percentages. For continuous variables, continuous variables with normal distribution were compared by one-way ANOVA, while non-normally distributed variables were analyzed using Kruskal-Wallis test. The Chi-square or Fisher’s exact test was used for categorical variables. Multiple logistic regression analysis was performed to calculate the Odds Ratio (OR) and the 95% confidence interval (CI) for the RAR and incidence of AKI within each subgroup. Model 1 was unadjusted; model 2 was adjusted for sex, age and race; and model 3 was additionally adjusted for Wbc, Hct, CRRT, map, creatinine, Bun and Ast. Restricted cubic spline (RCS) was performed to further validate the potential nonlinear relationships between RAR and the incidence of AKI. To verify the stability of our results, stratified analyses were conducted based on the following factors, including age, sex, sepsis, MI, CHF, COPD, hypertension, diabetes, invasivevent and RRT. Statistical analyses were performed with Empowerstats software (version 6.0). P-value < 0.05 was considered statistically significant.

According to the inclusion and exclusion criteria, 600 patients were included in the study, and the patient screening flow chart is shown in Fig 1 . According to occurrence of AKI or not, the involved AP patients were divided into AKI group and non-AKI group. There were no significant differences in sex, race, HR, Wbc, Plt, Hb, MI, COPD, diabetes, and CPR between the groups. However, significant differences were observed in age, ICU time, hospital time, SOFA score, Map, Rdw, Rbc, Alb, Cr, Bun, PT, INR, Alt, Ast, RAR, sepsis, CHF, hypertension, demand for invasive ventilation, and CRRT requirement (P < 0.05, Table 1 ). According to univariate analysis, age, hospital time, ICU time, sepsis, CHF, hypertension, SAPS II, SOFA score, invasive ventilation, RDW, Cr, Alb, Bun, and RAR were shown to be the primary confounders which influenced AKI incidence (P < 0.05) ( S1 Table ). Based on multiple logistic regression analysis, RAR showed a positive correlation with AKI as either a continuous (OR 1.46, 95% CI 1.22–1.76, P < 0.05) or categorical variable (OR 2.79, 95% CI 1.69–4.59, P < 0.05, Table 2 ). Model1: Unadjusted. Model2: adjusted for Gender, Age, Race. Model3: adjusted for Gender, Age, Race, Wbc; Hct; CRRT; Map; Creatinine; Bun; Ast. We analyzed the predictive value of RAR, RDW and albumin for AKI in AP patients using ROC curve analysis. The result showed that the area under the curve (AUC) of RAR was superior to those of RDW and Albumin ( Fig 2 ). In addition, we analyzed differences of RAR in different AKI stages based on the KDIGO guideline. The results reflected that RAR in the AKI stage 3 was significantly higher than others (P < 0.05, S1 Fig ). ROC curves of RAR (green line), Rdw (blue line), albumin (red line). RAR, red blood cell distribution width-to-albumin ratio; Rdw, red blood cell distribution. The RCS analysis was adjusted for the effects of sex, age, race, Wbc, CRRT, map, creatinine, Bun and Ast. It indicated a linear association between RAR and AKI after adjusting for confounding factors (P non-linear = 0.128, Fig 3 ). RAR was entered as a continuous variable. Hazard ratios were adjusted for Gender, Age, Race, Wbc, Hct, CRRT, Map, Creatinine, Bun, Ast. RAR, red blood cell distribution width-to-albumin ratio; WBC, white blood cell; Hct, hematocrit; CRRT, continuous renal replacement therapy; MAP, mean arterial pressure; BUN, blood urea nitrogen; AST, aspartate aminotransferase. The red line represents the estimated values. The shaded area represents the corresponding 95% confidence intervals. Subgroup analysis was performed to assess the relationship between RAR and AKI among the various subgroups. No significant interaction effect was found in any subgroups after stratifying by sex, age, hypertension, DM, SOFA score, and severity of AP (all P for interaction > 0.05， Table 3 ).

High RAR serves as an independent predictor for AKI in AP patients. Early assessment of RAR may facilitate risk stratification to guide clinical management, thereby improving clinical outcomes.

(DOCX) (TIFF)