Machine Learning Approaches to Analyzing Adverse Events Following Durable LVAD Implantation

Abstract Background This study employed machine learning (ML) approaches to analyze sequences of adverse events (AE) following left ventricular assist device (LVAD) implantation. Methods Data on patients implanted with the HeartWare HVAD durable LVAD were extracted from the ENDURANCE and ENDURANCE Supplemental clinical trials, with follow-up through 5 years. Major AEs included device malfunction, major bleeding, major infection, neurological dysfunction, renal dysfunction, respiratory dysfunction, and right heart failure (RHF). Time interval and transition probability analyses were performed. A Sankey diagram was created to visualize transitions between AEs. Hierarchical clustering was applied to dissimilarity matrices based on the longest common subsequence to identify clusters of patients with similar AE profiles. Results A total of 568 patients underwent HVAD implantation with 3,590 AEs. Bleeding and RHF comprised the highest proportion of early AEs postoperatively whereas infection and bleeding accounted for the majority of AEs occurring after 3 months. The highest transition probabilities were observed with infection to infection (0.34), bleeding to bleeding (0.31), right heart failure to bleeding (0.31), right heart failure to infection (0.28), and bleeding to infection (0.26). Five distinct clusters of patients were generated, each with different patterns of time intervals between AEs, transition rates between AEs, and clinical outcomes. Conclusions ML approaches allow for improved visualization and understanding of AE burden following LVAD implantation. Distinct patterns and relationships provide insights that may be important for quality improvement efforts.