Antibiotic prescribing that aligns with a host-protein test at US acute care settings is associated with fewer subsequent hospitalisations.

backgroundDifferentiating bacterial from viral infections in acute care is challenging, often leading to inappropriate antibiotic use. MeMed BV (MMBV) is a host-protein test integrating TNF-related apoptosis-inducing ligand, induced protein-10 and C-reactive protein to distinguish infection aetiology. We evaluated whether alignment between MMBV results and antibiotic prescribing is associated with downstream clinical outcomes and costs.

methodsData were pooled from two prospective US studies (APOLLO and JUNO). Alignment was defined as: antibiotics prescribed for bacterial MMBV results or no antibiotics for viral results. The primary outcome was unplanned revisits and hospitalisations within 28 days. Exploratory analysis estimated healthcare costs using 2024 Healthcare Cost and Utilisation Project data.

resultsThe cohort included 297 adults (median age 36.4 years). 20.5% had bacterial, 68.0% viral and 11.4% equivocal MMBV results. In patients with bacterial MMBV results, antibiotic prescription (alignment) was significantly associated with fewer revisits (5.0% vs 33.3%; p=0.003) and hospitalisations (2.5% vs 19.0%; p=0.026). For patients with viral results, outcomes were comparable regardless of antibiotic use. In the exploratory analysis, alignment with MMBV was associated with lower estimated costs (£101 vs £1175/patient; difference £1074), driven mainly by fewer hospital days in the aligned group.

conclusionMisalignment between MMBV bacterial results and antibiotic prescribing in acute care settings was associated with higher revisit and hospitalisation rates. These findings identify a quality gap under diagnostic uncertainty and support a methodology which includes assessment of test versus clinical outcome alignment when evaluating new diagnostic tests.