基于CT评分与炎症因子的重症急性胰腺炎预后模型构建与评价
该研究纳入2019年3月至2021年12月在河北北方学院附属第一医院收治的128例重症急性胰腺炎(SAP)患者,均给予乌司他丁联合连续性血液净化治疗。治疗前及治疗第3天检测CRP、PCT、IL-6、IL-8、TNF-α、D-二聚体,治疗第3天行腹部CT评估MCTSI与EPIC。按28天生存预后分为存活组(n=94)与死亡组(n=34),Logistic回归筛选预后危险因素并构建列线图模型。由治疗前CRP、D-二聚体及治疗后IL-6、IL-8、TNF-α与MCTSI构成的模型2,C-index、校准及决策曲线均优于不含MCTSI的模型1,且优于APACHE II与BISAP。
要点
- 治疗前CRP>140.70 mg/L、D-二聚体>2.00 mg/L,治疗后IL-6>31.28 ng/L、IL-8>31.04 ng/L、TNF-α>31.04 ng/L及MCTSI>8分均为SAP预后独立危险因素(均P<0.05)。
- 治疗前死亡组CRP、PCT、IL-6、IL-8、D-二聚体高于存活组;治疗后死亡组IL-6、IL-8、TNF-α高于存活组;存活组MCTSI与EPIC评分低于死亡组。
- 模型2(治疗前CRP、D-二聚体+治疗后IL-6、IL-8、TNF-α+MCTSI)的C-index为0.995,高于模型1的0.988,也高于APACHE II的0.833与BISAP的0.751。
- 模型2的MAE(0.017)与MSE(0.001)低于模型1(0.034、0.003);阈值概率0-0.66或0.72-1.00时模型1净获益低于模型2,0.66-0.72时模型1净获益高于模型2。
- 模型2的MAE(0.017)低于APACHE II(0.041)与BISAP(0.025),净获益高于APACHE II与BISAP。
适用人群与场景
- 2019年3月至2021年12月河北北方学院附属第一医院收治的128例确诊重症急性胰腺炎(SAP)患者,均给予乌司他丁联合连续性血液净化治疗。
- 按入院后28天生存预后分为存活组(n=94)与死亡组(n=34)。
诊断标准与关键阈值
- 治疗前CRP>140.70 mg/L为SAP预后独立危险因素(OR 8.939,95%CI 1.792-44.575,P<0.05)。
- 治疗前D-二聚体>2.00 mg/L为SAP预后独立危险因素(OR 6.369,95%CI 1.368-29.640,P<0.05)。
- 治疗后IL-6>31.28 ng/L为SAP预后独立危险因素(OR 8.546,95%CI 1.664-43.896,P<0.05)。
- 治疗后IL-8>31.04 ng/L为SAP预后独立危险因素(OR 5.239,95%CI 1.108-24.769,P<0.05)。
- 治疗后TNF-α>31.04 ng/L为SAP预后独立危险因素(OR 4.808,95%CI 1.126-20.525,P<0.05)。
- MCTSI>8分为SAP预后独立危险因素(OR 18.569,95%CI 3.931-87.725,P<0.05)。
处置流程与药物剂量
- 所有患者均给予乌司他丁联合连续性血液净化治疗。
- 治疗前及治疗第3天检测CRP、PCT、IL-6、IL-8、TNF-α、D-二聚体水平。
- 治疗第3天行腹部CT评估改良CT严重指数(MCTSI)与胰腺外炎症CT评分(EPIC)。
研究设计与关键结果
- 研究类型为预后模型构建与评价研究,单中心,纳入128例SAP患者,均接受乌司他丁联合连续性血液净化治疗;按28天生存预后分为存活组(n=94)与死亡组(n=34)。
- 治疗前死亡组CRP、PCT、IL-6、IL-8、D-二聚体水平高于存活组;治疗后死亡组IL-6、IL-8、TNF-α水平高于存活组;存活组MCTSI与EPIC评分低于死亡组。
- Logistic回归显示治疗前CRP>140.70 mg/L(OR 8.939,95%CI 1.792-44.575,P<0.05)、D-二聚体>2.00 mg/L(OR 6.369,95%CI 1.368-29.640,P<0.05)、治疗后IL-6>31.28 ng/L(OR 8.546,95%CI 1.664-43.896,P<0.05)、IL-8>31.04 ng/L(OR 5.239,95%CI 1.108-24.769,P<0.05)、TNF-α>31.04 ng/L(OR 4.808,95%CI 1.126-20.525,P<0.05)及MCTSI>8分(OR 18.569,95%CI 3.931-87.725,P<0.05)均为SAP预后独立危险因素。
- 模型2(治疗前CRP、D-二聚体+治疗后IL-6、IL-8、TNF-α+MCTSI)的C-index为0.995,高于模型1(治疗前CRP、D-二聚体+治疗后IL-6、IL-8、TNF-α)的0.988。
- 模型2的MAE(0.017)与MSE(0.001)低于模型1(0.034、0.003);阈值概率0-0.66或0.72-1.00时模型1净获益低于模型2,0.66-0.72时模型1净获益高于模型2。
- 模型2的C-index(0.995)高于APACHE II(0.833)与BISAP(0.751);MAE(0.017)低于APACHE II(0.041)与BISAP(0.025);MSE(0.001)低于APACHE II(0.002);净获益高于APACHE II与BISAP。
证据类型、等级与局限
- 证据类型:预后模型构建与评价研究(单中心),原文未给出证据等级或推荐强度。
- 局限:原文未报告局限性。
对床旁实践的改变
- 作者认为,由治疗前CRP、D-二聚体及治疗后IL-6、IL-8、TNF-α与MCTSI构成的SAP预后评估模型具有较高的区分度、精确度和临床应用价值,优于APACHE II和BISAP。
提炼说明
- 本条依据论文摘要整理;摘要未涉及「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 construct a prognostic model for severe acute pancreatitis (SAP) based on CT scores and inflammatory factors, and to evaluate its efficacy.
METHODS: 128 patients with SAP diagnosed admitted to the First Hospital Affiliated to Hebei North College from March 2019 to December 2021 were enrolled and given Ulinastatin combined with continuous blood purification therapy. The levels of C-reactive protein (CRP), procalcitonin (PCT), interleukins (IL-6, IL-8), tumor necrosis factor-α (TNF-α), and D-dimer were measured before and on the third day of treatment. An abdominal CT was performed on the third day of treatment to assess the modified CT severity index (MCTSI) and extra-pancreatic inflammatory CT score (EPIC). Patients were divided into the survival group (n = 94) and the death group (n = 34) according to the 28-day survival prognosis after admission. The risk factors for the SAP prognosis were analyzed using Logistic regression, which was then used to build nomogram regression models. The value of the model was evaluated using the concordance index (C-index), calibration curves and decision curve analysis (DCA).
RESULTS: Before treatment, the levels of CRP, PCT, IL-6, IL-8 and D-dimer in the death group were higher than those in the survival group. After treatment, the levels of IL-6, IL-8 and TNF-α in the death group were higher than those in the survival group. MCTSI and EPIC scores in the survival group were lower than those in the death group. Logistic regression analysis shows that, pre-treatment CRP > 140.70 mg/L, D-dimer > 2.00 mg/L, and post-treatment IL-6 > 31.28 ng/L, IL-8 > 31.04 ng/L, TNF-α > 31.04 ng/L, and MCTSI > 8 points were all independent risk factors for SAP prognosis [odds ratios (OR) and 95% confidence intervals (95%CI) were 8.939 (1.792-44.575), 6.369 (1.368-29.640), 8.546 (1.664-43.896), 5.239 (1.108-24.769), 4.808 (1.126-20.525), 18.569 (3.931-87.725), all P < 0.05]. Model 1 (consisting of pre-treatment CRP, D-dimer, and post-treatment IL-6, IL-8 and TNF-α) had a lower C-index than that model 2 (consisting of pre-treatment CRP, D-dimer, and post-treatment IL-6, IL-8 and TNF-α, and MCTSI; 0.988 vs. 0.995). The mean absolute error (MAE) and mean square error (MSE) of model 1 (0.034, 0.003) were higher than those of model 2 (0.017, 0.001). When the threshold probability was in the range of 0-0.66 or 0.72-1.00, the net benefit of model 1 was lower than that of model 2. When the threshold probability was in the range of 0.66-0.72, the net benefit of model 1 was higher than that of model 2. In addition, model 2 had a higher C-index than acute physiology and chronic health evaluation II (APACHE II) and bedside index of acute pancreatitis severity (BISAP, 0.995 vs. 0.833, 0.751). Model 2 had a lower MAE (0.017) and MSE (0.001) than APACHE II (0.041, 0.002). Model 2 had a lower MAE than BISAP (0.025). Model 2 had a higher net benefit than both APACHE II and BISAP.
CONCLUSIONS: The prognostic assessment model of SAP consisting of pre-treatment CRP, D-dimer, and post-treatment IL-6, IL-8 and TNF-α, and MCTSI has high discrimination, precision and clinical application value, and is superior to APACHE II and BISAP.
参考来源
- Han B, Yang W, Wang H, et al. [Construction and evaluation of a prognostic model for severe acute pancreatitis based on CT scores and inflammatory factors]. Zhonghua Wei Zhong Bing Ji Jiu Yi Xue. 2023. PMID: 36880244. doi:10.3760/cma.j.cn121430-20220411-00351
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。