UNOS redesigned lung allocation with MIT's models — adopted nationwide on March 9, 2023

The United Network for Organ Sharing has operated the US Organ Procurement and Transplantation Network since 1984, which means its allocation formulas decide, candidate by candidate, who is offered a donor lung. In 2019 the organization set out to redesign the national lung allocation policy around a new point system, with an explicit goal: prioritize waiting-list candidates in a way that yields both more efficient and more equitable outcomes — more transplants, more fairly distributed.

A joint Massachusetts Institute of Technology and UNOS team joined the OPTN Lung Transplantation Committee's policy design effort. The MIT group brought an analytical framework, developed at MIT, that combines optimization, regression and simulation. Its role was not to hand the committee an answer but to illuminate the salient trade-offs: for any proposed set of point-attribute weights, the analysis showed how outcomes would shift across the waiting list, making the human cost of each weighting choice visible before adoption.

The committee selected allocation-formula weights that the team's analysis had highlighted, and on March 9, 2023 the proposal was implemented as the national lung allocation policy across the United States. The work was recognized in the INFORMS Journal on Applied Analytics' special issue for the 2022 Daniel H. Wagner Prize, the practice award for advanced analytics and operations research, where the MIT–UNOS collaboration appeared among the finalists.

Optimize the weights of the waiting-list formula itself

Every weighting in an allocation formula is a moral and medical judgment, so the honest way to decide is to simulate what each weighting does, not debate adjectives.

Allocation policies are revised rarely and nation-wide, so a decision tool that quantifies trade-offs before adoption is worth far more than post-hoc analysis of a bad rollout.

Embedding the analysts inside the policy committee, rather than reporting to it, meant the committee's questions became the model's next run — the analysis shaped the choice at the moment of choice.

The framework's three techniques divided the work: regression captured the clinical data, optimization explored the weight space, and simulation stress-tested proposed formulas on realistic arrivals.

When an allocation rule is the product, treat the rule itself as the design space: simulate its weights before adopting them, and keep the approvers inside the analysis.

The weighted point system the committee selected went live as the United States' national lung allocation policy on March 9, 2023, replacing the previous priority rules for every waiting-list candidate. The collaboration was documented in 'Applying Analytics to Design Lung Transplant Allocation Policy' in the INFORMS Journal on Applied Analytics special issue associated with the 2022 Wagner Prize, giving transplant policy a worked example of model-guided rule design that other organ systems and countries can copy.