PRM59 - DEVELOPING A MODELLING FRAMEWORK TO ASSESS THE COST-EFFECTIVENESS OF DIAGNOSTIC TESTS FOR DISTINGUISHING BETWEEN BACTERIAL AND VIRAL INFECTIONS

Abstract

Bacterial and viral infections are often clinically indistinguishable, particularly in patients with respiratory tract infections. It is important to set a framework for evaluating diagnostic tests aiming to differentiate between these infections, since misdiagnosis can lead to antibiotic misuse and complications. The objective is to demonstrate the development of a cost-effectiveness model framework that integrates diagnostic pathways with antibiotic treatment decisions and its consequences, and accounts for the cost of antimicrobial resistance (AMR). A health-economic model was developed, which compares a new diagnostic test to standard-of-care (SoC), by simulating patients’ diagnostic and treatment pathways, quality-adjusted life years and associated costs. The model consists of three parts: 1.) A decision tree describing patients’ diagnostic pathways upon arrival at the emergency department; 2.) A Markov model simulating the therapeutic decision and its consequences, such as antibiotic adverse events; and 3.) An optional functionality that accounts for the cost of AMR. In a hypothetical scenario in which bacteria was the causative agent in 15% of pediatric patients and the new diagnostic had 15% higher sensitivity and specificity compared to SoC, holding all other model inputs equal, the antibiotic prescription rate was reduced by more than 30%, and costs due to antibiotic treatment plus related side effects, not including AMR, were reduced by approximately 30%. Additional results will be presented. This modelling framework is useful for comparing diagnostic tests to SoC, taking diagnostic accuracy, treatment decisions and treatment-related consequences into account, and for assessing the impact on antibiotic consumption. Furthermore, this framework accounts for the cost of AMR. Since this cost is a negative externality resulting from the consumption of antibiotics, it is usually not included in health economic evaluations. In the future the model will be used to assess the cost-effectiveness, and antibiotic-related aspects, of a new diagnostic in comparison to SoC.