中国血液净化

中国血液净化 ›› 2025, Vol. 24 ›› Issue (11): 919-923,928.doi: 10.3969/j.issn.1671-4091.2025.11.009

• 临床研究 • 上一篇    下一篇

个体化预测维持性血液透析患者便秘患病风险的列线图预测模型构建

赵莹莹   仲丽丽    高雪萍    王 琪    杨 静   顾红叶   

  1. 223002 淮安，1淮安市第二人民医院血液净化中心
  • 收稿日期:2025-01-21 修回日期:2025-08-01 出版日期:2025-11-12 发布日期:2025-11-12
  • 通讯作者: 赵莹莹 E-mail:1002345169@qq.com

Construction of a column chart prediction model for individualized prediction of constipation risk in maintenance hemodialysis patients

ZHAO Ying-ying, ZHONG Li-li, GAO Xue-ping, WANG Qi, Yang Jing, GU Hong-ye   

  1. Blood Purification Center, Huai’an Second People's Hospital, Huai’an 223002, China
  • Received:2025-01-21 Revised:2025-08-01 Online:2025-11-12 Published:2025-11-12
  • Contact: 赵莹莹 223002 淮安，1淮安市第二人民医院血液净化中心 E-mail:1002345169@qq.com

摘要： 目的  探讨维持性血液透析(maintenance hemodialysis，MHD)患者便秘的影响因素，并构建列线图预测模型。 方法  选取2021年4月─2023年6月于淮安市第二人民医院行MHD治疗的135例患者为建模组，另选取2023年7月─2024年11月58例MHD患者作为验证组。根据MHD期间便秘发生情况将建模组分为便秘组(n=56)和未便秘组(n=79)。MHD患者便秘的影响因素采用多因素Logistic回归分析；MHD患者便秘风险的列线图预测模型在R软件中构建，模型预测效能通过受试者工作特征(receiver operating characteristic，ROC)曲线、Hosmer-Lemeshow检验和校准曲线评估。 结果  193例MHD患者中共78例发生便秘，总发生率为40.41%。便秘组患者糖尿病肾病、透析龄≥5年及日常以卧床休息为主的占比高于未便秘组(χ2=5.539、5.053、4.293，P=0.019、0.025、0.038)，Alb水平低于(t=4.161，P＜0.001)。Logistic多因素回归显示糖尿病肾病(OR=5.739，95%CI：2.089～15.767，P=0.001)、透析龄≥5年(OR=6.287，95%CI：2.246～17.598，P＜0.001)、日常以卧床休息为主(OR=8.396，95%CI：2.348～30.022，P=0.001)是MHD患者便秘的独立危险因素，Alb为MHD患者便秘的保护因素(OR=0.865，95%CI：0.788～0.950，P=0.002)。列线图显示MHD患者总得分越高，便秘风险越高。ROC曲线显示建模组、验证组的曲线下面积(area under the curve，AUC)分别为0.806(95%CI：0.733～0.879)、0.879(95%CI：0.796～0.963)，表明预测区分度较高。Hosmer-Lemeshow 检验和校准曲线表明建模组、验证组模型预测一致性较高(χ2=4.286、4.090，P=0.830、0.849)。 结论 MHD患者便秘风险与原发疾病、透析龄、日常生活情况和Alb水平有关，基于这4个因素建立的列线图模型预测MHD患者便秘风险的效能良好，有助于临床及时预测，为便秘的预防提供帮助。

关键词: 维持性血液透析, 便秘, 影响因素, 列线图

Abstract: Objective  To explore the influencing factors of constipation in maintenance hemodialysis (MHD) patients and to construct a column chart prediction model.  Methods  A total of 135 patients who underwent MHD in Huai’an Second People's Hospital from April 2021 to June 2023 were regarded as the modeling group. A total of 58 MHD patients from July 2023 to November 2024 were recruited as the validation group. According to the occurrence of constipation during MHD, the modeling group was assigned into constipation group (n=56) and non-constipation group (n=79). The influencing factors of constipation in the MHD patients were analyzed using multivariate logistic regression. A column chart prediction model for constipation risk in MHD patients was then constructed using the R software. Predictive performance of the model was evaluated through receiver operating characteristic (ROC) curve, Hosmer-Lemeshow test, and calibration curve.  Results Among the 193 MHD patients, a total of 78 cases experienced constipation, with an incidence rate of 40.41%. Compared with the non-constipation group, the constipation group had higher rates of diabetes nephropathy, dialysis age≥5 years, and lying in bed in most daytime (c2=5.539, 5.053 and 4.293; P=0.019, 0.025 and 0.038), while plasma albumin was lower (t=4.161, P＜0.001). Multivariate logistic regression found that diabetes nephropathy (OR=5.739, 95% CI: 2.089～15.767, P=0.001), dialysis age ≥5 years (OR=6.287, 95% CI: 2.246～17.598, P＜0.001), and lying in bed in most daytime (OR=8.396, 95% CI: 2.348～30.022, P=0.001) were the independent risk factors for constipation in MHD patients (P＜0.05), and plasma albumin was a protective factor for constipation in MHD patients (OR=0.865, 95% CI: 0.788～0.950, P=0.002). The column chart showed that the higher the total score of MHD patients, the higher the risk of constipation. ROC curve showed that the area under the curve (AUC) of the modeling group and the validation group were 0.806 (95% CI: 0.733～0.879) and 0.879 (95% CI: 0.796～0.963) respectively, indicating higher predictive discrimination of the column chart. Hosmer-Lemeshow test and calibration curve indicated that the model had higher prediction consistency.  Conclusion   The risk of constipation in MHD patients is related to primary disease, dialysis age, daily living conditions and plasma albumin level. The nomogram model established based on the four risk factors has good efficacy in predicting the risk of constipation in MHD patients and is helpful for clinical prediction and prevention of constipation.

Key words: Maintenance hemodialysis, Constipation, Influencing factor, Column chart

中图分类号: