Accuracy of a Rapid Messenger Ribonucleic Acid Host Response Test: Results from the SEPSIS-SHIELD Registrational Trial
Recommended Citation
Michelson EA, Liesenfeld O, Arora S, Aufderheide TP, Clements CM, DeVos EL, Fischer M, Giamarellos-Bourboulis EJ, House SL, Humphries R, Gill J, Liu E, Mace SE, May LS, Osborn T, Panacek EA, Rothman RE, Self WH, Smithline H, Steingrub J, Van Heukelom P, Weissman AJ, Wolk D, Wright DW, Shapiro NI. Accuracy of a Rapid Messenger Ribonucleic Acid Host Response Test: Results from the SEPSIS-SHIELD Registrational Trial. Acad Emerg Med 2025; 32(S1):18-19.
Document Type
Conference Proceeding
Publication Date
5-13-2025
Publication Title
Acad Emerg Med
Keywords
biological marker, C reactive protein, messenger RNA, procalcitonin, adult, antimicrobial stewardship, bacterial infection, cohort analysis, conference abstract, controlled study, diagnosis, diagnostic test accuracy study, disease severity, disease severity assessment, emergency ward, human, immune response, intensive care unit, intervention study, leukocyte count, major clinical study, male, middle aged, nasopharyngeal swab, prognosis, rapid test, respiratory tract infection, sepsis, severity of illness index, treatment outcome
Abstract
Background and Objectives: Rapid tests are needed to distinguish between bacterial, viral and non-infectious causes of illness along with predicting illness severity in «gray zone» patients in emergency departments (ED). We investigated the accuracy of a new 29-mrna host response (29HR); a novel ~30-min test using host mRNAs, for diagnosing acute infection type and predicting illness severity. Methods: We enrolled 1,222 patients with suspected acute infection and/or sepsis from 22 EDs (SEPSIS-SHIELD, NCT04094818). Whole blood was collected for 29HR testing and other biomarkers; nasopharyngeal swabs were collected in those with suspected respiratory tract infection. Patients were managed as per local standard of care and followed for 28 days. 29HR produces 3 scores for the likelihood of: (1) a bacterial infection, (2) a viral infection, and (3) the need for «ICU-level» interventions within 7 days. Each score falls into one of five interpretation bands ranging from Very low to Very high. We validated bacterial and viral results against blinded 3-person expert clinical adjudication, and illness severity results against the need for ICU-level interventions within 7 days. Results: The 29HR Bacterial Score had AUROC of 0.83 (80% CI 0.81-0.85) for the diagnosis of bacterial infection, significantly (p < 0.0001) higher than those for C-reactive protein (0.72, 80% CI 0.68-0.77), procalcitonin (0.69, 80% CI 0.65-0.72) and white blood cell count (0.75, 80% CI 0.72-0.79). The 29 HR Viral Score had AUROC of 0.91 (80% CI 0.90-0.93) for the diagnosis of viral infection. The 29HR Illness Severity Score alone had AUROC of 0.77 (80% CI 0.77-0.81) for predicting illness severity. However, combined with qSOFA scores it improved re-classification for the risk for «ICU-level care». For the Bacterial, Viral, and Severity Scores, the 'Very Low' band sensitivity was 97%, 95%, and 92%, respectively, while the 'Very High' band specificity was 95%, 98%, and 99%, respectively. Conclusion: 29HR provided rapid, accurate and potentially actionable results for the diagnosis, treatment decisions and prognosis of patients with suspected acute infection and/or sepsis. Future interventional studies and real-world experiences will determine whether addition of the 29HR test improves individualized management of patients with suspected acute infections and/or sepsis for enhanced antibiotic stewardship and better patient outcomes.
Volume
32
Issue
S1
First Page
18
Last Page
19
