Futile organ procurement remains a costly and persistent challenge in liver transplantation. In donation after circulatory death (DCD) cases, surgical teams often prepare for organ recovery only to abandon the procedure because the donor does not die within the required timeframe.
These failed attempts consume ICU resources, operating room time, and staff effort, driving up healthcare costs that ultimately impact insurers through higher claims and prolonged hospital stays.
A recent multicenter study published in The Lancet highlights a promising solution: a machine-learning model built on the light gradient boosting machine (LightGBM) framework. By analyzing complex physiological, cardiovascular, and neurological data, the model predicts, with remarkable accuracy, whether a donor will progress to death within the necessary window. Validated across six US transplant centers, the tool reduced futile procurements by up to 60%, outperforming traditional calculators and even surgeon judgment.
This breakthrough represents more than a clinical win. By minimizing wasted interventions, hospitals can redirect resources to cases with higher success potential, reducing unnecessary costs and improving overall system efficiency and performance. For health insurers, this directly translates into fewer claims tied to failed transplant attempts. Additionally, better donor management accelerates successful transplants, improving survival rates for patients on waiting lists, potentially reducing death benefit payouts. These factors may influence underwriting and pricing for critical illness, transplant-related coverage, and even life policies.
Although the LightGBM model is still evolving, AI-driven decision tools like this are poised to reshape healthcare economics. Insurers who adapt their risk models to reflect these efficiencies will be better positioned in a market increasingly defined by precision and resource optimization.
RGA experts are eager to engage with clients to better understand and tackle the industry’s most pressing challenges together. Contact us to learn more about RGA's capabilities, resources, and solutions.
Development and validation of a machine-learning model to reduce futile procurements in donations after circulatory death in liver transplantation in the USA: a multicentre study – The Lancet Digital Health. https://www.thelancet.com/journals/landig/article/PIIS2589-7500(25)00100-1/fulltext
