Cost-Effectiveness of AI in Breast Cancer Screening and Lung Cancer Diagnostics: A Case Study From Costa Rica.
Authors
Affiliations (4)
Affiliations (4)
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada; Department of Informatics & Statistics, École Polytechnique de Lille, Lille, France.
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada; Institute of Medical Science, University of Toronto, Toronto, ON, Canada; School of Medicine, University of Costa Rica, San José, Costa Rica.
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada.
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada; Institute of Medical Science, University of Toronto, Toronto, ON, Canada; School of Medicine, University of Costa Rica, San José, Costa Rica; Department of Statistical Sciences, University of Toronto, Toronto, ON, Canada. Electronic address: [email protected].
Abstract
This study evaluated the cost-effectiveness of integrating artificial intelligence (AI) into breast and lung cancer diagnostic workflows in Costa Rica, addressing the scarcity of health economic evidence for AI adoption in low- and middle-income countries and focusing on context-specific systemic barriers. Two state-transition Markov models compared AI-assisted workflows (AI as a first reader) with conventional radiologist-only strategies over a 10-year horizon. AI-supported mammography screening was evaluated in asymptomatic women aged 50 to 70 years. For lung cancer, symptom-driven pathways were modeled using GLOBOCAN 2022 incidence data. Inputs included region-specific costs (USD), quality-adjusted life-years (QALYs), and progression probabilities, calibrated to Costa Rica's cancer registry. Methodological rigor adhered to the Consolidated Health Economic Evaluation Reporting Standards 2022 guidelines. AI in breast cancer screening demonstrated economic dominance: it improved QALYs by 49% (10.46 vs 7.04) and reduced costs by 49% (USD 93 102 vs USD 183 278), yielding a net monetary benefit of USD 212 119 at Costa Rica's willingness-to-pay threshold. For lung cancer, the AI strategy also proved dominant: it reduced total costs to USD 1 218 507 (vs USD 1 337 640) while increasing QALYs to 6.88 (vs 4.15), driven by earlier detection that averts expensive late-stage palliative care. AI significantly improved cost-effectiveness in both breast and lung cancer diagnostics, demonstrating economic dominance by mitigating diagnostic delays and reducing late-stage treatment burdens. These findings stress AI's context-dependent impact in low- and middle-income countries and the necessity of early detection investments and equity-focused implementation to align innovation with healthcare system realities.