Biomarkers to guide antibiotic timing and administration in infected patients presenting to the emergency department

Antibiotics are often prescribed in the emergency department (ED) to patients presenting with a suspected infection before any definitive diagnosis can be made [1]. However, increasing antibiotic resistance and detrimental effects on the microbiota require their use to be limited to those with a high likelihood of bacterial infection or the potential for further clinical deterioration. Conversely, withheld or delayed treatment in higher severity patients may lead to increased morbidity and mortality rates [2]. Thus, an accurate assessment of antibiotic requirement and speed of administration is crucial.

Current tools to aid clinical decision-making include the use of procalcitonin (PCT) and C-reactive protein (CRP). However, recent interventional evidence in the ED has shown few differences between conventional biomarker-guided therapy and standard practice [1, 3], despite protocol compliance, patient selection and cut-off concerns. This post hoc analysis of a patient subset (Malmö, Sweden) from our previous investigation [4] compared the use of PCT, CRP and lactate to the novel biomarker mid-regional proadrenomedullin (MR-proADM) in guiding antibiotic administration during treatment within the ED.

Within this subset (N = 213), 26 (12.2%), patients were prescribed antibiotics < 48 h prior to presentation, whilst 187 (87.8%) were administered antibiotics during ED assessment. Of these patients, 164 (77.0%) were treated with intravenous (i.v.) and 23 (10.8%) with oral antibiotics. The median time to initial administration was 93 [28–160] min, with 71 (43.8%) patients receiving therapy within 60 min. Univariate and multivariate logistic regression found that MR-proADM had the strongest association with the requirement for antibiotic administration during ED treatment (Table 1). Interestingly, MR-proADM (Spearman ρ = − 0.31, p < 0.001) and lactate (Spearman ρ = − 0.25, p = 0.002) were the only parameters to be significantly negatively correlated with the time to antibiotic administration, with significant differences found at optimised MR-proADM cut-offs for antibiotic administration (1.27 nmol/L: 139 [76–211] vs 43 [26–135] min; p < 0.001) or pre-established [4] cut-offs for mortality prediction (1.54 nmol/L: 124 [33–199] vs 42 [26–122] min; p = 0.002). Similar results were also found for MR-proADM within previously established PCT concentration ranges [5] (Table 2), with an absence of ICU admission or 28-day mortality in patients with low MR-proADM concentrations, despite lower antibiotic administration rates and a significantly longer time to administration.

Results suggest that delayed antibiotic administration in patients with low MR-proADM concentrations may result in few adverse effects, potentially allowing for a more detailed clinical assessment prior to any subsequent initiation. Further studies in larger patient populations are required to confirm these initial findings.