实用肝脏病杂志 ›› 2023, Vol. 26 ›› Issue (2): 242-245.doi: 10.3969/j.issn.1672-5069.2023.02.023

• 肝衰竭 • 上一篇    下一篇

常用血液学指标评定慢加急性肝衰竭患者并发细菌感染和预后价值研究*

邓珂欣, 罗红春   

  1. 400016 重庆市 重庆医科大学附属第一医院感染病科
  • 收稿日期:2022-09-28 出版日期:2023-03-10 发布日期:2023-03-21
  • 通讯作者: 罗红春,E-mail:wdlz1299@163.com
  • 作者简介:邓珂欣,女,26岁,硕士研究生,住院医师。E-mail:2981301284@qq.com
  • 基金资助:
    *重庆市自然科学基金资助项目(编号:cstc2021jcyj-msxmX0202);重庆市科卫联合医学科研项目(编号:2020MSXM099)

Evaluation of bacterial infection by common hematological indexes in patients with acute-on-chronic liver failure

Deng Kexin, Luo Hongchun   

  1. Department of Infectious Diseases, First Affiliated Hospital, Chongqing Medical University, Chongqing 400016, China
  • Received:2022-09-28 Online:2023-03-10 Published:2023-03-21

摘要: 目的 探讨常用血液学指标判断慢加急性肝衰竭(ACLF)并发细菌感染和预后的价值。方法 2015年7月~2022年7月我院收治的ACLF患者208例,给予内科综合治疗和人工肝治疗,回顾性分析其临床资料,采用二元多因素Logistic回归分析影响细菌感染的血液学指标,应用受试者工作特征曲线(ROC)评估各指标对细菌感染的判定价值,应用Cox回归分析预后。结果 在本组208例ACLF患者中,并发细菌感染143例(68.8%),其中腹腔感染占51.5%,肺部感染占22.5%,尿路感染占2.4%,败血症占0.6%,其他部位感染23.0%。119例为单部位感染,24例为多部位感染;108例(51.9%)好转出院,100例(48.1%)死亡;感染组年龄为(48.4±12.6)岁,显著大于未感染组【(43.9±12.4)岁,P<0.05】,外周血WBC、中性粒细胞计数和血清PCT水平分别为7.0(4.8,8.9)×109/L、5.0(3.3,6.7) ×109/L和0.9(0.4,1.2)ng/ml,均显著高于未感染组【分别为6.0(4.6,7.3) ×109/L、4.1(2.9,5.1) ×109/L和0.5(0.3,0.9) ng/ml,P<0.05】;感染组血清TBIL、国际标准化比值(INR)和D二聚体均显著高于未感染组(P<0.05),而RBC、Hb、白蛋白、ALT、AST、胆碱酯酶、Na+、PTA和AFP水平显著低于未感染组(P<0.05);经二元多因素Logistic回归分析发现,Hb、中性粒细胞计数、TBIL、Na+和PTA为影响ACLF患者发生细菌感染的独立因素;ROC曲线分析发现,分别以Hb≤117.5 g/L、中性粒细胞计数≥5.1×109/L、TBIL≥261.0μmol/L、Na+≤137.5 mmol/L和PTA≤34.5%为截断点,判定ACLF患者细菌感染的AUC分别为0.645、0.627、0.724、0.658和0.645,上述指标联合判断的AUC最大,为0.810,其灵敏度和特异度分别为70.6%和79.4%;经Cox多因素回归分析发现TBIL和PTA是影响ACLF患者预后的独立危险因素。结论 常用的血液学指标可辅助判断ACLF患者感染和不良预后,补充常用感染指标的不足,有助于临床诊治。

关键词: 慢加急性肝衰竭, 细菌感染, 血液学指标, 血清指标, 预后评估

Abstract: Objective The aim of this study was to explore the evaluation of bacterial infection by common hematological indexes in patients with acute-on-chronic liver failure (ACLF). Methods 208 patients with ACLF were admitted to our hospital between July 2015 and July 2022, and were given conventional supporting therapy. The clinical materials of all patients were retrospectively analyzed. The factors impacting bacterial infection was evaluated by binary multivariate Logistic analysis. The judging performance of infection by each index was evaluated by receiver operating characteristic (ROC) curves, and the Cox regression analysis was applied to predict the prognosis. Results Out of the 208 patients with ACLF in our series,the incidence of bacterial infection was 68.8%, and the common infection sites were abdominal infection in 87 cases(51.5%), pulmonary infection in 38 cases (22.5%), urinary tract infection in 4 cases (2.4%), blood infection in 1 case (0.6%) and other site infection in 39 cases (23.0%), and there were 119 cases with single site infection and 24 cases with multi-site infection; 108 patients (51.9%) were discharged recovered, and 100 patients (48.1%) died during hospitalization or after discharge; the average age in patients with bacterial infection was (48.4±12.6) years, significantly greater than in those without infection [(43.9±12.4) years, P<0.05]; the white blood cells (WBC) count, neutrophil count and serum procalcitonin (PCT) level in patients with infection were 7.0 (4.8, 8.9)×109/L, 5.0 (3.3,6.7)×109/L and 0.9 (0.4,1.2) ng/ml, all significantly higher than [6.0 (4.6, 7.3)×109/L, 4.1 (2.9, 5.1)×109/L and 0.5 (0.3, 0.9) ng/ml, respectively, P<0.05] in those without infection, while there were no significant differences as respect to serum C-reaction protein (CRP) and hypersensitive C-reactive protein (hsCRP) levels in the two groups (P > 0.05); serum total bilirubin (TBIL), international standardization ratio (INR) and D-dimer levels were significantly higher than those in patients without infection (P<0.05), while blood red blood cell (RBC) count, hemoglobin (Hb), albumin, ALT, AST, cholinesterase, Na+, PTA and alpha fetoprotein (AFP) in patients with infection were significantly lower than those in patients without infection (P<0.05); the Logistic analysis showed that Hb, neutrophil count, TBIL, Na+ and PTA were the independent influencing factors for bacterial infection in patients with ACLF; the ROC curves analysis showed that the AUCs were 0.645, 0.627, 0.724, 0.658 and 0.645 by Hb, neutrophil count, TBIL, Na+ and PTA, predicting bacterial infection in patients with ACLF, when the cut-off-value were set as 117.5 g/L, 5.1×109/L, 261.0 μmol/L, 137.5 mmol/L and 34.5%, while the combination of the above indicators had the largest AUC of 0.810, with the sensitivity and specificity of 70.6% and 79.4%, respectively; the Cox multivariate analysis showed that TBIL and PTA were the independent risk factors for the poor prognosis of patients with ACLF. Conclusion Under the circumstance of ACLF, some common blood parameters might help judge infection, and warrants emphasis in clinical practice.

Key words: Acute-on-chronic liver failure, Bacterial infection, Hematological index, Serum index, Auxiliary diagnosis, Prognosis