天津医药 ›› 2026, Vol. 54 ›› Issue (3): 249-253.doi: 10.11958/20252207

• 临床研究 • 上一篇    下一篇

基于息肉特征和血清学指标构建结直肠息肉复发风险预测模型及效能分析

李斌(), 谭振刚(), 张华清   

  1. 中国人民解放军联勤保障部队第九六六医院胃肠科(邮编118000)
  • 收稿日期:2025-06-05 修回日期:2025-10-09 出版日期:2026-03-15 发布日期:2026-03-17
  • 通讯作者: E-mail:18840550966@163.com
  • 作者简介:李斌(1983),男,副主任医师,主要从事胃肠疾病及消化内镜治疗方面研究。E-mail:bilyli519486@163.com
  • 基金资助:
    辽宁省卫生健康委员会科研基金资助项目(2022LK031513)

Development of a predictive model for colorectal polyp recurrence risk based on polyp characteristics and serum indicators

LI Bin(), TAN Zhengang(), ZHANG Huaqing   

  1. Department of Gastroenterology, 966 Hospital of Joint Service Support Force of the Chinese People's Liberation Army, Dandong 118000, China
  • Received:2025-06-05 Revised:2025-10-09 Published:2026-03-15 Online:2026-03-17
  • Contact: E-mail:18840550966@163.com

摘要:

目的 构建基于结直肠息肉特征和血清学指标的复发风险预测模型,以提高结直肠息肉复发的早期识别和个性化治疗的能力。方法 回顾性纳入235例接受结直肠息肉切除患者并根据息肉是否复发分为复发组(56例)和无复发组(179例)。收集全部患者的一般资料、息肉特征和相关血清学指标,利用Logistic回归分析筛选出与结直肠息肉复发相关的独立预测因子,构建结直肠息肉复发风险的列线图模型,并通过受试者工作特征(ROC)曲线、校准曲线对模型的预测效能进行评估。结果 与无复发组相比,复发组年龄、男性比例、体质量指数(BMI)、家族史比例、息肉大小、腺瘤性息肉的比例、高级别病变比例、血清癌胚抗原(CEA)和C-反应蛋白(CRP)水平升高(P<0.05)。Logistic回归分析显示,年龄(OR=1.032,95%CI:1.003~1.062)、BMI(OR=1.064,95%CI:1.000~1.131)、家族史(OR=2.340,95%CI:1.039~5.267)、病理学分级(OR=6.323,95%CI:2.184~18.430)、息肉大小(OR=3.161,95%CI:1.434~6.965)、CEA(OR=1.486,95%CI:1.160~1.904)和CRP(OR=1.132,95%CI:1.022~1.254)为结直肠息肉复发的重要预测因子。基于上述因素构建结直肠息肉复发的列线图,模型的曲线下面积为0.893(95%CI:0.836~0.950),敏感度为81.6%,特异度为87.5%,具有较好的区分度和校准度(Hosmer-Lemeshow χ2=2.396,P=0.743)。结论 结直肠息肉复发风险预测模型具有较高的预测效能,可以有效评估结直肠息肉复发风险。

关键词: 肠息肉病, 复发, 列线图, 息肉特征, 血清学指标

Abstract:

Objective To develop a recurrence risk prediction model for colorectal polyps based on polyp characteristics and serological indicators, aiming to improve early recognition of colorectal polyp recurrence and personalized treatment. Methods A retrospective analysis was conducted on clinical data from 235 patients who underwent colorectal polyp resection. Based on whether recurrence occurred, the patients were divided into the recurrence group (56 cases) and the non-recurrence group (179 cases). Clinical data, polyp characteristics and related serological indicators were collected. Logistic regression analysis was used to identify independent predictors for colorectal polyp recurrence. Based on the results of Logistic regression, a nomogram model for predicting colorectal polyp recurrence risk was constructed, and the model’s predictive performance was evaluated using ROC and calibration curves. Results Compared with the non-recurrence group, the recurrence group had higher age, male proportion, body mass index (BMI), family history proportion, polyp size, proportion of adenomatous polyps, proportion of high-grade lesions, and elevated serum levels of carcinoembryonic antigen (CEA) and C-reactive protein (CRP) (P < 0.05). Logistic regression analysis showed that age (OR = 1.032, 95%CI: 1.003-1.062), BMI (OR = 1.064, 95%CI: 1.000-1.131), family history (OR = 2.34, 95%CI: 1.039-5.267), high-grade pathological lesions (OR = 6.323, 95%CI: 2.184-18.430), polyp size (OR = 3.161, 95%CI: 1.434-6.965), CEA (OR = 1.486, 95%CI: 1.160-1.904), and CRP (OR = 1.132, 95%CI: 1.022-1.254) were important predictors for colorectal polyp recurrence. Based on these factors, a nomogram model for predicting colorectal polyp recurrence risk was constructed. Model evaluation showed that the nomogram had an area under the curve (AUC) of 0.893 (95%CI: 0.836-0.950), sensitivity of 81.6%, specificity of 87.5%, and good discrimination and calibration (Hosmer-Lemeshow χ2 = 2.396, P = 0.743). Conclusion The recurrence risk prediction model based on polyp characteristics and serological indicators developed in this study has high predictive efficacy and can effectively assess the risk of colorectal polyp recurrence.

Key words: intestinal polyposis, recurrence, nomograms, polyp characteristics, serological indicators

中图分类号: