【摘要】 目的 探究门诊结肠镜检查患者肠道准备不充分的预测风险模型，且根据风险模型制定针对性优化策略。方法 选取2023年1月至2025年3月南方医科大学南方医院门诊结肠镜检查患者714例，按7∶3的比例随机分为训练集（n=500）和验证集（n=214）。收集入组患者资料，采用单因素及多因素Logistic回归分析肠道准备不充分的影响因素，构建风险预测模型。结果 训练集患者中，肠道准确合格者382例，不合格者118例。单因素分析结果表明，两组年龄、身高、BMI、体重、文化程度、婚姻状态、费用类型、饮酒史、衰弱、检查前粗纤维饮食、呕吐、对比差异无统计学意义（P > 0.05）。两组吸烟史、腹部手术史、便秘史、等待间隔时长、末次排便性状、实际服用泻药剂量对比差异有统计学意义（P<0.05）。Logistic回归分析显示，吸烟史、腹部手术史、便秘史、等待间隔时长、末次排便性状、实际服用泻药剂量是肠道准备失败的影响因素（P<0.05）。ROC曲线显示，基于影响因素构建的预测模型在训练集（AUC=0.894,95%CI:0.869～0.919）和验证集（AUC=0.831,95%CI:0.788～0.874）中均能很好地识别肠道准备不充分者。利用Bootstrap法重复取样1 000次进行验证，模型在训练集（AUC 0.894,95%CI:0.869～0.919）和验证集的拟合度良好（P均> 0.05）。结论 基于吸烟史、腹部手术史、便秘史、等待间隔时长、末次排便性状、实际服用泻药剂量构建门诊结肠镜检查患者肠道准备不充分的预测模型具有良好的区分度，有利于尽早筛查肠道准备不充分的高危风险人群，并制定针对性优化策略，进而能够有效预防肠道准备不充分导致的不必要重复检查，提升患者体验感。更多还原

【Abstract】 Objective To explore the predictive risk model for insufficient intestinal preparation in outpatients undergoing colonoscopy, and to formulate targeted optimization strategies based on the risk model. Methods A total of 714 patients who underwent colonoscopy in the outpatient department of Nanfang Hospital, Southern Medical University from January 2023 to March 2025 were selected and randomly divided into the training set(n = 500) and the validation set(n = 214) at a ratio of 7∶3. The data of the enrolled patients were collected. Univariate and multivariate Logistic regression analyses were used to analyze the influencing factors of insufficient intestinal preparation, and a risk prediction model was constructed. Results Among the patients in the training set, 382 cases had accurate and qualified intestinal conditions, and 118 cases had unqualified ones. The results of univariate analysis indicated that there were no statistically significant differences between the two groups in terms of age, height, BMI, weight, educational level, marital status, type of expenses, history of alcohol consumption, frailty, crude fiber diet before examination, and vomiting(P > 0.05). There were statistically significant differences in the comparison of smoking history, abdominal surgery history, constipation history, waiting interval length, last defecation characteristics, and actual dosage of laxatives between the two groups(P < 0.05). Logistic regression analysis showed that smoking history, abdominal surgery history, constipation history, waiting interval length, last defecation characteristics, and actual dosage of laxatives were the influencing factors of intestinal preparation failure(P < 0.05). The ROC curve shows that the prediction model constructed based on influencing factors can well identify those with insufficient intestinal preparation in both the training set(AUC = 0.894, 95% CI: 0.869 ～ 0.919) and the validation set(AUC = 0.831, 95% CI: 0.788 ～ 0.874), as shown in Figure 1. The Bootstrap method was used for repeated sampling 1,000 times for verification. The model was in the training set(AUC = 0.894) The 95% CI(0.869 ～ 0.919) and the fit of the validation set were good(all P > 0.05). Conclusion The predictive model for insufficient intestinal preparation in outpatients undergoing colonoscopy was constructed based on smoking history, abdominal surgery history, constipation history, waiting interval length, last defecation characteristics, and actual dosage of laxatives, which had good discrimination. It is beneficial for early screening of high-risk groups with insufficient bowel preparation, formulation of targeted optimization strategies, effective prevention of unnecessary repeated examinations caused by insufficient bowel preparation, and improvement of patients’ experience.更多还原