慢性萎缩性胃炎患者肠化或异型增生的影响因素分析及预测模型的建立

Influencing Factors of Intestinal Metaplasia or Atypical Hyperplasia in Chronic Atrophic Gastritis Patients and Prediction Model Establishment

目的:探讨慢性萎缩性胃炎(CAG)患者发生肠化或异型增生的影响因素,同时构建其预测模型。方法:收录2016年6月至2021年6月在安徽中医药大学第二附属医院脾胃科收治的335例CAG患者的临床病例资料及实验室检查资料。通过SPSS 26.0中单因素及多因素Logistic回归分析方法确定CAG患者发生肠化或异型增生的影响因素,根据影响因素构建CAG患者发生肠化或异型增生的预测模型;另选取2019年6月至2021年6月于安徽医科大学第一附属医院就诊并确诊的115例CAG患者作为外部验证样本对构建的预测模型的预测效能进行验证评估。结果:多因素Logistic回归分析显示,胃蛋白酶原Ⅰ[比值比(OR)0.994,95%置信区间(95%CI)(0.990,0.999),P<0.05]、病灶数量[OR 6.765,95%CI(3.831,11.945),P<0.01]及幽门螺杆菌(Hp)感染[OR 0.546,95%CI(0.335,0.888),P<0.05]是CAG患者发生肠化或异型增生的独立影响因素。构建出CAG患者发生肠化或异型增生的预测模型公式为P=-1.558+0.606×Hp感染-0.006×胃蛋白酶原Ⅰ含量+1.912×病灶数量。受试者工作特征(ROC)曲线显示,曲线下面积为0.76,约登指数为0.443,最佳截断值为0.52,敏感度为0.533,特异度为0.910。将验证组患者相关因素的数据代入预测模型进行验证,利用决策曲线分析法(DCA)检验该模型的预测效能,结果证实,模型拟合优度好,预测价值高。结论:胃蛋白酶原Ⅰ、病灶数量及Hp感染是CAG患者肠化或异型增生的独立影响因素,据此构建的预测模型拟合优度好,预测价值高,可以为CAG患者病情分级及其个体化治疗方案的制定提供依据。

Objective: To investigate the influencing factors of intestinal metaplasia or atypical hyperplasia in chronic atrophic gastritis(CAG)patients and establish a prediction model. Method: The clinical records and laboratory examination data of 335 CAG patients treated in the department of gastroenterology of the Second Affiliated Hospital of the Anhui University of Chinese Medicine from June 2016 to June 2021 were collected. Single and multiple Logistic regression analyses were used to explore the influencing factors of intestinal metaplasia or atypical hyperplasia in CAG patients by SPSS 26.0. A prediction model was constructed based on the data of the related influencing factors. In addition, 115 CAG patients diagnosed in the First Affiliated Hospital of Anhui Medical University from June 2019 to June 2021 were selected as external validation samples to verify and evaluate the prediction efficiency of the constructed prediction model. Result: Multiple Logistic regression analysis showed that pepsinogen Ⅰ[odds ratio(OR)0.994,95% confidence interval(CI)(0.990,0.999),P<0.05],the number of focus[OR 6.765,95% CI(3.831,11.945),P<0.01],and Helicobacter pylori(Hp)infection[OR 0.546,95% CI(0.335,0.888),P<0.05] were independent risk factors for intestinal metaplasia or atypical hyperplasia in CAG patients(P<0.05). The formula of the prediction model is as follows:P=-1.558+0.606×Hp infection-0.006×pepsinogen Ⅰ +1.912×the number of focus. The receiver operating characteristic(ROC)curve showed the specific parameters as below:the area under the ROC curve of 0.76,the Youden index of 0.443,the best cut-off value of 0.52,sensitivity of 0.533,and specificity of 0.910. The prediction model was applied to the data of patients in the validation group for validation,and the predictive efficiency of the model was tested by decision curve analysis(DCA). The results showed that the model had a good fit and high predictive value. Conclusion: Pepsinogen Ⅰ,the number of focus,and Hp infection are independent risk factors for intestinal metaplasia or atypical hyperplasia in CAG patients. The prediction model constructed based on these factors has a good fit and high predictive value,which can provide references for the classification of CAG patients and the formulation of individual treatment protocols.