文献速递消化、代谢与肾脏急症重症综合管理与镇静镇痛

重症急性胰腺炎合并肠功能障碍多预测模型的建立

[Establishment of multiple predictor models of severe acute pancreatitis with intestine functional disturbance].
PubMed 文献 · Zhonghua Wei Zhong Bing Ji Jiu Yi Xue · 2019-01-01
一句话结论重症急性胰腺炎患者入院4小时内hs-CRP、BUN、SCr、Ca、PCT、NLR为合并肠功能障碍的独立预测因素,加权评分模型AUC 0.959,预测效能优于各单一指标。

该回顾性研究纳入2015年1月至2019年3月北京世纪坛医院消化科连续收治的149例重症急性胰腺炎(SAP)患者,按发病48小时是否发生肠功能障碍(IFD)分为IFD组45例和对照组104例,比较入院4小时内临床指标。单因素及多分类Logistic回归显示hs-CRP、BUN、SCr、Ca、PCT、NLR为SAP合并IFD的独立预测因素。基于上述指标建立非加权评分模型(unwScore,界值3分,AUC 0.944,敏感度95.6%,特异度94.2%)和加权评分模型(wScore,AUC 0.959,敏感度95.9%,特异度96.2%),wScore预测效能优于各单一指标及unwScore模型。

要点

  • SAP患者入院4小时内hs-CRP、BUN、SCr、Ca、PCT、NLR为合并IFD的独立预测因素。
  • 非加权评分模型界值3分,AUC 0.944,敏感度95.6%,特异度94.2%,阳性预测值87.8%,阴性预测值98.0%。
  • 加权评分模型AUC 0.959,敏感度95.9%,特异度96.2%,阳性预测值91.5%,阴性预测值98.1%,预测效能优于各单一指标及非加权模型。
  • 非加权模型评分越高,发生IFD的概率越高。

适用人群与场景

  • 2015年1月至2019年3月北京世纪坛医院消化科连续收治的重症急性胰腺炎(SAP)患者,共149例(男87例、女62例,年龄52.8±18.1岁)。
  • 按发病48小时是否发生肠功能障碍(IFD)分为IFD组45例和对照组104例。

诊断标准与关键阈值

  • 以发病48小时发生肠功能障碍(IFD)为分组标准。
  • 非加权评分模型预测IFD的界值为3分。

研究设计与关键结果

  • 回顾性收集149例SAP患者临床资料,比较IFD组45例与对照组104例入院4小时内指标。
  • 单因素及多分类Logistic回归分析显示hs-CRP、BUN、SCr、Ca、PCT、NLR为SAP合并IFD的独立预测因素。
  • 非加权评分模型(unwScore)界值3分,AUC 0.944,敏感度95.6%,特异度94.2%,阳性预测值87.8%,阴性预测值98.0%;评分越高IFD概率越高。
  • 加权评分模型(wScore)AUC 0.959,敏感度95.9%,特异度96.2%,阳性预测值91.5%,阴性预测值98.1%,预测效能优于各独立指标及unwScore模型。

证据类型、等级与局限

  • 证据类型:回顾性临床研究。
  • 局限:原文未报告局限性。

对床旁实践的改变

  • 作者认为,SAP合并IFD的多预测模型具有较好的预测效能,可能为预测和治疗提供有价值的临床参考。

提炼说明

  • 本条依据论文摘要整理;摘要未涉及「处置流程与药物剂量」「红旗征与不能漏的情况」「指南或共识出处」,故未列出这些栏目;全文请点「阅读原文」。
英文原文(摘要)· 点击展开

OBJECTIVE: To investigate the factors related to severe acute pancreatitis (SAP) with intestine functional disturbance (IFD) and to establish the multiple predictor models of SAP with IFD.

METHODS: Clinical data of consecutive SAP patients admitted to department of gastroenterology of Beijing Shijitan Hospital, Capital Medical University from January 2015 to March 2019 were retrospectively collected and analyzed. According to the occurrence of IFD at 48 hours after onset, the patients were divided into IFD group and control group. The clinical indicators within 4 hours after admission were compared between the two groups, and the independent predictive factors for SAP with IFD were screened by single factor analysis and multiple classified Logistic regression analysis. The unweighted predictive score (unwScore) and weighted predictive score (wScore) models were constructed by combining the independent predictors. The receiver operating characteristic (ROC) curves of SAP patients with IFD were plotted by independent predictive factors and predictive models, and the clinical predictive effect of each independent predictive index and predictive models were analyzed.

RESULTS: A total of 149 patients with SAP were enrolled, including 87 males and 62 females, with age of (52.8±18.1) years old. There were 45 patients in IFD group and 104 patients in control group.Univariate analysis and multiple classified Logistic regression analysis showed that high sensitive C-reactive protein (hs-CRP), blood urea nitrogen (BUN), serum creatinine (SCr), serum calcium (Ca), procalcitonin (PCT) and neutrophil-lymphocyte ratio (NLR) were independent predictive factors of SAP with IFD. The ROC curve was used to calculate the cut-off value of the above indexes to predict IFD, and unwScore model was established. The cut-off score of IFD prediction by the unwScore model was 3 points, and the probability of IFD increased with the increase of the score. The area under ROC curve (AUC) of unwScore was 0.944, the sensitivity was 95.6%, the specificity was 94.2%, the positive predictive value (PPV) was 87.8%, and the negative predictive value (NPV) was 98.0%. The binary Logistic regression analysis of hs-CRP, BUN, Ca, SCr, PCT and NLR were carried out, and wScore model was established. The AUC of wScore was 0.959, the sensitivity was 95.9%, the specificity was 96.2%, the PPV was 91.5%, and the NPV was 98.1%; predictive value was superior to each independent index and unwScore model.

CONCLUSIONS: hs-CRP, BUN, SCr, Ca, PCT and NLR were independent predictive factors of SAP with IFD. The multiple predictor models of SAP with IFD have a good predictive efficiency which may provide valuable clinical reference for prediction and treatment.

来源:Zhonghua Wei Zhong Bing Ji Jiu Yi Xue,查看原文。版权归原作者所有。

参考来源

  1. Guo C, Liu H, Tai W, et al. [Establishment of multiple predictor models of severe acute pancreatitis with intestine functional disturbance]. Zhonghua Wei Zhong Bing Ji Jiu Yi Xue. 2019. PMID: 31771726. doi:10.3760/cma.j.issn.2095-4352.2019.10.016

供执业医师抢救与诊疗参考,不替代临床判断;剂量请结合体重、肝肾功能和本院药品规格复核,以指南原文为准。