中国实用儿科杂志

中国实用儿科杂志 ›› 2024, Vol. 39 ›› Issue (10): 768-773.DOI: 10.19538/j.ek2024100612

• 论著 • 上一篇    下一篇

儿童重症急性胰腺炎的预测因素分析

  

  1. 国家儿童医学中心（北京）  首都医科大学附属北京儿童医院消化科，北京  100045
  • 出版日期:2024-10-06 发布日期:2024-11-06
  • 通讯作者: 吴捷，电子信箱：wujie_0901@163.com
  • 基金资助:
    北京市医院管理中心“登峰”计划（DF20221003）

Predictive factors of severe acute pancreatitis in children

  1. Department of Gastroenterology，Beijing Children′s Hospital Affiliated to Capital Medical University；National Center for Children′s Health，Beijing  100045，China
  • Online:2024-10-06 Published:2024-11-06

摘要： 目的    寻找预测儿童重症急性胰腺炎（severe acute pancreatitis，SAP）有价值的临床指标，建立预测SAP的模型。方法    纳入2016年1月至2022年6月于首都医科大学附属北京儿童医院确诊的172例急性胰腺炎患儿，参照《中国急性胰腺炎诊治指南（2019年，沈阳）》，将172例患儿分为SAP组（43例）和非重症急性胰腺炎（NSAP）组（129例），比较两组一般资料（性别、年龄、体重指数、起病时间、病因、体征、住院时间、预后）、实验室指标［血常规、C反应蛋白（CRP）、血生化、凝血功能、降钙素原、红细胞沉降率等］，影像学检查指标（胰腺肿胀、胰周积液、胸腔积液、腹腔积液深度等）。比较两组间有统计学意义的临床指标，并进行多因素Logistic回归分析筛选出预测SAP的独立因子，建立Logistic回归模型，通过绘制ROC曲线、计算曲线下面积评估其预测价值。结果    172例胰腺炎患儿中，饮食不当63例（36.6%），胰胆管发育异常23例（13.4%），基因突变12例。170例（98.8%）患儿存在腹痛，SAP组更易出现发热、恶心、呕吐、腹胀以及嗜睡、烦躁等神经系统症状。血常规方面，SAP组白细胞（WBC）、中性粒细胞、淋巴细胞、中性粒细胞与淋巴细胞比值、血红蛋白、红细胞分布宽度（RDW）均更高；血生化方面，血尿素氮、肌酐、血糖、胆固醇、甘油三酯、乳酸脱氢酶（LDH）、丙氨酸氨基转移酶、天冬氨酸氨基转移酶均更高，白蛋白、血钙水平更低；凝血功能方面，凝血酶原时间、活化部分凝血活酶时间均延长，D-二聚体更高；炎症指标方面，CRP、降钙素原更高，红细胞沉降率更快。两组在胰腺肿胀、腹腔积液深度、胸腔积液、胰周积液的异常率差异有统计学意义，且SAP组后期出现胰腺假性囊肿发生率更高。应用Logistic回归分析最终筛选出CRP、WBC、RDW、LDH为SAP的独立预测因子，logit（P）=-19.151+0.043 CRP（mg/L）+0.295WBC（×109/L）+0.007LDH（U/L）+0.694RDW（%）。该预测模型曲线下面积为0.987（95%CI 0.973～1.000）。结论    CRP、WBC、RDW、LDH可作为SAP的独立预测因子，构建的模型能够预测儿童SAP发生，模型有助于临床医生早期的干预。

关键词: 儿童, 重症急性胰腺炎, 临床特征, 预测模型

Abstract: Objective    To find valuable clinical indicators for severe acute pancreatitis（SAP）and establish a predictive model for SAP. Methods    A total of 172 patients with acute pancreatitis diagnosed in Beijing Children's Hospital Affiliated to Capital Medical University from January 2016 to June 2022 were enrolled.According to the Chinese guidelines for the diagnosis and treatment of pancreatitis（2019·Shenyang），the 172 patients were divided into SAP group（43 cases）and non-severe acute pancreatitis（NSAP）group（129 cases）. The general data ［gender，age，body mass index（BMI），onset time，etiology，signs，hospitalization time，prognosis］，laboratory indicators［blood routine，C-reactive protein（CRP），blood biochemistry，coagulation function，procalcitonin，erythrocyte sedimentation rate，etc.］ and imaging examination（pancreatic swelling，peripancreatic effusion，pleural effusion，ascites depth，etc.）of the two groups were compared. Compare the clinical indicators with statistically significant differences between the two groups.Multivariate Logistic regression analysis was performed on these indicators to screen out independent factors for predicting SAP.A Logistic regression model was established and its ability to predict SAP was evaluated by plotting the ROC curve and calculating the area under the curve. Results    In all the 172 patients，63 cases（36.6%）had improper diet，23 cases（13.4%）had pancreaticobiliary dysplasia，and 12 cases had gene mutations. Totally 170 cases（98.8%）had abdominal pain，and the SAP group was more likely to have fever，nausea，vomiting，abdominal distension and nervous system manifestations such as drowsiness and irritability.SAP group had higher values of white blood cell（WBC），neutrophil，lymphocyte，neutrophil-lymphocyte ratio（NLR），hemoglobin，and red blood cell distribution width（RDW）in blood routine；SAP group had higher blood urea nitrogen，creatinine，blood glucose，cholesterol，triglyceride，lactate dehydrogenase（LDH），alanine aminotransferase and aspartate aminotransferase， and lower albumin and blood calcium levels in blood biochemistry；SAP group had longer PT and APTT，and higher D-dimer levels in coagulation function；SAP group had higher CRP and procalcitonin levels，and faster erythrocyte sedimentation rate in inflammatory indicators. There were statistically significant differences in the abnormal rates of pancreatic swelling，ascites depth，pleural effusion，and peripancreatic effusion between the two groups.The incidence of pancreatic pseudocyst was higher in the late stage in the SAP group.Logistic regression analysis showed that CRP，WBC，RDW and LDH were independent predictors of SAP，logit（P）=-19.151+0.043 CRP（mg/L）+0.295WBC（×109/L）+0.007LDH（U/L）+0.694RDW（%）.The area under the curve（AUC）of the prediction model was 0.987（95% confidence interval was 0.973-1.000）. Conclusion    CRP，WBC，RDW and LDH can be used as independent predictors of SAP.The established model can predict the incidence of acute severe pancreatitis，which is helpful for clinicians to make early intervention and treatment.

Key words: child, severe acute pancreatitis, clinical features, prediction model