急性胰腺炎合并急性肾损伤患者的预后评估

Prognostic evaluation of acute pancreatitis complicated with acute renal injury

目的:评估急性胰腺炎(AP)合并急性肾损伤(AKI)患者的短期预后,并建立一个可预测AP合并AKI患者院内死亡风险的模型,探索其预后影响因素。方法:从重症监护医学信息数据库(MIMIC)Ⅳ 0.4中提取AP患者的临床数据。其后采用单因素及多因素logistic回归方法探索影响AP合并AKI患者预后的危险因素,在此基础上构建列线图。通过绘制受试者工作特征(ROC)曲线,判断其准确性。结果:共纳入1 213例AP患者,根据AP患者是否合并AKI,将其分为AKI组(739例)和非AKI组(474例),两组患者院内病死率分别为9.9%(73/739)和2.1%(10/474)。然后根据患者预后情况,将739例AP合并AKI患者分为存活组(666例)和死亡组(73例)。经单因素及多因素logistic回归分析,结果提示年龄[比值比(OR)= 1.033,95%置信区间(CI)(1.004,1.062),P=0.024],入院类型中相对于急诊入院,紧急入院[OR=2.203,95% CI(1.079,4.499),P=0.030]、择期入院[OR=0.134,95%CI(0.050,0.358),P < 0.001],高脂血症[OR=0.241,95%CI(0.124,0.468),P < 0.001],简化急性生理功能评分Ⅱ(SAPSⅡ)[OR=1.038,95%CI(1.014,1.062),P=0.002],碳酸氢盐[OR=0.910,95%CI(0.853,0.970),P=0.004],休克[OR=2.273,95%CI(1.166,4.429),P=0.016]及腹腔间隔室综合征(ACS)[OR=3.873,95%CI(1.173,12.788),P=0.026]为AP合并AKI患者死亡的影响因素。根据影响因素构建预测模型,计算曲线下面积为0.885[95%CI(0.844,0.926),P < 0.001]。结论:AP患者中,合并AKI者病情更严重,病死率明显升高,预后更差。基于年龄、入院类型、高脂血症、SAPSⅡ评分、碳酸氢盐、ACS、休克7项指标构建的列线图,对AP合并AKI患者具有较好的预测性能,有助于临床医生及时识别患者的死亡风险,从而采取更有效的治疗措施。

Objective:To explore the short-term prognosis of patients with acute pancreatitis (AP) complicated with acute kidney injury (AKI), and to establish a model that can predict the risk of in-hospital death in these patients so as to explore prognostic factors.Methods:The clinical data of patients with AP were extracted from the medical information mart for intensive care (MIMIC) Ⅳ 0.4. Afterwards, single-factor and multi-factor logistic regression methods were used to explore the risk factors that affect the prognosis of patients with AP combined with AKI. A nomogram was constructed on this basis and a receiver operating characteristic (ROC) curve was drawn to judge its accuracy.Results:Depending on whether they had concomitant AKI, a total of 1 213 patients with AP were included and separated into an AKI group (739 patients) and a non-AKI group (474 patients). The in-hospital mortality for the two groups was 9.9% (73/739) and 2.1% (10/474), respectively. Then, 739 patients with AP and AKI were separated into survival (666 patients) and death (73 patients) groups based on their prognosis. The single-factor and multi-factor logistic regression analysis showed that age [odds ratio (OR)=1.033, 95% confidence interval (CI) (1.004, 1.062), P=0.024], urgent admission [OR=2.203, 95%CI (1.079, 4.499), P=0.030] and selective admission [OR=0.134, 95%CI (0.050, 0.358), P < 0.001] compared with emergent admission, hyperlipidemia [OR=0.241, 95%CI (0.124, 0.468), P < 0.001], simplified acute physiology score Ⅱ (SAPSⅡ) score [OR=1.038, 95%CI (1.014, 1.062), P=0.002], bicarbonate [OR=0.910, 95%CI (0.853, 0.970), P=0.004], shock [OR=2.273, 95%CI (1.166, 4.429), P=0.016], and abdominal compartment syndrome (ACS) [OR=3.873, 95%CI (1.173, 12.788), P=0.026] were influencing factors for in-hospital death in patients with AP complicated with AKI. A predictive model was constructed based on these influencing factors, and the area under the curve (AUC) was calculated to be 0.885 [95%CI (0.844, 0.926), P < 0.001].Conclusions:Among AP patients, those with AKI are more seriously ill, with a significantly higher mortality rate and a worse prognosis. The nomogram constructed based on seven indicators of age, admission type, hyperlipidemia, SAPSⅡ score, bicarbonate, ACS, and shock has good predictive performance for patients with AP and AKI, which helps clinicians identify the risk of mortality in time and take more effective treatment measures.