Evaluation of a host-protein signature score for differentiating between bacterial and viral infections: real-life evidence from a German tertiary hospital

Le résumé fourni par la source

PURPOSE: A host-protein signature score, consisting of serum-concentrations of C-reactive protein, tumour necrosis factor-related apoptosis-inducing ligand, and interferon gamma-induced protein 10, was validated for distinguishing between bacterial and viral infections as an antimicrobial stewardship measure for routine clinical practice among adult patients in a German tertiary hospital. METHODS: This single-centre, explorative study prospectively assessed the host-protein signature score, comparing it with serum procalcitonin (PCT) in patients with blood stream infections (BSI) and evaluating its efficacy in patients with viral infections against the standard of care (SOC) to assess the need for antibiotics due to suspected bacterial super/coinfection. Manufacturer-specified threshold scores were used to differentiate viral (< 35) and bacterial (> 65) infections. RESULTS: Ninety-seven patients (BSI [n = 56]; viral infections [n = 41]) were included. The score (cut-off score > 65) tended to detect BSI with higher sensitivity than did PCT (cut-off > 0.5 ng/mL) (87.5% vs. 76.6%). Three patients (5.4%) with BSI had a score < 35. One patient with BSI did not receive antibiotic treatment following SOC prior to positive blood culture results. Among patients with viral infections, 29 (70.7%) had scores > 65, indicating bacterial superinfections. Additionally, 11 patients (26.8%) had scores < 35, indicating no bacterial superinfections. In total, the antibiotic treatment discrepancy in the viral group between the SOC and a host-protein signature score guided approach was 2/41 patients (4.9%). CONCLUSION: The score tended towards a higher sensitivity in detecting BSI than that with PCT. However, its impact on reducing antibiotic use in viral infections was minor compared with that of SOC.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.