Left Ventricular Assist Devices (LVADs) are used to support blood flow in people with advanced heart failure. They improve quality of life and survival and are used in some cases as a bridge to cardiac transplantation. However, many LVAD recipients experience undesired adverse events, including acute right heart failure (RHF). To improve the understanding of underpinnings and optimal management of post-LVAD right heart failure.
Dr. Bahl and colleagues examined over 19,000 individuals who had LVAD placement between 2008 and 2017 from the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) database. Using an unbiased explainable machine learning approach, they examined relationships between 186 pre-LVAD variables and post-LVAD outcomes.
Severe RHF developed in 19% of study subjects. Dr. Bahl et al. found many factors associated with RHF development, such as MELD score, inotrope use, and hemoglobin level. He graphically demonstrated trends between levels of candidate risk factors and the strength of association as well as interactions between risk factors. This novel knowledge of risk factors and relationships between them is progress toward predicting which patients are most likely to develop post-LVAD RHF and implementing risk-reduction strategies for those at highest risk.
Dr. Bahl presented this research at the American Society for Artificial Intelligence Organs (ASAIO) annual conference in Chicago, where it won the Top Cardiac Abstract award and the Kolff Award. The abstract was published in ASAIO Journal [June 2022;68(Supp 2):2] and the manuscript is forthcoming in the same journal.