高级检索

维持性血液透析患者发生营养不良的风险预测模型构建

Construction of a risk prediction model for malnutrition in maintenance hemodialysis patients

    摘要
  • 摘要:
    目的 探讨维持性血液透析患者发生营养不良的影响因素，并建立列线图预测模型。
    方法 回顾性分析行维持性血液透析255例患者的资料，按照6∶4比例将患者分为建模组153例及验证组102例。对建模组患者资料行单因素及多因素Logistic回归分析; 根据维持性血液透析患者发生营养不良的影响因素建立列线图预测模型。绘制校准曲线及受试者工作特征(ROC)曲线评估模型的有效性及区分度。
    结果 工作状况、有无合并糖尿病、营养知识知晓度、有无蛋白质能量摄入不足、透析频率、总尿素清除指数(Kt/v)、透析龄、年龄、体质量指数(BMI)、上臂中部周径(MAC)、超敏C反应蛋白、血肌酐(Scr)、降钙素原(PCT)、碳酸氢根(HCO3-)、甲状旁腺激素(PTH)为维持性血液透析患者发生营养不良的因素(P < 0.05)。多因素Logistic回归分析显示，蛋白质能量摄入不足、Kt/v、年龄、透析龄、超敏C反应蛋白为维持性血液透析患者发生营养不良的影响因素(P < 0.05)。校准曲线评估模型有效性结果显示，建模组的χ2=7.425, P=0.492; 验证组的χ2=6.398, P=0.603。ROC曲线评估模型区分度结果显示，建模组曲线下面积为0.877(95%CI: 0.818~0.936), 敏感度为84.34%, 特异度为80.00%; 验证组曲线下面积为0.825(95%CI: 0.747~0.903), 敏感度为89.09%, 特异度为61.70%。
    结论 蛋白质能量摄入不足、Kt/v、年龄、透析龄、超敏C反应蛋白为维持性血液透析患者发生营养不良的影响因素。建立的列线图预测模型区分度及有效性较好，可作为维持性血液透析患者早期筛选营养不良发生的工具。

     

    Abstract:
    Objective To explore the influencing factors of malnutrition in maintenance hemodialysis patients, and establish the prediction model of the nomogram.
    Methods Data of 255 patients undergoing maintenance hemodialysis were retrospectively analyzed, and the patients were divided into modeling group (153 cases) and verification group (102 cases) according to a vatio of 6∶4. Univariate and multivariate Logistic regression analysis was performed on data of patients in the modeling group; according to the influencing factors of malnutrition in maintenance hemodialysis patients, the prediction model of the nomogram was established. Calibration curve and receiver operating characteristic (ROC) curve were drawn to evaluate the validity and differentiation of the model.
    Results Working status, whether there was diabetes mellitus or not, nutritional knowledge awareness, whether there was insufficient protein energy intake or not, dialysis frequency, urea clearance index (Kt/v), dialysis age, age, body mass index (BMI), mid-upper arm circumference (MAC), hypersensitive c-reactive protein, serum creatinine (Scr), procalcitonin (PCT), HCO3- and parathyroid hormone (PTH) were the factors of malnutrition in maintenance hemodialysis patients (P < 0.05). Multivariate Logistic regression analysisshowed that insufficient protein energy intake, Kt/v, age, dialysis age and hypersensitive c-reactive protein were the influencing factors for malnutrition in maintenance hemodialysis patients (P < 0.05). Calibration curve evaluation results showed that χ2 of the modeling group was 7.425, and P value was 0.492; the χ2 of the verification group was 6.398, and P value was 0.603. The ROC curve evaluation model differentiation results showed that the area under the curve of the modeling group was 0.877 (95%CI, 0.818 to 0.936), the sensitivity was 84.34%, and the specificity was 80.00%; the area under the curve of the verification group was 0.825 (95%CI, 0.747 to 0.903), the sensitivity was 89.09%, and the specificity was 61.70%.
    Conclusion Insufficient protein energy intake, Kt/v, age, dialysis age and hypersensitive C-reactive protein are the influencing factors of malnutrition in maintenance hemodialysis patients. The established nomograph prediction model has good differentiation and effectiveness, and can be used as a tool for early screening of malnutrition in maintenance hemodialysis patients.

     

    • HTML全文
    • 参考文献(18)
    • 相关文章
    • 施引文献
  • 资源附件(0)

/

返回文章
返回