Adherence to SVS Abdominal Aortic Aneurysm Guidelines Among Patients Detected by AI-Based Algorithm

Ann Vasc Surg. 2025 May 9:120:108-114. doi: 10.1016/j.avsg.2025.05.007. Online ahead of print.

Abstract

Background: This study evaluates adherence to the latest Society for Vascular Surgery (SVS) guidelines on imaging surveillance, physician evaluation, and surgical intervention for abdominal aortic aneurysm (AAA).

Methods: Artificial intelligencebased (AI-based) natural language processing was applied to imaging scans from retrospectively identified AAA patients at a tertiary care center between January and March 2019 and 2021, excluding the pandemic period. Retrospective chart review assessed demographics, comorbidities, imaging, and follow-up adherence. Statistical significance was set at P < 0.05.

Results: Among 479 identified patients, 279 remained in the final cohort following exclusion of deceased patients. Imaging surveillance adherence was 67.7% (189/279), with males comprising 72.5% (137/189). The mean age for adherent patients was 73.9 (SD ± 9.5) vs. 75.2 (SD ± 10.8) for nonadherent patients. Adherent females were significantly younger than nonadherent females (76.7 vs. 81.1 years; P = 0.003) with no significant age difference in adherent males. Adherent patients were more likely to be evaluated by a vascular provider within 6 months (P < 0.001), but aneurysm size did not affect imaging adherence: 3.0-4.0 cm (P = 0.24), 4.0-5.0 cm (P = 0.88), > 5.0 cm (P = 0.29). Based on SVS surgical criteria, 18 males (AAA > 5.5 cm) and 17 females (AAA > 5.0 cm) qualified for intervention and repair rates increased in 2021. Thirty four males (20 in 2019 vs. 14 in 2021) and 7 females (2021 only) received surgical intervention below the threshold for repair.

Conclusion: Despite consistent SVS guidelines, adherence remains moderate. AI-based detection and follow-up algorithms may enhance adherence and long-term AAA patient outcomes, however further research is needed to assess the specific impacts of AI.