48小时内血清降钙素原与动脉乳酸清除率对判断脓毒性休克患者预后的价值

Predictive value of the clearance rate of serum procalcitonin and arterial lactate in 48 hours in predicting the prognosis of patients with septic shock

目的探讨48 h内血清降钙素原(PCT)与动脉乳酸(LAC)清除率对判断脓毒性休克患者预后的预测价值。方法采用回顾性分析方法收集并分析2012年3月至2017年3月本院综合重症监护病房(ICU)收治的脓毒性休克患者的临床资料。结果按照第28天患者的生存情况分为存活组63例(经ICU治疗病情好转出院或转入其他科室继续治疗者)和死亡组41例(经治疗无效,在观察期间死亡者)。两组病例入住ICU时序贯性器官衰竭评估(SOFA)、急性生理与慢性健康评分Ⅱ(APACHEⅡ)及PCT、动脉LAC检测差异均无统计学意义。治疗48 h后,存活组较死亡组患者血清PCT(PCT_(48 h))显著下降[(3.31±1.05)ng/ml vs.(4.37±2.13)ng/ml,t=3.375、P=0.001]、动脉LAC水平(LAC_(48 h))显著下降[(2.11±0.55)mmol/L vs.(2.52±0.92)mmol/L,t=2.845,P=0.006]。存活组与死亡组患者PCT及动脉LAC清除率差异均有统计学意义:PCT_(c-48 h)[(49.62±5.85)%vs.(42.29±4.99)%,t=-6.607、P<0.001)及LAC_(c-48 h)[(51.27±3.44)%vs.(44.86±5.55)%,t=-7.277、P<0.001]。多因素二元logistic回归分析结果显示,PCT_(c-48 h)和LAC_(c-48 h)对脓毒性休克的预后具有一定预测价值,其对良好预后的OR值分别为1.212和1.570。以PCT_(c-48 h)和LAC_(c-48 h)作为参数拟合Logistic二元回归方程,拟合优度高(χ~2=7.986、P=0.435),得出PCT_(c-48 h)和LAC_(c-48 h)对脓毒性休克预后的预测方程为logit(P)=0.430+0.251×(PCT_(c-48 h))+0.472×(LAC_(c-48 h))。ROC曲线图显示PCT_(c-48 h)曲线下面积为0.842,LAC_(c-48 h)曲线下面积为0.901。而PCT_(c-48 h)和LAC_(c-48 h)联合预测价值最大,曲线下面积为0.944,其最大约登指数为0.819,最佳临界值为35.32%,对应的灵敏度和特异度分别为97.56%和80.95%。结论 PCT_(c-48 h)和LAC_(c-48 h)联合检测可作为脓毒性休克预后判断的简便有效指标。

Objective To evaluate the predictive value of the clearance of serum procalcitonin（PCT） and arterial lactate（LAC） in patients with septic shock in 48 hours. Methods The clinical data of patients with septic shock hospitalized in ICU in our hospital from March 2012 to March 2017 were analyzed, retrospectively. Results All patients were divided into survival group（63 cases, after treatment in ICU, patients who discharged or transferred to other departments for continue treatment） and death group（41 cases, patients with ineffective treatment and died during the observation）, according to the twenty-eighth days survival. The sequential organ failure assessment（SOFA）, acute physiology and chronic health evaluation（APACHE） Ⅱ, PCT and arterial LAC before treatment of patients in the two groups were without significant differences. The levels of PCT [（3.31 ± 1.05） ng/ml vs.（4.37 ± 2.13） ng/ml; t = 3.375, P = 0.001] and arterial LAC [（2.11 ± 0.55） mmol/L vs.（2.52 ± 0.92） mmol/L; t = 2.845, P = 0.006 ] of patients in survival group and death group after treatment of 48 hours were significantly different. The clearance rates of PCT and arterial LAC： PCTc-48 h [（49.62 ± 5.85）% vs.（42.29 ± 4.99）%; t =-6.607, P 0.001）, LACc-48 h [（51.27 ± 3.44）% vs.（44.86 ± 5.55）%; t =-7.277, P 0.001] of patients in survival group and death group weresignificantly different. The predictive value for septic shock were evaluated by multivariate Logistic regression analysis, showing that PCTc-48 h and LACc-48 h had a certain predictive value, while the OR values for good prognosis were 1.212 and 1.570. Logistic two regression equation was fitted using PCTc-48 h and LACc-48 h as parameters, and the fitting degree was high（χ2 = 7.986, P = 0.435）, the equation for septic shock was： logit（P） = 0.430 ＋ 0.251 ×（PCTc-48 h） ＋ 0.472 ×（LACc-48 h）. ROC curve was figured, and the area under PCTc-48 h curve was 0.842, and the area under LACc-48 h curve was 0.901, meanwhile the combination of PCTc-48 h and LACc-48 h for good prognosis of patients with septic shock had a higher predictive value, the area under curve was 0.944, the maximum Youden index was 0.819, the cutoff value was 35.32%, and the corresponding sensitivity and specificity were 97.56% and 80.95%. Conclusion The combined detection of PCTc-48 h and LACc-48 h could be used as a simple and effective indicator for the prognosis of patients with septic shock.