标题：脓毒症患者继发持续性炎症-免疫抑制-分解代谢综合征的影响因素及预测模型构建

      作者：杨蓉 1，王鹏 1，屈文静 2    延安市人民医院重症医学科 1、急诊科 2，陕西 延安 716000

      卷次： 2021年32卷17期

      【摘要】 目的 探讨脓毒症患者继发持续性炎症-免疫抑制-分解代谢综合征(PICS)的影响因素，并构建PICS发生风险的列线图预测模型，以评估其预测效果。方法 选取2017年9月至2020年9月在延安市人民医院重症加强护理病房(ICU)住院治疗的脓毒症患者420例，根据PICS诊断标准并以28 d作为观察终点，将117例发生PICS的患者纳入PICS组，余303例患者纳入非PICS组。收集两组患者的临床资料并进行单因素及多因素分析，以确定脓毒症患者发生PICS的独立危险因素，并将其纳入R3.6.3软件以构建预测脓毒症患者发生PICS的列线图模型；绘制ROC曲线及校准曲线图，用于评估列线图模型预测脓毒症患者发生PICS的区分度和一致性。结果 PICS组患者的年龄、急性生理学和慢性健康状况评价Ⅱ(APACHEⅡ)、住 ICU时间及机械通气比例明显高于非 PICS组，而CD4+/CD8+比值明显低于非 PICS组，差异均有统计学意义(P<0.05)；多因素分析结果显示，APACHEⅡ评分(OR=1.094，95% CI=1.042~1.149)、住 ICU时间 (OR=1.111，95% CI=1.046~1.180)、CD4 +/CD8 +比值 (OR=1.224，95% CI=1.055~1.421)、机械通气(OR=1.682，95%CI=1.064~2.659)均是影响脓毒症患者发生PICS的独立危险因素(P<0.05)；列线图模型预测脓毒症患者PICS发生的曲线下面积为0.755 (95%CI=0.704~0.806)，Hosmer-Lemeshow拟合优度检验=8.994，P=0.306，且预测PICS的校准曲线斜率接近 1。结论 本研究基于APACHEⅡ评分、住 ICU时间、CD4+/CD8+比值、机械通气构建的列线图模型，其预测脓毒症患者PICS发生的效果较好。
      【关键词】 脓毒症；持续性炎症-免疫抑制-分解代谢综合征；危险因素；列线图
      【中图分类号】 R631 【文献标识码】 A 【文章编号】 1003—6350（2021）17—2182—04

Influencing factors and predictive model construction of persistent inflammation, immunosuppression andcatabolism syndrome secondary to patients with sepsis.
YANG Rong 1, WANG Peng 1, QU Wen-jing 2. Department ofIntensive Medicine 1, Department of Emergency 2, Yan'an People's Hospital, Yan'an 716000, Shaanxi, CHINA
【Abstract】 Objective To explore the influencing factors of persistent inflammation, immunosuppression andcatabolism syndrome (PICS) secondary to patients with sepsis, and construct a nomogram prediction model for PICSrisk to evaluate its predictive effect. Methods A total of 420 patients with sepsis were selected, who were hospitalizedin the intensive care unit (ICU) of Yan'an People's Hospital from September 2017 to September 2020. According to thediagnostic criteria of PICS and 28 days as the observational endpoint, 117 patients with PICS were included in the PICSgroup, and the remaining 303 patients were included in the non-PICS group. The clinical data of the two groups were col-lected, and the independent risk factors for PICS secondary to patients with sepsis were determined by univariate andmultivariate analysis, then the data were incorporate into R3.6.3 software to construct a nomogram model to predict theoccurrence of PICS secondary to patients with sepsis. ROC curve and calibration curve were drawn to evaluate the dis-crimination and consistency of nomogram model to predict the occurrence of PICS secondary to patients with sepsis.Results The age, acute physiology and chronic health evaluation (APACHE) Ⅱ, ICU stay, and the proportion of me-chanical ventilation in the PICS group were significantly higher than those in the non-PICS group (P<0.05), and the ratioof CD4 +/CD8 + was significantly lower in the PICS group than in the non-PICS group (P<0.05). Multivariate analysisshowed that APACHEⅡ score (OR=1.094, 95% CI =1.042-1.149), ICU stay (OR=1.111, 95% CI=1.046-1.180), CD4+/CD8+ ratio (OR=1.224, 95% CI =1.055-1.421) and mechanical ventilation (OR=1.682, 95% CI = 1.064-2.659) were in-dependent risk factors for PICS secondary to sepsis patients (P<0.05). The nomogram model predicts that the area undercurve of PICS secondary to patients with sepsis was 0.755 (95% CI=0.704-0.806), Hosmer-Lemeshow goodness of fittest showed 8.994, P=0.306, and the slope of calibration curve for predicting PICS was close to 1. Conclusion In thisstudy, a nomogram model constructed based on APACHEⅡ score, ICU stay, CD4+/CD8+ ratio, and mechanical ventila-tion has a better effect on predicting the occurrence of PICS secondary to patients with sepsis.
      【Key words】 Sepsis; Persistent inflammation immunosuppression catabolism syndrome; Risk factors; Nomogram