A retrospective cohort study on early antibiotic use in vaccinated and unvaccinated COVID-19 patients

Document Type



Biological and Biomedical Sciences


The bacteriophage behavior of SARS-CoV-2 during the acute and post-COVID-19 phases appears to be an important factor in the development of the disease. The early use of antibiotics seems to be crucial to inhibit disease progression-to prevent viral replication in the gut microbiome, and control toxicological production from the human microbiome. To study the impact of specific antibiotics on recovery from COVID-19 and long COVID (LC) taking into account: vaccination status, comorbidities, SARS-CoV-2 wave, time of initiation of antibiotic therapy and concomitant use of corticosteroids and nonsteroidal anti-inflammatory drugs (NSAIDs). A total of 211 COVID-19 patients were included in the study: of which 59 were vaccinated with mRNA vaccines against SARS-CoV-2 while 152 were unvaccinated. Patients were enrolled in three waves: from September 2020 to October 2022, corresponding to the emergence of the pre-Delta, Delta, and Omicron variants of the SARS-CoV-2 virus. The three criteria for enrolling patients were: oropharyngeal swab positivity or fecal findings; moderate symptoms with antibiotic intake; and measurement of blood oxygen saturation during the period of illness. The use of antibiotic combinations, such as amoxicillin with clavulanic acid (875 + 125 mg tablets, every 12 h) plus rifaximin (400 mg tablets every 12 h), as first choice, as suggested from the previous data, or azithromycin (500 mg tablets every 24 h), plus rifaximin as above, allows healthcare professionals to focus on the gut microbiome and its implications in COVID-19 disease during patient care. The primary outcome measured in this study was the estimated average treatment effect, which quantified the difference in mean recovery between patients receiving antibiotics and those not receiving antibiotics at 3 and 9 days after the start of treatment. In the analysis, both vaccinated and unvaccinated groups had a median illness duration of 7 days (interquartile range [IQR] 6-9 days for each; recovery crude hazard ratio [HR] = 0.94, p = 0.700). The median illness duration for the pre-Delta and Delta waves was 8 days (IQR 7-10 days), while it was shorter, 6.5 days, for Omicron (IQR 6-8 days; recovery crude HR = 1.71, p < 0.001). These results were confirmed by multivariate analysis. Patients with comorbidities had a significantly longer disease duration: median 8 days (IQR 7-10 days) compared to 7 days (IQR 6-8 days) for those without comorbidities (crude HR = 0.75, p = 0.038), but this result was not confirmed in multivariate analysis as statistical significance was lost. Early initiation of antibiotic therapy resulted in a significantly shorter recovery time (crude HR = 4.74, p < 0.001). Concomitant use of NSAIDs did not reduce disease duration and in multivariate analysis prolonged the disease (p = 0.041). A subgroup of 42 patients receiving corticosteroids for a median of 3 days (IQR 3-6 days) had a longer recovery time (median 9 days, IQR 8-10 days) compared to others (median 7 days, IQR 6-8 days; crude HR = 0.542, p < 0.001), as confirmed also by the adjusted HR. In this study, a statistically significant reduction in recovery time was observed among patients who received early antibiotic treatment. Early initiation of antibiotics played a crucial role in maintaining higher levels of blood oxygen saturation. In addition, it is worth noting that a significant number of patients who received antibiotics in the first 3 days and for a duration of 7 days, during the acute phase did not develop LC.


Pagination is not provided by author/publisher

Publication (Name of Journal)

Journal of Medical Virology