老年无痛胃肠镜检查患者肠道准备失败的影响因素及预测模型构建

Influencing factors of bowel preparation failure in elderly patients undergoing painless gastroenteroscopy and construction of a predictive model

【摘要】目的 探讨老年无痛胃肠镜检查患者肠道准备失败的影响因素并构建风险预测模型。方法 回顾性分析2020年6月至2023年9月在联勤保障部队第九四〇医院进行无痛胃肠镜检查的412例老年患者的临床资料,按照2∶1的比例将患者分为测试组(n=275)及验证组(n=137)。根据肠道准备质量,将测试组中患者分为两个亚组,肠道准备成功者纳入成功组(n=178),肠道准备失败者纳入失败组(n=97)。采用SPSS 20.0统计软件进行数据分析。根据数据类型,分别采用t检验或χ²检验进行组间比较。采用多因素logistic回归模型分析老年无痛胃肠镜检查患者肠道准备失败的影响因素,并以危险因素构建风险预测模型。采用Hosmer-Lemeshow检验验证预测效能。结果 成功组与失败组体质量指数、行走情况、便秘、结直肠手术史、首次进行结肠镜检查、服用三环类抗抑郁药、大量腹水、腹部/盆腔手术史、基础病种类资料比较,差异有统计学意义(P＜0.05)。多因素logistic回归分析发现,便秘(OR=4.384,95%CI 1.136～16.919)、结直肠手术史(OR=4.043,95%CI 1.392～11.743)、服用三环类抗抑郁药(OR=5.013,95%CI 1.237～20.317)及基础病种类≥2种(OR=4.973,95%CI 1.335～18.526)是老年无痛胃肠镜检查患者肠道准备失败的独立危险因素。以危险因素代入风险预测模型,发现上述因素在老年无痛胃肠镜检查患者肠道准备失败中的预测灵敏度为67.18%、特异度为85.69%,曲线下面积(AUC)为0.739。验证组检验结果发现,灵敏度为66.54%、特异度为82.37%,AUC为0.709。结论 便秘、结直肠手术史、服用三环类抗抑郁药及基础病种类≥2种均会对老年胃肠镜检查患者肠道准备质量产生影响,且以上述因素构建的风险预测模型在老年胃肠镜检查患者肠道准备失败中预测效能较佳,临床需根据实际情况提出针对性干预方案,提高肠道准备质量。

基金项目:甘肃省自然科学基金(21JR1RA183)；甘肃省卫生健康委员会资助项目(GSWSKY2021-044)

【Abstract】Objective To investigate the factors influencing bowel preparation failure in elderly patients with painless gastroentero-scopy and establish a risk prediction model. Methods A retrospective analysis was made of the clinical data of 412 elderly patients who underwent painless gastrointestinal endoscopy at 940th Hospital of Joint Logistics Support Force of Chinese People′s Liberation Army from June 2020 to September 2023. The patients were divided into a test group (n=275) and a validation group (n=137) at a ratio of 2∶1. According to the quality of bowel preparation, the test group were divided into two subgroups:a success group (n=178) and a failure group (n=97). SPSS 20.0 was used for data analysis. According to the data type, t test orχ² test was used for data comparison between groups. Multivariate logistic regression analysis was used to analyze the influencing factors of bowel preparation failure, a risk prediction model was established based on the risk factors, and tailored intervention strategies were proposed. Hosmer-Lemeshow test was used to verify the prediction efficiency. Results Significant differences were seen between the success group and the failure group in body mass index, walking status, constipation, history of colorectal surgery, first colonoscopy, use of tricyclic antidepressants, massive ascites, history of abdominal/pelvic surgery, and type of underlying diseases (P<0.05). Logistic regression analysis showed that constipation (OR=4.384,95%CI 1.136-16.919), history of colorectal surgery (OR=4.043,95%CI 1.392-11.743), and use of tricyclic antidepressants (OR=5.013,95%CI 1.237-20.317), and underlying diseases ≥2 kinds (OR=4.973,95%CI 1.335-18.526) were independent risk factor for bowel preparation failure in elderly patients with painless gastroenteroscopy. The risk prediction model incorporating these factors demonstrated a predictive sensitivity of 67.18%, a specificity of 85.69%, and an area under the receiver operating characteristic curve (AUC) of 0.739. The validation test showed a sensitivity of 66.54%, a specificity of 82.37%, and an AUC of 0.709. Conclusion Constipation, history of colorectal surgery, use of tricyclic antidepressants, and underlying disease ≥2 kinds affect the quality of bowel preparation in elderly patients with gastroenteroscopy, and the risk prediction model based on the above factors exhibits a good predictive efficacy for bowel preparation failure in this population. Tailored intervention plans should be proposed according to the actual situations in clinical practice to improve the quality of bowel preparation.

This work was supported by Natural Science Foundation of Gansu Province (21JR1RA183) and the Project of Health Commission of GansuProvince (GSWSKY2021-044).