Emergency Department Sepsis Alert Admission Decisions After Introduction of a Host-Protein Test for Differentiating Bacterial vs Viral Infection

Emergency Department sepsis alerts prompt frequent admission before infection etiology is known. An FDA-cleared host-response assay (MeMed BV) that differentiates bacterial from viral infection was added as a default order to the sepsis alert set in 2024. We evaluated whether result availability reduced admissions among patients with viral results.

We performed a single-center, pre/post cohort study, where MMBV was automatically included in the ED sepsis alert order set for adults. In the control period, clinicians were blinded to MMBV results. In the intervention period (encompassing two serial intervention phases that improved visibility of the results in the EMR), clinicians received the MMBV result (score 0-100) and simple interpretive guidance (low scores: viral/non-bacterial; high scores: bacterial/possible coinfection). The primary outcome was hospital admission. We performed a single-center, pre/post cohort study, where MMBV was automatically included in the ED sepsis alert order set for adults. In the control period, clinicians were blinded to MMBV results. In the intervention period (encompassing two serial intervention phases that improved visibility of the results in the EMR), clinicians received the MMBV result (score 0-100) and simple interpretive guidance (low scores: viral/non-bacterial; high scores: bacterial/possible coinfection). The primary outcome was hospital admission.

In this preliminary analysis of the first 499 eligible encounters (control n=164; intervention n=335), demographics and immune status were similar between periods; approximately 41% were immunocompromised. For viral results, admissions were 93.5% (29/31) vs 78.9% (71/90) in control vs intervention (p=0.064). For nonviral results, admissions were 91.7% (122/133) vs 94.7% (232/245) (p=0.260). Mortality did not differ (p=0.197). The viral vs nonviral admission difference widened by 17.6 percentage points with test availability (p=0.008).

Our preliminary analysis supports that MMBV introduction to the ED sepsis alert order set was associated with significantly fewer admissions for patients with MMBV viral results without an increase in mortality.