Objective: Several studies showed the substantial use of antibiotics and increased risk of antimicrobial resistant infections in patients with COVID-19. The impact of COVID-19-related treatments and antibiotics on gut dysbiosis has not been clarified.

Design: The prospective cohort study included hospitalized COVID-19 patients (April-December 2020). The gut microbiome composition was analysed by 16S sequencing. The gut diversity and changes in opportunistic bacteria (OBs) or symbionts were analysed according to clinical parameters, laboratory markers of disease progression, type of non-antibiotic COVID-19 treatments (NACT) and type, WHO AWaRe group, and duration of antibiotic therapy (AT).

Results: A total of 82 patients (mean age 66 ± 13 years, 70% males) were enrolled. The relative abundance of Enterococcus was significantly correlated with duration of hospitalization, intensive care unit stay, O2 needs, and D-dimer, ferritin, and IL-6 blood levels. The presence of Enterococcus showed the highest number of correlations with NACT, AT, and AT + NACT (e.g., hydroxychloroquine ± lopinavir/ritonavir) and increased relative abundance with AWaRe Watch/Reserve antibiotics, AT duration, and combinations. Abundance of Dorea, Agathobacter, Roseburia, and Barnesiella was negatively correlated with AT and corticosteroids use. Patients with increased IL-6, D-dimer, and ferritin levels receiving AT were more likely to show dysbiosis with increased abundance of Enterococcus and Bilophila bacteria and decreased abundance of Roseburia compared with those not receiving AT.

Conclusion: Microbiome diversity is affected by COVID-19 severity. In this context, antibiotic treatment may shift the gut microbiome composition towards OBs, particularly Enterococcus. The impact of treatment-driven dysbiosis on OBs infections and long-term consequences needs further study to define the role of gut homeostasis in COVID-19 recovery and inform targeted interventions.

COVID-19 seroprevalence amongst healthcare workers: potential biases in estimating infection prevalence