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.
End-stage kidney disease (ESKD) patients treated with hemodialysis (HD) have a high disease burden and high mortality rate. ESKD is a remarkable risk factor across different health domains in itself. Combined with this, most patients have multiple comorbidities and an array of clinical abnormalities that act as risk factors for hospitalization and death events. The extent to which various types of demographic, clinical, and other risk factors may cluster and have additive prognostic value is unknown. Unsupervised machine learning (ML) methods can reveal hidden structures in complex clinical data. We applied this method across multiple laboratory and treatment domains in dialysis patients to identify risk factor profiles and evaluated whether these profiles were associated with patient outcomes.
Data were extracted from a globally distributed anonymized dialysis database, ApolloDialDb™ (Wolf M, 2025, KI Reports). We included incident patients from countries worldwide who started dialysis treatment in 2018. K-means clustering was performed using the FASTCLUS procedure in SAS (SAS Institute Inc., USA). A 5-cluster solution was chosen as it allowed for the most distinct and clinically meaningful characterization of patient profiles. The first three months of dialysis treatment were defined as baseline. We restricted the analysis to patients with complete data on all selected relevant laboratory and treatment variables. Patients were followed up for one year after baseline.
A total of 31,523 adult patients were included who started dialysis treatment in 2018. Cluster analysis based on five clusters (Table 1) yielded one small (Cluster 5: n=21) and four large clusters (Cluster 1: n=1,045; Cluster 2: n=18,050; Cluster 3: n=11,103; Cluster 4: n=1,304). Cluster 5 showed the highest mortality rate and was characterized by anemia and an increase in inflammatory parameters such as white blood cell count, followed by cluster 4 which was characterized by more pronounced differences in nutritional parameters, as reflected by lower serum creatinine and BUN levels. Cluster 1 comprised patients with severe overweight, hyperphosphatemia, higher blood pressure and high ultrafiltration (UF) rates suggesting fluid overload; Cluster 2 included patients without specific abnormalities, whereas cluster 3 consisted of patients characterized by parameters reflecting fluid overload and hyperphosphatemia, with a comparable mortality rate as cluster 2 despite being significantly younger.
Distinct risk patterns based on meaningful differences in laboratory and treatment parameters could be identified using cluster analysis. Recognizing these risk patterns may support more personalized treatment strategies in dialysis patients. Development and testing of prognostic cluster scores appear warranted and should consider both electronic integration as well as pragmatic approaches for calculation in real-world settings.
