构建脓毒症诱发肠黏膜屏障功能损伤的风险因素模型，并验证其应用效果。

纳入东莞市人民医院2021年3月至2023年9月收治的351例脓毒症患者，按照7∶3比例分别纳入训练组（n=246）、验证组（n=105）。使用急性胃肠损伤分级（AGI）评估其肠黏膜屏障功能损伤情况，记录患者胃肠功能异常率，并比较训练组胃肠功能正常、胃肠功能异常患者临床资料差异。使用Logistic多因素回归分析，归纳脓毒症诱发肠黏膜屏障功能损伤的风险因素，并将风险因素纳入Nomogram预测模型；绘制预测模型预测验证组患者肠黏膜屏障功能损伤的受试者工作特征曲线，使用Bootstarp法和临床决策曲线分析（DCA）验证模型校准度。

训练组、验证组肠黏膜功能损伤发生率分别为52.85%、54.29%，差异无统计学意义（P>0.05）。Logistic多因素分析示，急性生理与慢性健康状况评分Ⅱ、C反应蛋白、降钙素原、血乳酸升高，以及并发心力衰竭、低血压、休克和机械通气，均为影响脓毒症诱发肠黏膜屏障功能损伤的独立危险因素（P<0.05）。基于风险因素构建的Nomogram模型应用于训练组、验证组的曲线下面积分别为0.854、0.810，应用于验证组的灵敏度、特异性分别为80.70%、81.25%；DCA分析示模型临床净收益率较高。

脓毒症诱发肠黏膜屏障损伤风险较高，且与患者病情、炎症反应、并发症、治疗策略等因素有关，基于上述因素建立的预测模型能够为患者肠黏膜屏障功能损伤风险评估提供可靠参考。

To construct a risk factor model of intestinal mucosal barrier function damage induced by sepsis and verify its application effect.

A total of 351 patients with sepsis admitted to Dongguan People's Hospital from March 2021 to September 2023 were divided into training group (n=246) and verification group (n=105) according to the ratio of 7∶3. Acute Gastrointestinal Injury Classification (AGI) was used to evaluate the damage of intestinal mucosal barrier function, and the abnormal rate of gastrointestinal function was recorded, and the clinical data of patients with normal gastrointestinal function and abnormal gastrointestinal function in training group were compared. Logistic multivariate regression analysis was used to summarize the risk factors of intestinal mucosal barrier function damage induced by sepsis, and the risk factors were included in Nomogram prediction model. A prediction model to predict the receiver operating characteristic curve (ROC) of intestinal mucosal barrier function damage in the validation group was drawn, Bootstarp method and clinical decision curve analysis (DCA) were used to verify model calibration.

The incidence of intestinal mucosal dysfunction in training group and validation group was 52.85% and 54.29%, respectively, and there was no statistically significant difference between the two groups (P>0.05). Logistic multivariate analysis showed that elevated acute physiological and chronic health status scores Ⅱ score, C-reactive protein, procalcitonin, blood lactic acid, heart failure, hypotension, shock and mechanical ventilation were all independent risk factors for intestinal mucosal barrier function damage induced by sepsis (P<0.05). The area under the curve of Nomogram model based on risk factors was 0.854 and 0.810 when applied to training group and verification group, and the sensitivity and specificity when applied to verification group were 80.70% and 81.25%, respectively. DCA analysis shows that the clinical net yield of the model was high.

The risk of intestinal mucosal barrier injury induced by sepsis is high, which is related to the patient's condition, inflammatory reaction, complications and treatment strategies. The prediction model based on the above factors can provide reliable reference for the risk assessment of intestinal mucosal barrier function injury in patients.