BACKGROUND: Many trials, reviews, and meta-analyses have been performed on the comparison of short versus long antibiotic treatment in respiratory tract infections, generally supporting shorter treatment. The aim of this umbrella review is to assess the soundness of the current evidence base for optimal antibiotic treatment duration.

METHODS: A search in Ovid MEDLINE, Embase, and Clarivate Analytics Web of Science Core Collection was performed on May 1, 2024, without date and language restrictions. Systematic reviews addressing treatment durations in community-acquired pneumonia (CAP), acute exacerbation of chronic obstructive pulmonary disease (AECOPD), hospital-acquired pneumonia (HAP), acute sinusitis, and streptococcal pharyngitis, tonsillitis, or pharyngotonsillitis were included. Studies from inpatient and outpatient settings were included; reviews in paediatric populations were excluded. Outcomes of interest were clinical and bacteriological cure, microbiological eradication, mortality, relapse rate, and adverse events. The quality of the reviews was assessed using the AMSTAR 2 tool, risk of bias of all included randomised controlled trials (RCTs) using the Cochrane risk-of-bias tool (version 1), and overall quality of evidence according to GRADE.

FINDINGS: We identified 30 systematic reviews meeting the criteria; they were generally of a low to critically low quality. 21 reviews conducted a meta-analysis. For CAP outside the intensive care unit (ICU; 14 reviews, of which eight did a meta-analysis) and AECOPD (eight reviews, of which five did a meta-analysis), there was sufficient evidence supporting a treatment duration of 5 days; evidence for shorter durations is scarce. Evidence on non-ventilator-associated HAP is absent, despite identifying three reviews (of which one did a meta-analysis), since no trials were conducted exclusively in this population. For sinusitis the evidence appears to support a shorter regimen, but more evidence is needed in the population who actually require antibiotic treatment. For pharyngotonsillitis (eight reviews, of which six did a meta-analysis), sufficient evidence exists to support short-course cephalosporin but not short-course penicillin when dosed three times a day.

INTERPRETATION: The available evidence for non-ICU CAP and AECOPD supports a short-course treatment duration of 5 days in patients who have clinically improved. Efforts of the scientific community should be directed at implementing this evidence in daily practice. High-quality RCTs are needed to underpin even shorter treatment durations for CAP and AECOPD, to establish the optimal treatment duration of HAP and acute sinusitis, and to evaluate shorter duration using an optimal penicillin dosing schedule in patients with pharyngotonsillitis.

FUNDING: None.

General Internal Medicine-Primary Care(US) rater

There appears to be ample evidence that shorter courses are adequate. I agree with the authors that efforts now should be directed toward implementing these ideas into widespread clinical practice.

Infectious Disease rater

This is a methodologically clever review of antibiotic treatment duration for both upper and lower respiratory infections outside of the ICU setting. The authors observe that although recommendations from authorities generally favor shorter courses of antibiotics, clinicians generally prescribe longer courses. Much of the evidence is based on reviews that support a shorter course and are of moderate quality. The authors suggest that more reviews are unnecessary, but attention should focus on RCTs comparing shorter courses.

Pediatrics (General) rater

This study discusses an old and important topic for how long antibiotics should be given. It falls short on defining when it is too long and when too short

Respirology/Pulmonology rater

This meta-analysis suggests the possibility that shorter-course antibiotic therapy is non-inferior to longer courses for non-ICU CAP and AECOPD. The authors acknowledge that the quality of most of the studies was not good, so the conclusions should have little impact on current practice.