DOACT.Vasc Study

Abstract

The project was conceived and developed based on an automated computational flowchart structured through a decision tree using the CART (Classification and Regression Tree) method, and it was approved by the Ethics Committee of Santa Casa de São Paulo. The resulting algorithm, named DOACT, was fully documented and made available in a private GitHub repository, ensuring scientific transparency and reproducibility. The code repository associated with this study can be accessed at: https://github.com/italoeugenioabreu/doact . The flowchart was structured according to the most recent recommendations from leading international clinical guidelines, including those from the American College of Chest Physicians (CHEST) and the European Society for Vascular Surgery (ESVS). Study Design This was an observational, comparative, and exploratory single-blind study conducted across three analytical groups: Vascular surgery specialists Non-vascular medical specialists Artificial intelligence (AI) models based on Large Language Models (LLMs) - (ChatGPT-4o, Gemini 2.5, Grok 3.0, and Claude Sonnet 4.0) The total sample consisted of 53 participants, and the main objective was to evaluate the correct responses and diagnostic performance of the DOACT tool across 15 clinical case vignettes involving superficial venous thrombosis, deep vein thrombosis, and pulmonary thromboembolism.Case