Effects of antibiotic treatment on clinical outcomes in patients with COVID-19: a systematic review and meta-analysis

BACKGROUND: The effect of antibiotics on patient-important outcomes in non-severe (NS) and severe or critical (SC) COVID-19 is unclear. We aimed to systematically review and meta-analyse the available evidence on the impact of antibiotic use on clinical outcomes in these patient groups. METHODS: We included randomised controlled trials (RCTs) published between 2019 and 2025 comparing antibiotics plus standard of care (SoC) versus SoC alone in COVID-19 patients, reporting at least one prespecified outcome. Risk of bias was assessed using the Cochrane RoB2.0 tool. Absolute effects were generally estimated by applying pooled relative risk to baseline risks; for low-event outcomes, pooled risk differences (RDs) were used. Mortality in the NS-COVID-19 group was stratified by baseline risk as high (6%), moderate (3%), or low (0·5%). Certainty of evidence was assessed using GRADE. This review was prospectively registered in PROSPERO (CRD42023454933). FINDINGS: Of 54 918 identified studies, 11 RCTs met our eligibility criteria. For NS-COVID-19, antibiotic use did not reduce mortality (RD zero fewer per 1000; high risk 95% CI -27 to 27; moderate risk -19 to 19, both low certainty; low risk -8 to 8, moderate certainty), hospitalisation (two fewer per 1000; -16 to 18, moderate certainty), or invasive mechanical ventilation (six fewer per 1000; -12 to 12, high certainty), and serious adverse events (SAEs) did not differ. For SC-COVID-19, antibiotic use might have reduced mortality (RD six fewer per 1000, 95% CI -21 to 13, low certainty) but did not affect invasive mechanical ventilation (seven fewer per 1000; -19 to 7, moderate certainty) or SAEs (three fewer per 1000; -10 to 6, high certainty). INTERPRETATION: Use of antibiotics is unlikely to improve patient-important outcomes in individuals with NS-COVID-19 or SC-COVID-19. Consistent with these findings, the 2025 WHO guidelines recommend against antibiotics in patients with low suspicion of bacterial infection and emphasise judicious use in COVID-19 management. FUNDING: None.