泛免疫炎症指数在入院早期与延迟阶段对急性胰腺炎严重程度的预测价值研究

吴依婷; 王雅琦; 魏文俊; 丁晓春

泛免疫炎症指数在入院早期与延迟阶段对急性胰腺炎严重程度的预测价值研究

Study on the predictive value of pan-immune inflammation value for the severity of acute pancreatitis in the early admission and delayed admission stages

摘要

摘要:
目的： 探讨泛免疫炎症指数（PIV）在入院早期与延迟阶段（入院24 h内和入院24 h后）对急性胰腺炎（AP）严重程度的预测价值。
方法： 501例AP病人根据病情严重程度分为轻症急性胰腺炎组（MAP）与非轻症组（non–MAP，包括中度重症及重症）。收集入院24 h内及24 h后的血常规与生化数据，计算相关炎症指标，将全身免疫炎症指数（SII）、C反应蛋白（CRP）、中性粒细胞计数/淋巴细胞计数（NLR）、中性粒细胞/血小板比值（NPR）、C反应蛋白/白蛋白比值（CAR）、C反应蛋白/淋巴细胞比值（CLR）、淋巴细胞/单核细胞比值（LMR）在入院早期与延迟阶段对AP严重程度的预测价值进行比较，采用二元logistic回归分析严重程度的独立影响因素，并通过受试者工作特征（ROC）曲线评估各指标及联合模型的预测效能。
结果： non–MAP组年龄显著高于MAP组（P ＜ 0.01）。入院24 h内，non–MAP组PIV水平高于MAP组（P ＜ 0.01），但多因素回归显示PIV并非独立危险因素（P = 0.97）。入院24 h后，PIV成为独立危险因素（P ＜ 0.01）。ROC分析显示，PIV预测non–MAP的AUC为0.734（95%CI：0.683 ~ 0.784），灵敏度73.43%，特异度64.01%，截断值为426.56。入院24 h后，联合模型CAR + LMR + NLR的AUC为0.750（95%CI：0.703 ~ 0.797），进一步纳入PIV后（CAR + LMR + NLR + PIV）AUC提升至0.760（95%CI：0.713 ~ 0.807），为所有模型中最优。
结论： PIV在AP入院24h后具备良好的病情严重程度预测价值，联合多指标可进一步提升预测效能，可作为临床评估的新工具。

Abstract:
Objective To investigate the predictive value of the pan-immune inflammation value (PIV) for the severity of acute pancreatitis (AP) in the early and delayed stages of admission (within 24 hours and after 24 hours of admission).
Methods A total of 501 patients with AP were divided into the mild acute pancreatitis group (MAP) and non-mild group (non-MAP, including moderate severe and severe cases) according to the severity of disease. The blood routine and biochemical data of patients within 24 hours and after 24 hours of admission were collected, and the relevant inflammatory indicators were calculated. The predictive value of the systemic immune inflammatory index (SII), C-reactive protein (CRP), neutrophil count/lymphocyte count (NLR), neutrophil/platelet ratio (NPR), C-reactive protein/albumin ratio (CAR), C-reactive protein/lymphocyte ratio (CLR) and lymphocyte/monocyte ratio (LMR) for the severity of AP in the early stage and delayed stage of admission were compared. Binary logistic regression was used to analyze the independent influencing factors of severity, and the predictive efficacy of each index and combined model were evaluated using the receiver operating characteristic (ROC) curve.
Results The age of the non-MAP group was significantly higher than that of the MAP group (P ＜ 0.01). Within 24 hours of admission, the PIV level in the non-MAP group was higher than that in the MAP group (P ＜ 0.01), but the results of multivariate regression showed that the PIV was not an independent risk factor (P = 0.97). After twenty-four hours of admission, the PIV became an independent risk factor (P ＜ 0.01). The results of ROC analysis showed that the AUC of PIV for predicting non-MAP was 0.734 (95%CI: 0.683–0.784), with a sensitivity of 73.43%, a specificity of 64.01% and a cut-off value of 426.56. After twenty-four hours of admission, the AUC of the combined model CAR + LMR + NLR was 0.750 (95%CI: 0.703–0.797). After further inclusion of PIV (CAR + LMR + NLR + PIV), the AUC increased to 0.760 (95%CI: 0.713–0.807), which was the best among all models.
Conclusions The PIV has a good predictive value for the severity of AP after 24 hours of admission, and its combination with multiple indicators can further improve the predictive efficacy. Therefore, PIV can be used as a novel tool for clinical severity assessment of AP.

HTML全文

参考文献(22)

施引文献

资源附件(0)