Optimizing Anticoagulation Efficiency in Pediatric Heart Disease

Following a surgical intervention, pediatric patients with heart disease have increased risk of forming damaging and potentially fatal blot clots. Anticoagulation therapy is a common strategy used to prevent potential organ dysfunction caused by these clots. Warfarin, the current principal medication used for oral anticoagulation, has a narrow therapeutic window and long half-life, making it a difficult and time-intensive drug to manage. Patient responses to the drug vary significantly and are particularly understudied in pediatrics. As a result, there is no standardized approach to warfarin dosing, which necessitates prescription tweaks and extended hospital stays. The goal of our project is to leverage data analysis to provide initial warfarin dosing recommendations based on given patient parameters.

In approaching this project, we conducted an extensive literature review and consulted pediatric cardiologists, advanced practice providers, and pharmacists to better understand the factors that may impact a patient’s metabolism of warfarin. Through these interviews, we found that children with heart disease often undergo intensive procedures and require several different medications throughout their hospital stay. As a result, we focused on examining patient sensitivity to warfarin with varying levels of hepatic impairment and the use of concomitant drugs. We compiled and cleaned many years of inpatient warfarin administration data and established binaries to easily identify patients that were on prescriptions that enhanced anticoagulation or had lab tests that indicated poor liver function. Through our analysis, we found that patients that were simultaneously taking medications with that enhanced anticoagulation were up to 5 times more sensitive to warfarin (delta INR/mg warfarin) than those that were not. We also found that patients with hepatic impairment were up to 6 times more sensitive to warfarin than patients with normal liver function.

Using these sensitivities, we hope to be able to construct a mixed-effects modeling program to provide initial warfarin dosing recommendations based on patient health parameters. In providing a more optimal starting dose for each patient, we aim to reduce the time it takes for patients to reach their therapeutic anticoagulation level. In doing so, we would also improve patient turnover and efficiency within Lucile Packard Children’s Hospital.

Abby Fleischli

M.S. Management Science & Engineering

B.S. Biology


Annie Vesey

M.S. Management Science & Engineering

B.S. Mechanical Engineering


David Scheinker

Founder & Director, SURF Stanford Medicine