2224. Reducing Antibiotic Overuse in Adult Lower Respiratory Tract Infections Using Novel Host–Response-Based Diagnostics

Abstract Background Antibiotic overuse in LRTI is a major healthcare care problem, contributing to antimicrobial resistance. A novel assay that integrates blood levels of three immune-proteins TRAIL\IP-10\CRP was developed to assist in differentiating bacterial from viral disease. The assay exhibited high performance in blinded validation studies focusing on children. We performed a preliminary analysis of the ongoing OBSERVER study, evaluating the assay’s potential to reduce antibiotic misuse in adult patients presenting with suspicion of LRTI. Methods OBSERVER (NCT03011515) is an EU Horizon 2020 funded study (grant No. 684589), the first to validate the signature in adult LRTI patients. For every participant recruited at the emergency departments of three hospitals in Israel, we collected medical history, physical examination, routine lab, imaging, and respiratory multiplex PCR data. The assay outcomes are bacterial, viral or equivocal. Reference standard outcome of bacterial, viral, indeterminate or noninfectious, was assigned by expert panel majority adjudication. Indeterminates were excluded from the analysis. Results In this preliminary analysis, we included the first 218 patients with locked data (Figure 1). Age ranged from 18 to 96 years (mean 59.5). Clinical syndromes included: 21% pneumonia, 13% acute bronchitis, 6% COPD exacerbation, 32% upper respiratory tract infection and 16% unspecified LRTI or viral infections. The assay demonstrated high diagnostic performance for distinguishing bacterial from viral disease (Figure 2). Assay equivocal rate was 8%. In this cohort, antibiotics were prescribed to 41 of 105 patients with viral reference outcomes indicating an overuse rate of 39%, of these, 34 yielded viral index test outcomes, supporting the potential of the assay to reduce overuse by ~83%. Conclusion The TRAIL/IP-10/CRP assay demonstrated high diagnostic performance for differentiating between bacterial and viral disease. Medical literature shows that there is a big gap between guidelines antibiotic prescription recommendations and reported prescribing rates (~25% vs. 40%-50%) for suspected LRTI in adults. The use of this new assay, which has a specificity of 93% and NPV of 99%, can help to close the gap and improve adherence to the guidelines. Disclosures All authors: No reported disclosures.