Economic evaluations of healthcare technologies that incorporate artificial intelligence in radiology: a systematic review

Résumé fourni par la source

The integration of artificial intelligence (AI) in radiology has become increasingly relevant due to its potential to improve diagnostic accuracy and optimize healthcare workflows. This study reviews health economic evaluations of AI in radiology, following Cochrane international standards in the field. A systematic search was conducted in PubMed, Embase, Web of Science, and Google Scholar for studies evaluating AI-based health technologies in radiology from a health economics perspective. A qualitative synthesis of the data was performed, and each study was assessed for quality using the Consolidated Health Economic Evaluation Reporting Standards 2022. The PROSPERO registration number is CRD42024616818. A total of 938 records were retrieved, with 14 included in the review. Most studies were published between 2021 and 2024. The United States and the United Kingdom had the highest number of studies, with most research focusing on high-income countries. Cost-effectiveness analysis was the most common type of economic evaluation used. X-ray imaging was the most frequently analyzed modality, particularly for cancer detection, with a focus on breast and lung cancer. The most commonly used software for modeling was TreeAge Pro, followed by R and Microsoft Excel. AI in radiology offers substantial clinical benefits, but comprehensive long-term economic evaluations are necessary to fully assess its value. High initial investments, infrastructure changes, and ongoing maintenance were identified as key obstacles, especially in low-resource settings. Future research should prioritize underrepresented regions and incorporate diverse patient demographics to enhance the generalizability of findings. Not applicable.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.