Optimizing the allocation of hearts from deceased donors - PROJECT ABSTRACT In 2018, heart allocation moved from using three statuses to six statuses, but the priority groups remain undifferentiated and large, only coarsely reflecting each candidate’s risk of death on the waiting list. The new system aimed to reduce exception requests and more accurately rank candidates; instead, exceptions, which might be subjectively or inappropriately assigned, have risen to comprise 30% of heart transplants. The system also incentivized dramatic changes in clinical practice: because intraortic balloon pumps yield higher priority for transplant, balloon pump usage increased three-fold even though balloon pumps are associated with increased risk of neurologic complications. Other disparities persist: sensitized candidates wait four times as long for transplants but get no increased priority for hearts, in contrast to increased kidney priority for sensitized candidates. Our study will first describe the landscape of disparities in heart transplant, diagnosing whether transplants are equally available to candidates who are sensitized, of smaller or larger size, and of different races and ethnicities. We will also determine whether exception statuses are justified by candidates’ individual risks of waitlist death, and whether exceptions are being requested and granted in an equitable fashion. We will use machine learning to build a MESH (Model for End State Heart disease) score that predicts death on the waitlist for heart transplant candidates from national data. An individualized MESH score based on hemodynamic criteria in the context of cardiac pathology, end organ function, cause of heart failure, and eligibility for mechanical circulatory support therapies would better prioritize heart transplant candidates to reduce waitlist deaths without distorting clinical practice. Implementing an analogous lung allocation score in 2005 reduced waitlist deaths from 500 to 300 per year, so this change is overdue in heart allocation. Finally, we will design a composite allocation score for hearts. The Organ Procurement and Transplantation Network has resolved to replace the 250 and 500 mile circles in current policy with continuous distribution by implementing a composite allocation score. A composite allocation score is a weighted combination of medical urgency (MESH), with distance between donor and candidate, and other priority considerations like blood type, candidate and donor size, sensitization, and priority for prior living donors. However, eliminating geographic boundaries in this way creates an enormous combinatorial design space with complex tradeoffs. We will use simulation optimization to explore many possible choices for the numerical weights using clinically detailed simulations, guided by a differential evolution algorithm. Our team is exceptionally well-suited to the task, with dedicated quantitative scientists who have years of collaborative experience advancing transplantation partnering with clinicians of a superior heart transplant program. The proposed research would correct an arbitrary and coarse prioritization scheme, by establishing a validated heart allocation scoring system that reduces waitlist deaths and increases equity in heart allocation.