The efficiency of artificial intelligence for management and clinical decision-making in the identification of patients with undiagnosed HCV infection (Intelligen-C strategy).

Abstract

Introduction: Artificial intelligence (AI) allows the optimization of diagnostic processes for hepatitis C virus (HCV) patients. Our objective was to evaluate the clinical, economic, and management benefits of an AI-based clinical decision support system (Intelligen-C strategy).

Methods: The Intelligen-C strategy consisted of (1) a retrospective phase (Dec 2013-Sep 2021), in which medical records were reviewed to search for anti-HCV-positive and/or HCV-RNA-positive patients lost in the system, and (2) a prospective phase (Feb 2022-Jan 2023), in which automated screening (40-70 years) and routine testing for risk factors were performed in patients who were admitted to the emergency department or were hospitalized. With the use of automated screening, the system identified patients without an HCV diagnosis among those requiring blood tests and requested HCV serology; if the results were positive, reflex testing for HCV RNA was performed. If a patient was HCV-RNA positive, an alert was generated and sent to the hepatology department. In addition, the prospective phase was compared with the previous period to evaluate its effectiveness and efficiency.

Results: In the retrospective phase, the Intelligen-C strategy allowed the identification of 272 anti-HCV- or HCV-RNA-positive patients who were lost to follow-up, of whom 11 were treated; in the prospective phase, after 7,312 serologies were performed, 28 HCV-RNA-positive patients were identified, 14 attended the appointment, and 9 were treated. In the prospective phase vs. the previous period, increased serology (7,312 vs. 909), HCV RNA-positive detection (28 vs. 3), and treated patients (9 vs. 1) generated savings to the health system related to medical visits. In addition, Intelligen-C was cost-effective.

Conclusions: The implementation of the Intelligen-C strategy allowed the identification of patients with undiagnosed infection, facilitated their diagnosis, reduced healthcare processes and associated hospital costs, and proved to be efficient.