A Study of the Effect of Hematocrit on Weaning in the 12 cc Penn State Pediatric Ventricular Assist Device

Gallagher, Maureen Brenna
Area of Honors:
Biomedical Engineering
Bachelor of Science
Document Type:
Thesis Supervisors:
  • Keefe Manning, Thesis Supervisor
  • Keefe Manning, Honors Advisor
  • William Hancock, Faculty Reader
  • William O Hancock, Faculty Reader
  • PVAD
  • artificial heart
  • hematocrit
  • weaning
  • fluid dynamics
Ventricular assist devices (VADs) are used to treat severe heart failure by aiding the heart mechanically to increase cardiac output. These devices are used to bridge a patient to transplant or even recovery, and as a method for stand-alone long-term survival. For cases in which long-term VAD use has allowed for the recovery of the patient’s heart, it is desirable to wean a patient from the VAD and allow the heart to strengthen and function on its own again. Stroke volume reduction is a promising way to reload the heart gradually and wean a patient from the device. Since a much smaller volume of research exists for pediatric VADs compared to the larger adult VADs, serious complications persist in their use. Past research has shown that certain flow conditions can promote or reduce occurrences of thrombosis. To yield a better understanding of the factors that cause complications with pediatric VADs, this study analyzed the fluid dynamics in the 12 cc Penn State pediatric VAD. Changes in both stroke volume and hematocrit have been shown to influence the fluid dynamics within the pediatric VAD. While hematocrit, or the ratio of packed red blood cells to whole blood, remains constant for adults at 40%, pediatric hematocrit ranges from 20%-60%. This study examined the effect of hematocrit on the fluid dynamics in the Penn State pediatric VAD during a stroke volume reduction weaning protocol. The experiments used a mock circulatory loop and blood analog fluid to mimic cardiovascular conditions and blood behavior, respectively. To cover the range of pediatric hematocrit, fluids mimicking blood at 20%, 40%, and 60% were used, with the 20% hematocrit fluid being the least viscoelastic fluid and the 60% hematocrit fluid being the most viscoelastic fluid. Stroke volume conditions of 100% (12 cc), 80% (9.6 cc), and 60% (7.2 cc) were created to simulate a stroke volume reduction weaning protocol. Particle image velocimetry (PIV) was used to image the flow fields within the VAD, with the results producing quantitative velocity data describing the flow conditions within the VAD for all three hematocrits at each step of the weaning procedure. The flow patterns for each set of conditions followed a similar sequence: inflow began at the start of diastole, inflow transitioned to a solid body rotation through diastole, and rotational flow dissipated as outflow began in systole. Varying the hematocrit and stroke volume, however, created specific dynamic differences in the flow characteristics. Increased viscoelasticity led to delayed formation of flow patterns, including the inlet jet and the rotational flow. A more viscoelastic fluid also created a stronger inflow, a stronger, late-diastolic rotational flow, and a stronger outflow. A reduction in stroke volume led to flow at lower speeds and more areas of stagnation throughout the cycle. Favorable flow patterns persisted with the use of 40% and 60% hematocrit fluids through all three steps of stroke volume reduction. The 60% hematocrit fluid created the strongest flow and the longest-lasting solid body rotation with the most potential for thrombus reduction. The 20% hematocrit fluid, however, created weaker rotational flow that lasted into systole, dissipated sooner and left more areas of low flow and stagnation. Care should be taken with patients with low hematocrit that adequate flow and wall washing is occurring throughout the cycle to prevent occurrences of thrombosis and patient complications.