降钙素原、超敏C反应蛋白、胰石蛋白联合检测对脓毒症患儿预后的预测价值

目的探讨降钙素原（procalcitonin，PCT）、超敏C反应蛋白（high sensitivity C-reactive protein，hs-CRP）、胰石蛋白（pancreatic stone protein，PSP）对评估脓毒症患儿预后的价值。方法收集我院ICU收住的确诊为脓毒症的患儿106例，平均（4.4±1.6）岁。记录患儿一般情况及小儿危重病例评分（PCIS）。入院后第1天抽取静脉血，检测所有患者血清中PCT、hs-CRP、PSP的表达情况。随访28 d，记录患儿生存情况。Pearson 相关分析检验PCT、hs-CRP、PSP与 PCIS 的相关性；多因素logistic回归分析PCT、hs-CRP、PSP以及其他危险因素与脓毒症患儿预后生存（28 d 死亡与否）的相关性。ROC曲线分析比较PCT、hs-CRP、PSP单独或联合检测在脓毒症患儿预后评估中的价值。结果 [随访28 d，34例患儿死亡，占整体的32.08%。相比于生存组，死亡组PCIS评分降低（P<0.001）；同时，死亡组患者血清中PCT、hs-CRP、PSP表达升高（P<0.001）。Pearson相关分析显示PCT与PCIS呈负相关（ r=-0.437 2，P<0.001）；hs-CRP与PCIS呈负相关（ r=-0.649 7，P<0.001）；PSP与PCIS呈负相关（ r=-0.645 3，P<0.001）。多因素logistic回归结果显示，高水平PCT〔标准偏回归系数（B’）=0.481，P=0.000〕、hs-CRP（B’=0.392，P=0.014）、PSP（B’=0.314，P=0.041），都是脓毒症患儿28 d死亡的独立危险因素。ROC曲线分析结果显示PCT、hs-CRP以及PSP曲线下面积（AUC）分别为0.86〔95%可信区间（CI），0.78～0.92〕、0.70（95%CI，0.61～0.79）以及0.69（95%CI，0.60～0.78）。PCT、hs-CRP和PSP三者联合(0.481×PCT+0.392×hs-CRP+0.314×PSP)检测的AUC值〔0.92（95%CI，0.85～0.96）〕优于单一检测PCT、hs-CRP或PSP（P<0.001）;联合检测值＜122.3时脓毒血症患儿预后好（28 d生存），≥122.3时预后差（28 d死亡）。结论血清PCT、hs-CRP、PSP均与PCIS呈负相关，且这3个血清学指标联合检测在评估脓毒症患儿预后时具有良好的预测价值。

英文摘要:

Objective To determine the value of procalcitonin (PCT), high sensitivity C-reactive protein (hs-CRP) and pancreatic stone protein(PSP) in predicting the prognosis of children with sepsis. Methods A total of 106 hospitalized children 〔(4.4±1.6) year-old〕 with sepsis were enrolled in this study. The expressions of PTC, hs-CRP and PSP in the serum samples of the children were detected on the first day of admission to hospital. Pearson correlation analyses were performed to test the correlations between pediatric critical illness score (PCIS) and PTC, hs-CRP and PSP. Logistic regression models were established to determine factors predicting death of children. The value of PTC, hs-CRP and PSP in predicting the prognosis of children with sepsis was determined using ROC curves. Results About 32% children (34 cases) died. Higher expressions of PTC, hs-CRP and PSP were found in those who died (P<0.001). Serum PTC, hs-CRP and PSP were negatively correlated with PCIS (P<0.001). The multivariate logistic regression showed that PTC, hs-CRP and PSP were independent predictors

查看全文查看/发表评论下载PDF阅读器

关闭