Back
For best output, select "Paper Size" as "A4" and "Margin" as "0" or "None".
To save or print to PDF, please select Print Destination > Save as PDF, enable Background Graphics under "More Settings", then click "Save".
During the congress, E-Posters will be accessible to all participants on the congress website 24/7, as well as in the E-poster stations in the congress center.
Preparing your E-Poster
Please review the E-Poster format requirements carefully when preparing your E-Poster. Should your E-Poster not meet the mentioned requirements, it may not be displayed as described above.
E-Poster Submission Deadline
Please prepare and upload your E-Poster no later than March 14, 2026 11.59PM CET. After this date, you will no longer be able to prepare and upload your E-poster and it will not be displayed and accessible on the congress website.
Please follow the instructions below to input your abstract title.
Abstract titles should be brief and reflect the content of the abstract.
To explore factors associated with chronic kidney disease-associated pruritus (CKD-aP) in patients with maintenance hemodialysis (MHD), and construct prediction models separately using logistic regression and machine learning methods for risk factors identification.
Data in this cross-sectional study were collected from 327 hemodialysis centers in Jiangsu Province, China. The participants were divided into training set and validation set with a 5:1 ratio. Multivariate logistic regression analysis was used to identify factors associated with CKD-aP and used for nomogram construction, evaluated by odds ratio (OR) and 95% confidence interval (CI). 3 Machine learning algorithms were used to construct CKD-aP prediction models with all factors, and the area under (AUC) the receiver operator characteristic curve (ROC) was calculated for performance evaluation. Shapley additive explanation (SHAP) values were used to represent impact of each feature in the best performed model.
Among 38,952 eligible participants, 3,379 had CKD-aP. 14 factors were identified to be significantly associated with CKD-aP, including age (OR=1.006; 95%CI: 1.003-1.009; P<0.001), TCa (OR=1.319; 95%CI: 1.103-1.577; P=0.003), Ca-P product (OR=1.005; 95%CI:1.000-1.010; P=0.041), HDF treatment times per week (OR=1.229; 95%CI:1.151-1.312; P<0.001), post-dialysis HR (OR=1.004; 95%CI: 1.001-1.007; P=0.004), Mg (OR=1.782; 95%CI: 1.106-2.870; P=0.025), P (OR=1.311; 95%CI: 1.137-1.512; P=0.002), male (OR=1.251; 95%CI: 1.074-1.456; P=0.006), spKt/V (OR=1.323; 95%CI: 1.002-1.746; P=0.049), TG (OR=1.079; 95%CI: 1.019-1.142; P=0.017), vascular access (OR=1.447; 95%CI: 1.180-1.775; P=0.003), HB (OR=0.996; 95%CI: 0.994-0.999; P=0.008), HP treatment duration per week (OR=0.831; 95%CI: 0.706-0.978; P=0.032) and UA (OR=0.999; 95%CI: 0.998-0.999; P<0.001). The logistic regression model showed a good performance on CKD-aP prediction, with AUCs of 0.640 and 0.636 in training and validation sets, respectively. The CatBoost categorical boosting (CatBoost) algorithm had the best prediction performance among the three machine learning methods (AUC for validation set was 0.804). In addition to the risk factors identified in the logistic regression, the SHAP values constructed from the CatBoost suggested that β2-MG is an important independent risk factor.
The study results provided information for risk prediction on CKD-aP in MHD patients. Compared to the logistic regression model, Machine learning methods can yield additional discoveries beyond traditional findings and provide additional insight for future research.
