BISAP、MEWS和常规检验指标的急性胰腺炎严重程度预测模型的临床意义

发布时间：2018-06-16 07:33

本文选题：急性胰腺炎 + 急性胰腺炎严重程度床边指数　；参考：《昆明医科大学》2017年硕士论文

【摘要】：[目的]探讨红细胞分布宽度(Red Cell Distribution Width,RDW)、血清Ca~(2+)水平、改良早期预警的评分(Modified Early Warning Score,MEWS)、急性胰腺炎严重程度床边指数(Bedside Index for Severity in Acute Pancreatitis,BISAP)对预测急性胰腺炎严重程度的价值,并构建一种更为精确的评分系统在发病早期来预测急性胰腺炎(Acute Pancreatitis,AP)的严重程度指导基层对AP进行分级,从而及时进行干预和转诊,降低病死率及重症率。[方法]分别统计302例急性胰腺炎患者红细胞分布宽度和血清Ca~(2+)水平,统计和计算302例AP病例的MEWS、BISAP,使用单因素logistic回归分析RDW、血清Ca~(2+)水平、MEWS、BISAP是否为AP严重程度的预测指标。将单因素logistic回归中有统计学意义的参数纳入多因素logistic回归,采用向前逐步回归,筛选变量,构建多因素预测模型。构建受试者工作特征曲线(receiver operating characteristic curve,ROC),通过曲线下面积,比较多因素预测模型和各单因素模型预测AP严重程度的意义,并采用bootstrap法对模型的内部效度进行验证。[结果]302例患者中MAP 209例,SAP 93例。单因素logistic回归分析后发现,血清Ca~(2+)水平、MEWS、BISAP均为AP严重程度的预测指标(P值均0.001),而RDW不是AP严重程度的预测指标(P0.05)。多因素logistic回归分析后发现,血清Ca~(2+)水平和BISAP是AP严重程度的独立预测指标(P值均0.001),而MEWS不是AP严重程度的独立预测指(P0.05),且血清Ca~(2+)水平和 BISAP 呈负相关(r =-0.330, P0.001)。各模型对SAP的预测能力为:联合血清Ca~(2+)水平和BISAP、新构建的预测模型血清Ca~(2+)水平BISAP,血清Ca~(2+)水平和BISAP的预测能力无统计学意义(P0.05);新构建的预测模型分别与单项血清Ca~(2+)水平、BISAP的预测能力有显著统计学意义(P0.01)。采用bootstrap法对各模型的内部效度进行验证后发现3个模型内部效度良好。[结论]血清Ca~(2+)水平和BISAP对AP严重程度的预测价值较高,但联合血清Ca~(2+)水平和BISAP构建的模型明显优于血清Ca~(2+)水平和BISAP,且简单易行,值得在临床推广。
[Abstract]:[objective] to investigate the value of red cell distribution (RDW2), modified early warning score (MEWS) and bedside Index for severity in Acute pancreatitis (BISAP) in predicting the severity of acute pancreatitis. Furthermore, a more accurate scoring system was established to predict the severity of acute pancreatitis (AP) at the early stage of the disease, to guide the basic units to grade AP, so that timely intervention and referral could be carried out to reduce the mortality rate and the severe rate of acute pancreatitis. [methods] the distribution width of erythrocyte and the level of Cafi2 in serum of 302 patients with acute pancreatitis were counted and calculated. The single factor logistic regression analysis was used to determine whether MEWS BISAP was a predictor of AP severity. The parameters with statistical significance in univariate logistic regression were incorporated into multivariate logistic regression and the multivariate prediction model was constructed by stepwise forward regression and screening of variables. The receiver operating characteristic curve was constructed. The significance of predicting AP severity by multi-factor prediction model and single-factor model was compared by the area under the curve, and the internal validity of the model was verified by bootstrap method. [results] among 302 patients, there were 209 cases with SAP and 93 cases with map. The results of univariate logistic regression analysis showed that the serum Caan2) level was a predictor of AP severity (P = 0.001), while RDW was not a predictor of AP severity (P 0.05). The results of multivariate logistic regression analysis showed that the serum Caan2) level and BISAP were both independent predictors of AP severity (P = 0.001), while Mews was not an independent predictor of AP severity (P 0.05), and the serum Caanzao (2) level was negatively correlated with BISAP (r = -0.330, P 0.001). The predictive ability of each model to SAP is as follows: combined serum Caan2) level and BISAP level, newly constructed prediction model serum Caanzao 2) level BISAP level, serum Caanzao 2) level and BISAP prediction ability have no statistical significance, the new prediction model and single prediction model have no statistical significance (P0.05). The predictive ability of BISAP was statistically significant (P 0.01). Bootstrap method was used to verify the internal validity of each model, and it was found that the internal validity of the three models was good. [conclusion] the level of serum Caanzao 2) and BISAP in predicting the severity of AP were higher, but the combined serum level of Caan2) and the model of BISAP were obviously superior to the level of serum Caan2) and BISAP, and were simple and easy to use, so it was worth popularizing in clinic.
【学位授予单位】：昆明医科大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R576

【参考文献】