Data-driven solutions for heart allocation policy challenges: the lvad dilemma and the status exception crisis

The National Heart, Lung, and Blood Institute awarded The Leland Stanford Junior University $105,476 for a Project Grant (CFDA 93.837, Cardiovascular Diseases Research) on January 25, 2027.

The award funds research addressing two critical shortcomings in the current donor heart allocation system. The project develops data-driven policy solutions to improve the prioritization of candidates with durable left ventricular assist devices (LVADs), which extend time on mechanical support but are associated with complications that reduce post-transplant survival, and to address the overuse of status exceptions in heart transplant listings, which now account for nearly 40 percent of all heart transplant allocations. The research applies rigorous methodological techniques—including mixed-effects Cox proportional hazards modeling—to novel data sources such as free-text clinical narratives to yield practical, evidence-based policy recommendations for the heart transplant community.

The performance period runs from January 25, 2027 through January 24, 2028, with work performed in California.

Abstract

Heart transplants are a potentially life-saving treatment for hundreds of thousands of patients living with advanced heart failure, but there are only enough donor hearts to perform a few thousand transplants annually. Under federal law, donor hearts must be allocated in order of medical urgency based on objective medical criteria. The current donor heart allocation system follows this mandate by ranking candidates based on six statuses, reflecting the acuity of therapies they are receiving. However, there are two critical shortcomings that threaten the fairness of donor heart allocation. First, durable left ventricular assist devices (lvads) are a treatment that allows select candidates to delay heart transplant until they stabilize clinically. In this setting, candidates with durable lvads are assigned low waitlist priority because they have low medical urgency; however, extended duration of lvad support is associated with complications that significantly reduce post-transplant survival. Solutions to appropriately prioritize these candidates before developing critical complications are severely lacking. Second, status exceptions, which are special requests to upgrade candidates' waitlist priority, now account for nearly 40% of all heart transplant listings. I have previously shown that exceptions are granted to candidates with low medical urgency and that the approval rate is near 100%. This proposal aims to address the critical and urgent need for solutions to these significant problems highlighted by the heart transplant community. The central hypothesis of this proposal is that applying methodologically rigorous techniques to novel sources of data, such as free-text narratives written by transplant clinicians, will yield practical policy-based solutions. In aim 1, i will construct a mixed-effects COX proportional hazards model incorporating restricted cubic splines to estimate the survival benefit of heart transplantation as a function of durable lvad support duration, identify the point at which the survival benefit is maximal, and determine which candidates should receive upgraded priority. In aim 2, i will employ a hybrid manual and large language model-based thematic analysis of more than 32,000 clinical narratives submitted with exception applications to identify distinct clinical phenotypes that correlate with medical urgency, providing standardized criteria for review of exception requests. This work will be completed at stanford university under the mentorship of DR. Kiran khush (primary sponsor), DR. William parker (CO-sponsor), and DR. Arden morris (CO-sponsor), with collaboration from DR. Suzanne tamang. The training plan emphasizes three learning aims: qualitative data analysis and large language model utilization, clinical medical ethics, and research dissemination to organ allocation stakeholders. Upon completion, this research will provide methodologically rigorous, evidence-based solutions to two maximum-priority policy challenges, directly informing national heart allocation policy to improve fairness and efficiency in donor organ distribution.