The Center for the Evaluation of Value and Risk in Health (CEVR) at Tufts Medical Center is pleased to announce the winners of our 8th annual Paper of the Year Award.
The best cost-effectiveness analysis published in 2025 was:
Rao IJ and ML Brandeau. Modeling health and economic outcomes of providing stable housing to homeless adults with opioid use disorder. JAMA Network Open, 2025, 8(6), e2517103. doi:10.1001/jamanetworkopen.2025.17103
This study assessed the cost-effectiveness of providing stable housing to people experiencing homelessness who have opioid use disorder (OUD), compared to the status quo of no housing in the United States. Using a dynamic model, the researchers found that providing housing, “even when no requirement to enter OUD treatment”, was cost-effective, resulted in “fewer deaths, and improved health outcomes”.
Rao and Brandeau’s paper was chosen from a competitive pool of nominations for its novelty and exemplary incorporation of an important social determinant of health into economic evaluation. Each paper underwent a blinded review by CEVR directors and faculty members, who evaluated the submissions for methodological quality, timeliness, and potential influence on policy or clinical decision-making. The top four papers with the highest faculty ratings were then reviewed by CEVR’s CEA Registry Scientific Advisory Board using the same criteria.
The runner-up for this category was:
• Wei X, Mansour L, Oxley S, Fierheller CT, Kalra A, Sia J, Ganesan S, Sideris M, Sun L, Brentnall A, Duffy S, Evans DG, Yang L, Legood R, Manchanda R. Defining Lifetime Risk Thresholds for Breast Cancer Surgical Prevention. JAMA Oncol. 2025;11(9):1072–1082. doi:10.1001/jamaoncol.2025.2203
CEVR is also thrilled to announce the winner of the Best CEA Registry Application category for our Paper of the Year awards.
In recognition of its outstanding importance, quality, and creativity, we selected the paper “Use of Large Language Models to Extract Cost-Effectiveness Analysis Data: A Case Study” as the winner.
Written by Xujun Gu et al. and published in Value in Health, this article evaluates the data extraction performance of GPT-4o against the CEA Registry. The authors found that GPT had success assisting with “simple variable extraction, filling missing data, and acting as a reference,” but it struggled with complex variables, such as utility and ratio-related variables. This exploration offers a valuable perspective on incorporating AI into health economics work.
