Non-invasive Tools to Diagnose Invasive Aspergillosis Infections in ICU Patients With COVID-19 and Other Conditions.

Invasive aspergillosis (IA) are difficult to diagnose in the ICU population, as the patients often do not present the conventional risks factors of immunocompromised patients (EORTC/MSG criteria). In the ICU population, patients often present other risk factors, such as cirrhosis, COPD, influenza and currently SARS-Cov2. The clinicians are thus currently missing precise criteria to distinguish colonization from IA in these patients, while they need to decide if an antifungal treatment is necessary or not. A new algorithm, entitled BM ASP ICU, based on investigators field experience and the scientific literature, which takes into account both EORTC/MSG criteria and a combination of fungal biomarkers, was proposed recently by Haman et al, Annals Intensive Care, 2021. Additional serological assays (immunoprecipitation and ELISA) showed since their interest, especially concerning SARS-Cov2 patients, a new population at risk of IA in the ICU, which emerged in the past months. The present study aims at prospectively implementing the BM ASP ICU algorithm during two years in the routine practice of six ICU units distributed in general and teaching hospitals situated northeast of France. The BM ASP ICU algorithm would be completed by serological assays aiming at assessing a sensitization towards Aspergillus fumigatus. The investigators plan to include 400 ICU patients at risk of IA; SARS-Cov2 patients will be part of the cohort. A weekly screening including culture of respiratory samples, galactomannan antigen, fungal qPCRS (targeting A. fumigatus), and A. fumigatus serology will be applied for all included patients. The performance (sensitivity and specificity, likelihood ratios) of each fungal biomarkers, alone and in combination with others, will be assessed, for all patients, and also within subgroups of patients with specific risk factors (such as SARS-Cov2 for example). These results should lead to solid understanding of which combination of tests is optimal to di