Blood pressure variability and cardiovascular risk assessment using machine learning clustering after arteriovenous fistula creation in hemodialysis patients.

Introduction: Hypertension (HTN) is highly prevalent among patients with chronic kidney disease (CKD) and end-stage kidney disease (ESKD), uncontrolled HTN as well as high blood pressure variability (BPV) can significantly impact cardiovascular health. This study aims to understand the impact of BPV in patients with ESKD following arteriovenous fistula (AVF) creation for hemodialysis, to better understand its association with age, sex, BMI, and cardiovascular risk.

Methods: A retrospective analysis was conducted on 83 patients who underwent AVF creation between February 2019 and October 2020 at University of Miami Hospital. During 20 months, blood pressure measurements were collected from medical records at four different times during AVF access creation process, and BPV was quantified using machine learning clustering techniques.

Results: The study identified three distinct clusters among the 83 patients with ESKD who underwent AVF creation. Cluster 1 (n = 28) was characterized by high blood pressure variability, with a mean systolic blood pressure standard deviation of 16.2 mmHg. This cluster exhibited a significantly higher cardiovascular risk rate of 42.9% compared to Cluster 2 (n = 30) and Cluster 3 (n = 25), which had lower variability with mean systolic blood pressure standard deviations of 10.3 and 8.7 mmHg, respectively, and cardiovascular risk rates of 16.7% and 12.0%. Age, sex, and BMI did not differ significantly across the clusters.

Conclusion: This study underscores the association between blood pressure variability and the increased risk of cardiovascular events, including myocardial infraction, stroke or heart failure in patients with ESKD undergoing AVF creation. Our findings highlight the critical need for vigilant monitoring of blood pressure fluctuations in this population and demonstrate how these variations are influenced by individual demographic factors. Further research is needed to develop targeted interventions to mitigate this risk.