临床外科杂志 ›› 2024, Vol. 32 ›› Issue (4): 425-428.doi: 10.3969/j.issn.1005-6483.2024.04.027

• 论著 • 上一篇    下一篇

急危重多发伤胃肠破裂急诊修复术后不良预后风险预测模型的构建与验证

沈辉 云天奇 代伟宏   

  1. 570311 海口,海南医学院第二附属医院急诊创伤外科
  • 收稿日期:2023-04-13 出版日期:2024-05-10 发布日期:2024-05-10
  • 基金资助:
    海南省卫生健康行业科研项目(22A200086)

Construction and validation of a risk prediction model for the poor prognosis after repair of traumatic perforation of gastrointestinal tract following emergency admission for acute and critical multiple injuries

SHEN Hui,YUN Tianqi,DAI Weihong   

  1. Department Emergency Traumatology Surgery,The Second Affiliated Hospital of Hainan Medical university,Haikou Hainan 570311,China
  • Received:2023-04-13 Online:2024-05-10 Published:2024-05-10

摘要: 目的 探讨急危重多发伤后胃肠破裂急诊修复术后不良预后风险预测模型的构建与验证。方法 选取2019年1月~2022年12月急诊收治急危重多发伤后胃肠破裂病人183例,均采取对症及修复手术方案治疗,院内随访30天,记录预后情况,死亡为主要终点,收集病人临床资料,运用logistic回归分析病人发生不良预后的风险因素,建立列线图模型,绘制ROC曲线检验模型的区分度,通过Hosmer-Lemeshow拟合优度检验和绘制校准曲线验证模型的校准度。结果 随访30天后,其中死亡28例,临床症状及体征较入院时加重38例,预后不良率为26.78%,然后分为预后不良组(49例)及预后良好组(134例)。多因素logistic回归分析结果显示,低格拉斯哥昏迷评分、休克、高D-二聚体、高损伤严重程度评分是急危重多发伤后胃肠破裂修复术后不良预后的危险因素(P<0.05)。基于此危险因素构建预测急危重多发伤后胃肠破裂修复术后发生不良预后的列线图模型,并验证模型的精准性与区分性,ROC曲线下面积为 0.986; H-L偏差度检验结果为0.984。结论 低格拉斯哥昏迷评分、休克、高D-二聚体、高损伤严重程度评分是急危重多发伤后胃肠破裂修复术后不良预后的独立危险因素,预测风险列线图模型区分度与精准性较高,在急危重多发伤后胃肠破裂预防及治疗中的应用价值较好。

关键词: 急诊; 急危重多发伤; 胃肠破裂修复术; 不良预后; 风险预测模型

Abstract: Objective To construct and verify a risk prediction model for the poor prognosis after repair of traumatic perforation of gastrointestinal tract following emergency admission for acute and critical multiple injuries.Methods 183 patients with traumatic perforation of gastrointestinal tract due to acute and critical multiple injuries underwent treatment in emergency department of our hospital from January 2019 to December 2022 were enrolled.All were treated with symptomatic and reparative surgical protocols.Prognosis was recorded at 30d of the hospital follow-up,with death as the primary endpoint.Patient clinical data were collected.Logistic regression was used to screen the risk factors of poor prognosis,and a line graph model was established.Receiver Operating Curve (ROC) was used to test model differentiation.Calibration of models was evaluated with the Hosmer-Lemeshow goodness-of-fit test and calibration curve.Results After a follow-up of 30 days,28 cases died,and 38 cases had worsening clinical symptoms and signs compared to when they were admitted.The poor prognosis rate was 26.78%,and then divided into a poor prognosis group (49 cases) and a good prognosis group (134 cases).Multivariate Logistic regression analysis denoted that low GCS score,presence of shock,elevated D-dimer level,and high ISS score were risk factors affecting the poor prognosis after repair of traumatic perforation of gastrointestinal tract (P<0.05).The above risk factors were used to construct a line graph model for predicting the occurrence of poor prognosis after surgery,and the accuracy and differentiation of the model were verified showing an area under the curve of 0.986 and Hosmer-Lemeshow deviation test result of 0.984.Conclusion The low GCS score,presence of shock,elevated D-dimer level,and high ISS score are independent risk factors for poor prognosis after repair of traumatic perforation of gastrointestinal tract following emergency admission for acute and critical multiple injuries,and the predictive risk line graph model has high differentiation and precision,which is of great value for application in the prevention and treatment of traumatic perforation of gastrointestinal tract.

Key words: emergency treatment; acute and critical multiple injuries; repair of traumatic perforation of gastrointestinal tract; adverse prognosis; risk prediction model

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!