The Organ Procurement and Transplantation Network (OPTN) is working to develop a more equitable system of allocating deceased donor organs, referred to as continuous distribution (CD). The design of a new CD allocation policy can be viewed as a multi-objective optimization problem, requiring a balance between multiple equity and utility objectives.

This Policy Design Tool leverages machine learning, simulation, and optimization to allow for dynamic exploration of transplant outcomes under different CD policies across a variety of different metrics. It was developed at the MIT Operations Research Center, in collaboration with the UNOS and the OPTN. The methodology was introduced in [1], while its application to the design of a new lung CD policy is described in [2], and to kidney and pancreas in [3].

Overview of Continuous Distribution

Under the continuous distribution framework, candidates for deceased donor transplant are awarded priority points based on a series of factors, referred to as attributes. Attributes are weighted and put into groups based on the goal of each attribute. Grouping the attributes into goals allows transparent comparisons of value-based decisions.

The proposed goals (numbered) and attributes (alphabetic) for the continuous distribution of kidney, pancreas and kidney-pancreas are outlined below:

Candidates are awarded a composite allocation score (CAS) that is the weighted sum of the attributes outlined above. Higher scores put patients closer to the top of the waiting list, making them more likely to receive an organ transplant. The Policy Design Tool allows users to explore the impact of different policy options by reconfiguring the relative weighting of attributes and compare resulting metrics.

Simulation Notes

Simulation results are generated in MIT's simulator, which uses acceptance and transplant outcome models from the most recently published versions of the SRTR's Simulated Allocation Models (SAMs). The MIT simulator mimics KPSAM's and LSAM's simulation methodology, and has been validated against SRTR's simulation results.

References

1. Papalexopoulos, T., Alcorn, J., Bertsimas, D., Goff, R., Stewart, D., & Trichakis, N. (2023). Reshaping national organ allocation policy. Operations Research.
2. Papalexopoulos, T., Alcorn, J., Bertsimas, D., Goff, R., Stewart, D., & Trichakis, N. (2023). Applying Analytics to Design Lung Transplant Allocation Policy. INFORMS Journal on Applied Analytics, 53(5), 350-358.