慢性阻塞性肺疾病患者发生胃食管反流的风险预测模型构建及验证
| 作者: |
1蔡书宾,
1张伟,
1李际强
1 广东省中医院大学城医院综合三科，广州 510006 |
| 通讯: |
李际强
Email: lijiqiangjizhen@163.com |
| DOI: | 10.3978/j.issn.2095-6959.2018.04.017 |
| 基金: | 广东省中医药局科研项目(20172073)。 |
摘要
目的：探讨慢性阻塞性肺疾病(chronic obstructive pulmonary disease，COPD)患者发生胃食管反流病(gastroesophageal reflux disease，GERD)的风险因素，并构建有效的风险预测模型。方法：收集2013年3月至2015年10月在广东省中医院大学城医院诊疗的335例COPD患者的病例资料，采用单因素分析和logistic回归分析建立风险预测模型，采用Hosmer-Lemeshow χ2检验和受试者工作特征(receiver operating characteristic，ROC)曲线验证风险预测模型的鉴别效度和校准度，并根据已建立的预测模型中各危险因素的β值及其变量类型建立GERD发生风险的评分标准。结果：共82例(24.48%)COPD患者发生GERD，多因素logistic回归分析结果显示，活动受限(OR=3.569，95%CI：1.327~9.602)、饮酒(OR=4.678，95%CI：1.744~12.548)、高脂肪饮食(OR=5.089，95%CI：1.235~5.939)、急性发作次数(OR=6.209，95%CI：2.226~17.315)、进食过饱(OR=6.464，95%CI：2.179~19.174)、病情分级(OR=7.481，95%CI：2.179~19.174)、便秘(OR=6.464，95%CI：2.179~19.174)、进食过快(OR=6.464，95%CI：2.586~21.647)、平卧位作息(OR=3.063，95%CI：1.010~9.387)、使用β2受体激动剂(OR=18.650，95%CI：4.312~83.230)、使用激素(OR=25.142，95%CI：5.799~61.419)和使用茶碱药物(OR=31.555，95%CI：14.473~72.452)为发生GERD的独立危险因素；Hosmer-Lemeshow χ2=3.079(P=0.563)，ROC曲线下面积为0.889(95%CI：0.846~0.936)，模型的鉴别效度和校准度较好；建立的评分标准可以将患者划分为低风险(0~11分)、中风险(12~23分)和高风险(≥24分)3个组别。结论：活动受限等12个因素是影响COPD发生GERD的独立危险因子，建立的风险模型具有良好的鉴别效度和校准度，通过该模型建立的评分标准可为COPD患者预防GERD的发生提供参考依据。
关键词：
慢性阻塞性肺疾病；胃食管反流；风险预测模型
Construction and validation of a risk prediction model for gastroesophageal reflux in patients with chronic obstructive pulmonary disease
CorrespondingAuthor: LI Jiqiang Email: lijiqiangjizhen@163.com
DOI: 10.3978/j.issn.2095-6959.2018.04.017
Foundation: This work was supported by the Scientific Research Project from Guangdong Provincial Administration of Traditional Chinese Medicine, China (20172073).
Abstract
Objective: To explore the risk factors of gastroesophageal reflux in chronic obstructive pulmonary disease (COPD) and to construct an effective risk prediction model. Methods: The clinical data of 335 patients with chronic obstructive pulmonary disease in our hospital from March 2013 to October 2015 were collected, single factor analysis and logistic regression analysis were used to establish a risk prediction model; Hosmer-Lemeshow Chi square test and receiver operating characteristic (ROC) curve were used to validate the risk prediction model of differential validity and calibration, and according to the established prediction model of the risk in the beta value and variable type factors to establish standard for evaluation GERD risk. Results: A total of 82 cases (24.5%) of gastroesophageal reflux in patients with chronic obstructive pulmonary disease, multi factor logistic regression analysis showed that limited activity (OR=3.569, 95% CI: 1.327, 9.602), drinking (OR=4.678, 95% CI: 1.744–12.548), high fat diet (OR=5.089, 95% CI: 1.235–5.939), the number of acute attack(OR=6.209, 95% CI: 2.226–17.315), eating too much (OR=6.464, 95% CI: 2.179–19.174), grade of disease (OR=7.481, 95% CI: 2.179–19.174), constipation (OR=6.464, 95% CI: 2.179–19.174), eating too fast (OR=6.464, 95% CI: 2.586–21.647), supine rest (OR=3.063, 95% CI: 1.010–9.387), beta 2 receptor agonist (OR=18.650, 95% CI: 4.312–83.230), the use of hormone (OR=25.142, 95% CI: 5.799–61.419) and the use of theophylline drugs (OR=31.555, 95% CI: 14.473–72.452) as the independent risk factors for the occurrence of gastroesophageal reflux; Hosmer-Lemeshow Chi square test value is 3.079 (P=0.563), the area under the ROC curve was 0.889(95% CI: 0.846–0.936). The model has good discriminant validity and calibration, and the patients were divided into low risk (0–11 points), moderate risk (12–23 points) and high risk (≥24 points) 3 groups. Conclusion: Twelve factors such as limited activity are independent risk factors affecting the occurrence of gastroesophageal reflux in chronic obstructive pulmonary disease. The established risk model has good discriminative validity and calibration. The scoring standard established by this model can provide reference for prevention of gastroesophageal reflux in patients with chronic obstructive pulmonary disease.
Keywords：
chronic obstructive pulmonary disease; gastroesophageal reflux; risk prediction model