Improving hard-to-place kidney allocation: A machine learning approach to center ranking
Sean Berry1, Berk Görgülü2, Sait Tunç3
1Department of Mechanical, Industrial and Mechatronics Engineering, Toronto Metropolitan University, 350 Victoria Street, Toronto, ON, M5B 2K3, Canada.
Abstract:
Kidney transplantation is the preferred treatment for end-stage renal disease, yet donor scarcity and inefficiencies in allocation systems create major bottlenecks, resulting in prolonged wait times and alarming mortality rates. Despite the severe shortage of donor kidneys, timely and effective interventions to prevent non-utilization of life-saving organs remain limited. Expedited out-of-sequence placement of hard-to-place kidneys to centers with a high likelihood of acceptance has been recommended in the literature as a strategy to improve placement success. However, in practice, this process remains nonstandardized and relies heavily on the subjective judgment of decision-makers. We propose a data-driven, machine learning-based ranking policy for out-of-sequence allocation of hard-to-place kidneys that prioritizes transplant centers using predicted center-level acceptance probabilities. Using national deceased-donor and kidney-offer data, we construct a unique offer-level dataset with donor- and center-specific features. We also employ machine learning interpretability tools to provide insight into the factors influencing kidney allocation decisions. Our analysis demonstrates that the proposed policy can reduce the average number of centers considered before placement by fourfold for all kidneys and tenfold for the subset of hard-to-place kidneys. These results highlight the potential of the proposed framework to improve the efficiency of expedited placement and support more timely utilization of hard-to-place kidneys.
Insights
A new machine learning policy can improve kidney allocation by prioritizing hard-to-place kidneys to transplant centers most likely to accept them, reducing wait times and organ waste.
Area of Science:
- Nephrology
- Transplantation Science
- Artificial Intelligence in Healthcare
Background:
- Kidney transplantation is crucial for end-stage renal disease, but donor organ scarcity and allocation inefficiencies lead to long waitlists and high mortality.
- Current methods for expediting the placement of difficult-to-allocate kidneys are often subjective and lack standardization, contributing to organ non-utilization.
Purpose of the Study:
- To develop and evaluate a data-driven, machine learning-based policy for the out-of-sequence allocation of hard-to-place donor kidneys.
- To improve the efficiency of organ placement by prioritizing transplant centers with higher predicted acceptance probabilities.
Main Methods:
- Construction of a unique offer-level dataset using national deceased-donor and kidney-offer data, incorporating donor- and center-specific features.
- Development of a machine learning ranking policy to predict center-level acceptance probabilities for kidney offers.
- Utilization of machine learning interpretability tools to identify key factors influencing kidney allocation decisions.
Main Results:
- The proposed machine learning policy significantly reduces the number of transplant centers considered before kidney placement.
- A fourfold reduction in the average number of centers evaluated was observed for all kidneys.
- A tenfold reduction was achieved for the subset of hard-to-place kidneys, indicating enhanced efficiency.
Conclusions:
- A data-driven, machine learning approach offers a standardized and effective strategy for the expedited out-of-sequence allocation of donor kidneys.
- This framework has the potential to substantially improve the utilization of hard-to-place kidneys, thereby increasing transplantation rates and reducing patient mortality.
Related Concept Videos
01:28Kidney Transplant I: Introduction
01:26Kidney Transplant II: Surgical Procedure