血尿素氮与血清白蛋白的比率对重症监护室中急性胰腺炎患者住院期间全因死亡风险的预测价值
Predictive Value of Blood Urea Nitrogen-to-Serum Albumin Ratio on the Risk of in-Hospital All-Cause Death in Patients with Acute Pancreatitis in the Intensive Care Unit

目的：探究血尿素氮与血清白蛋白的比率(BUN/Alb)对重症监护室急性胰腺炎(AP)患者住院期间全因死亡风险的预测价值。方法：回顾性分析MIMIC-IV数据库中904例重症AP患者临床资料。以BUN/Alb为暴露变量，院内全因死亡为结局变量，分为存活组(789例)与死亡组(115例)。对比两组基线，用Logistic回归、趋势性检验、限制性三次样条图以及亚组分析探索BUN/Alb与死亡率关系及预测价值。结果：死亡组年龄更高，合并房颤、急性心梗比例更大，血清阴离子间隙等多项指标更高，ICU住院时间更长，更易使用血管活性药物(P < 0.05)。Logistic多元回归显示，BUN/Alb是AP患者死亡率重要独立预测因子(OR = 1.05，95% CI: 1.03~ 1.08, P <0. 001)。将BUN/Alb分四分位数，随其升高死亡率上升。限制性三次样条曲线示二者呈正相关线性关系。亚组分析表明，高血压、糖尿病等与BUN/Alb和死亡率存在相关性相互作用。结论：对于重症监护室中的AP患者而言，住院期间的BUN/Alb是院内全因死亡发生的独立预测指标。
Objective: To investigate the predictive value of the blood urea nitrogen to serum albumin ratio (BUN/Alb) for in-hospital all-cause mortality risk in acute pancreatitis (AP) patients admitted to the intensive care unit (ICU). Methods: A retrospective analysis was conducted on clinical data from 904 critically ill AP patients in the MIMIC-IV database. BUN/Alb served as the exposure variable, and in-hospital all-cause mortality as the outcome variable. Patients were categorized into survival (n = 789) and non-survival (n = 115) groups. Baseline characteristics were compared between groups. Logistic regression, trend tests, restricted cubic spline curves, and subgroup analyses were employed to explore the relationship between BUN/Alb and mortality, as well as its predictive value. Results: Compared with the survival group, non-survivors were older, had higher proportions of atrial fibrillation and acute myocardial infarction comorbidities, and exhibited elevated serum anion gap, blood urea nitrogen, serum creatinine, blood glucose, white blood cell count, blood lactate, plasma crystalloid osmotic pressure, sodium and potassium levels, SAPS II scores, SOFA scores, and longer ICU stays. They were also more likely to receive vasoactive medications (all P < 0.05). Fully adjusted multivariate logistic regression identified BUN/Alb as a significant independent predictor of mortality in AP patients (OR = 1.05, 95% CI: 1.03~1.08, P < 0.001). When BUN/Alb was divided into quartiles, mortality rates progressively increased with ascending BUN/Alb quartiles (P for trend < 0.05). Restricted cubic spline analysis revealed a positive linear correlation between BUN/Alb and in-hospital mortality. Subgroup analyses demonstrated significant interaction effects between hypertension, diabetes, chronic kidney disease, vasoactive medication use, and the association of BUN/Alb with mortality (P for interaction <0.05). Conclusion: For AP patients