摘要
目的分析幽门螺杆菌(Helicobacter pylori,H.pylori)感染患者的临床和行为习惯特征,建立H.pylori感染患者预测模型。方法选取2021年7月至2022年4月在中山市中医院、中山市东凤人民医院、中山市南区医院共1425名接受H.pylori检测者为研究对象,以检测者的临床表征、慢性基础疾病、生活习惯和饮食习惯为主要研究变量,共计63个变量进行问卷调查。采用χ2检验和二阶聚类分别对变量进行降维,最终对差异有统计学意义的变量进行Logistic回归分析。将临床表征和危险行为因素构建预测模型和建立森林图。以ROC曲线下面积和Hosmer-Lemeshow检验验证模型的准确性。结果Logistic回归分析结果显示,居住的楼层、胃胀、口气口臭、午餐在家煮、使用公筷情况、使用公筷习惯、同居家人是否感染是H.pylori感染的独立影响因素(P<0.05)。Logistic回归模型训练组和验证组的ROC曲线下面积分别为0.7334、0.7153,截断值为0.52时,约登指数最大为0.389,灵敏度为73.8%,特异度为64.8%。Hosmer-Lemeshow拟合优度检验为14.45(P=0.071)。结论通过建立风险模型,筛选出的胃胀、口气口臭是感染的特征因素,一直都不习惯使用公筷、同居家人有感染是风险因素,午餐在家煮、家庭居住4~10层、外出有使用公筷是感染的保护因素,针对这些风险采取相应的对策,对人们的预防感染及保障生命健康有重要的意义。
Objective To analyze the clinical and behavioral characteristics of patients with H.pylori infection and to establish a predictive model for patients with H.pylori infection.Methods Using a retrospective randomized case-control,a total of 1425 patients who underwent H.pylori testing in Zhongshan Hospital of Traditional Chinese Medicine,Dongfeng People′s Hospital of Zhongshan,and Zhongshan South District Hospital from Jul.2021 to Apr.2022 were selected as the study population,and a total of 63 variables were questioned using the clinical characteristics,chronic underlying diseases,living habits and eating habits of the test subjects as the main study variables.Theχ2 test and second-order clustering were used to downscale the variables,respectively,and the final binary Logistic regression was performed on the statistically significant variables.The characteristics and risk behavior factors were used to build forest plots and construct prediction models.The accuracy of the model was verified by the area under the ROC curve and the Hosmer-Lemeshow test.Results Logistic regression analysis showed that floor of residence,gastric distension,bad breath,cooking lunch at home,whether to use public chopsticks,whether to have the habit of using public chopsticks,and whether family members had infection were independent characteristic factors of H.pylori(P<0.05),and the area under the ROC curve for the training group and validation group of the Logistic regression model were 0.7334 and 0.7153,respectively.When the cut-off value was 0.52,the maximum Yoden index was 0.389,the sensitivity was 73.8%and specificity was 64.8%.The Hosmer-Lemeshow goodness-of-fit test was 14.45(P=0.071).Conclusion By establishing a risk model,the screened out gastric distention,bad breath and bad breath are the characteristic factors of infection,and people who are not used to using public chopsticks all the time and the infection of the same family are the risk factors,cook lunch at home,live on the 4th to 10th floor of the family,and use it when going out.Public chopsticks are a protective factor against infection.Taking corresponding countermeasures against these risks is of great significance to people′s prevention of infection and protection of life and health.
作者
杜结玲
袁一鸣
洪慧斯
韦翠花
卢苑香
DU Jieling;YUAN Yiming;HONG Huisi;WEI Cuihua;LU Yuanxiang(Department of Endoscopy and Treatment,Zhongshan Hospital of Traditional Chinese Medicine,Zhongshan 528401;Department of Pharmacy,Zhongshan Hospital of Traditional Chinese Medicine;Department of Science and Education,Zhongshan Hospital of Traditional Chinese Medicine;Endoscopy Room,Zhongshan South District Hospital;Endoscopy Room,Dongfeng People′s Hospital of Zhongshan,China)
出处
《胃肠病学和肝病学杂志》
CAS
2022年第9期992-998,共7页
Chinese Journal of Gastroenterology and Hepatology
基金
中山市科学技术局资助项目(2021SYF01)。
作者简介
第一作者:杜结玲,副主任护师,研究方向:临床护理。E-mail:1689854687@qq.com;通讯作者:袁一鸣,硕士,主管药师,研究方向:机器学习,信息药学研究。E-mail:king_13602583@qq.com。
