脓毒症早期血小板变化与并发急性肾衰竭的关系

Correlation between early change in platelet counts and complication with acute renal failure in septic patients

目的:探讨脓毒症患者血小板的早期变化与并发急性肾衰竭的相关性。方法:回顾性分析2019年1月到2020年1月深圳市人民医院收治的60例脓毒症患者的临床资料,根据是否出现并发急性肾衰竭分为并发组(n=28)和无并发组(n=32),对比两组一般资料、血小板、Scr、Cys-C、APACHEⅡ评分、SOFA评分,以并发急性肾衰竭为因变量,以上述有统计学意义的变量为自变量,作多元线性回归,校正PLT对并发急性肾衰竭是否存在独立影响。以死亡和存活分为两个类别,以PLT水平为预测变量,绘制ROC曲线,分析PLT的预测并发急性肾衰竭的阈值。结果:并发组血清Scr、Cys-C均高于无并发组,并发组各个时相点血小板均低于无并发组(均P<0.05);并发组血小板表达水平与Scr、Cys-C及KIM-1表达水平均呈正相关(r=0.704、0.812、0.863,P<0.01);ROC曲线分析结果显示,血清Scr、Cys-C及血小板预测脓毒症患者并发急性肾衰竭的曲线下面积(0.962,95%CI:0.906~0.998)均分别大于上述各单项指标预测的曲线下面积(P<0.05),且敏感度(97.0%)和特异度(91.4%)最高。结论:脓毒症患者血小板早期下降对预测脓毒症患者并发急性肾衰竭具有重要价值。

Objective:To investigate the correlation between early changes in platelets(PLT)in patients with sepsis and complication with acute renal failure(ARF).Methods:A retrospective analysis was performed on clinical data from 60 patients with sepsis admitted to Shenzhen People’s Hospital between January 2019 and January 2020. The subjects was divided into the complicated group( n= 28)and a non-complicated group(n= 32)according to presence of ARF. The two groups were compared for demography,PLT counts,serum creatinine(Scr),Cys-C,APACHE Ⅱ and SOFA scores. Multivariate linear regression was performed using complication with ARF as the dependent variable and above-mentioned indicators with statistical significance in univariate tests as independent variables,so as to validate whether PLT count has an independent effect on complication with ARF. A ROC curve was generated with PLT count as the predictor for death and survival as two categories. The prediction of ARF and cut-off value by PLT count was investigated.Results:The serum Scr and Cys-C levels were higher,and the PLT count was lower,in the complicated group than those in the noncomplicated group at all time points(allP< 0.05). The PLT expression level in the complicated group was positively correlated with the expression levels of Scr(r= 0.704,P<0.01),Cys-C(r= 0.812,P<0.01)and KIM-1(r= 0.863,P<0.01). ROC curve analysis showed that the area under the curve(AUC)(0.962,95%CI:0.906~0.998)with serum Scr,Cys-C and platelet count combined in predicting ARF among septic patients was greater than the AUC with any of the indicators alone(P<0.05),and so were the sensitivity(97.0%)and specificity(91.4%).Conclusion:Early reduction in platelet counts is of important significance in predicting complication with ARF in septic patients.