For over 50 years, continuous flow ventricular assist devices (VADs) have greatly improved the longevity and quality of life of end-stage heart failure patients. However, these individuals remain at high risk for thrombosis, stroke, and gastrointestinal bleeding. Herein, this report attempts to understand and explain why these adverse events persist in both children and adults receiving a VAD despite decades of R&D. One objective was to develop a novel method of characterizing pump thrombosis in an axial-flow VAD, which can be used to better calibrate computational fluid dynamics models in blood-wetted devices. These thrombi were classified in five regions of the pump by their composition, severity, frequency, and physical distribution. There was a statistically significant correlation between thrombosis deposition on the forbearing and subsequent downstream thrombogenesis in the outlet bearing/stator region, suggesting upstream thrombosis may propagate and lead to downstream deposition.A second aim was to propose an alternative explanation for elevated stroke rates in the absence of adherent pump thrombosis; mainly, the ingestion of thrombi from the ventricle or atrial appendage. High-speed videography was used to capture the degree of disruption, adherence, and disintegration of thrombi in transparent VAD models. Although a large proportion of thrombi were macerated into infinitesimal pieces, some thromboemboli ejected from the pumps were of sufficient size to occlude an intracranial artery. Other ingested thrombi were observed physically adhering to the inner flow path, which may serve as a nidus for further thrombogenesis. A third objective was to explain the differences in non-surgical bleeding rates between mechanical circulatory devices in the pediatric setting, which is commonly attributed to the mechano-enzymatic cleavage of von Willebrand factor (vWF). At a 1.5 LPM flow condition, the loss of high molecular weight vWF multimers was comparable (-7.5% and -7.9%) between the PediaFlow™, a magnetically levitated pediatric ventricular assist device, and a standard of care device (PediaMag™). The vWF: CB/Ag ratio decreased slightly in the PediaFlow™ compared to PediaMag™ (-8% vs +3%), but the results were not statistically significant. Collectively, these results support alternative hypotheses for persistent adverse events observed in both pediatric and adult VAD patients.