摘要
背景据统计住院重症社区获得性肺炎(SCAP)患者病死率超过50%,而早期识别死亡风险高的SCAP患者并及时干预对改善患者预后具有重要意义。目前,临床上尚未构建预测SCAP患者死亡风险的Nomogram模型。目的基于白蛋白(Alb)-红细胞分布宽度(RDW)评分、肺炎严重程度指数(PSI)危险分级构建SCAP患者入院后90 d内预后的Nomogram模型并评估其预测价值。方法选取2017年1月—2019年1月青海省第五人民医院重症监护室收治的SCAP患者216例,根据入院后90 d内预后分为存活组(n=158)和死亡组(n=58)。收集患者入院24 h内临床资料(包括年龄、性别、吸烟情况、基础疾病、气管插管情况、脓毒性休克发生情况)和实验室检查指标〔包括白细胞计数(WBC)、中性粒细胞与淋巴细胞比值(NLR)、血小板与淋巴细胞比值(PLR)、单核细胞与淋巴细胞比值(MLR)、C反应蛋白(CRP)、血红蛋白(Hb)、血肌酐(Scr)、Alb、RDW〕,计算患者Alb-RDW评分并评估PSI危险分级。绘制受试者工作特征曲线(ROC曲线)以评估相关连续变量预测SCAP患者入院后90 d内预后的最佳截断值;采用多因素Logistic回归分析探讨SCAP患者入院后90 d内预后的影响因素,采用R版3.5.2中的“rms”软件包构建SCAP患者入院后90 d内死亡风险的Nomogram模型,并绘制ROC曲线以评估其预测价值。结果ROC曲线分析结果显示,年龄、WBC、NLR、PLR、MLR、CRP、Hb、Scr、Alb、RDW预测SCAP患者入院后90 d内预后的最佳截断值分别为65岁、12.58×10^9/L、6.24、220.54、0.61、118.52 mg/L、103.58 g/L、91.72μmol/L、29.98 g/L、14.15%。死亡组患者中年龄≥65岁、吸烟、气管插管、脓毒性休克、WBC≥12.58×10^9/L、NLR≥6.24、Hb<103.58 g/L、Scr≥91.72μmol/L、Alb<29.98 g/L、RDW≥14.15%、Alb-RDW评分1分及2分、PSI危险分级Ⅲ级及Ⅳ~Ⅴ级者所占比例均高于存活组(P<0.05)。多因素Logistic回归分析结果显示,年龄≥65岁、气管插管、NLR≥6.24、Alb<29.98 g/L、RDW≥14.15%、Alb-RDW评分1分及2分、PSI危险分级Ⅲ级及Ⅳ~Ⅴ级是SCAP患者入院后90 d内死亡的危险因素(P<0.05)。将年龄、气管插管、NLR、Alb-RDW评分、PSI危险分级作为预测因子,构建SCAP患者入院后90 d内死亡风险的Nomogram模型。ROC曲线分析结果显示,Nomogram模型预测SCAP患者入院后90 d内死亡的曲线下面积(AUC)为0.862〔95%CI(0.729,0.916)〕,分别高于Alb-RDW评分的0.742〔95%CI(0.682,0.796)〕及PSI危险分级的0.755〔95%CI(0.696,0.808)〕(P<0.05)。结论基于Alb-RDW评分、PSI危险分级构建的Nomogram模型对SCAP患者入院后90 d内死亡风险的预测价值高于单独Alb-RDW评分及PSI危险分级。
Background According to statistics,the mortality rate of hospitalized patients with severe communityacquired pneumonia(SCAP)is more than 50%,early identification of patients with high risk of death and timely intervention is of great significance to improve the prognosis of patients.At present,there is no Nomogram model to predict the death risk of patients with SCAP.Objective To construct Nomogram model for prognosis within 90 days after admission in patients with SCAP based on Alb-RDW score and pneumonia severity index(PSI)risk classification,and estimate its predictive value.Methods A total of 216 patients with SCAP in Intensive Care Unit,the Fifth People's Hospital of Qinghai Province from January 2017 to January 2019 were selected,and they were divided into survival group(n=158)and death group(n=58)according to the prognosis within 90 days after admission.The clinical data(including age,gender,smoking,underlying diseases,tracheal intubation,septic shock)and laboratory examination results(including WBC,NLR,PLR,MLR,CRP,Hb,Scr,Alb,RDW)within 24 hours after admission were collected,the Alb-RDW score was calculated and the PSI risk classification was evaluated.ROC curve was drawn to evaluate the optimum cutoff values of relevant continuous variables in predicting the prognosis within 90 days after admission in patients with SCAP;multivariate Logistic regression was used to analyze the influencing factors of prognosis within 90 days after admission in patients with SCAP,the Nomogram model of death risk within 90 days after admission was constructed by"rms"package of R(version 3.5.2),and ROC curve was drawn to evaluate its predictive value.Results ROC curve results showed that,the optimum cutoff values of age,WBC,NLR,PLR,MLR,CRP,Hb,Scr,Alb and RDW were 65 years old,12.58×10^9/L,6.24,220.54,0.61,118.52 mg/L,103.58 g/L,91.72μmol/L,29.98 g/L and 14.15%,respectively.The proportion of patients with age≥65 years,smoking,tracheal intubation,septic shock,WBC≥12.58×10^9/L,NLR≥6.24,Hb<103.58 g/L,Scr≥91.72μmol/L,Alb<29.98 g/L,RDW≥14.15%,Alb-RDW score 1 and 2,PSI risk classificationⅢandⅣ-Ⅴin death group were higher than those in survival group(P<0.05).Multivariate Logistic regression analysis showed that,age≥65 years old,tracheal intubation,NLR≥6.24,Alb<29.98 g/L,RDW≥14.15%,Alb-RDW score 1 and 2,PSI risk classificationⅢandⅣ-Ⅴwere risk factors for death within 90 days after admission in patients with SCAP(P<0.05).Age,tracheal intubation,NLR,Alb-RDW score and PSI risk classification were used as predictors to construct Nomogram model of death risk within 90 days after admission in patients with SCAP.ROC curve showed that,the AUC of Nomogram model in predicting death within 90 days after admission in patients with SCAP was 0.862〔95%CI(0.729,0.916)〕,which was higher than that of Alb-RDW score〔0.742,95%CI(0.682,0.796)〕and PSI risk classification〔0.755,95%CI(0.696,0.808)〕,respectively(P<0.05).Conclusion The predictive value of Nomogram model constructed based on Alb-RDW score and PSI risk classification is higher than that of Alb-RDW score and PSI risk classification alone for prognosis within 90 days after admission in patients with SCAP.
作者
余长升
马艳红
辛晓婷
YU Changsheng;MA Yanhong;XIN Xiaoting(Intensive Care Unit,the Fifth People's Hospital of Qinghai Province,Xining 810000,China)
出处
《实用心脑肺血管病杂志》
2020年第10期72-78,共7页
Practical Journal of Cardiac Cerebral Pneumal and Vascular Disease