Amoxicillin-Clavulanate vs Amoxicillin for Acute Sinusitis in Adults
In adults with uncomplicated acute sinusitis, amoxicillin-clavulanate did not reduce treatment failure but increased later yeast and Clostridioides difficile infections.
*Retrospective matched cohort study; Level 2b (Oxford).
*Retrospective matched cohort study; Level 2b (Oxford).
Citation
Savage TJ, Butler AM, Kronman MP, et al. Amoxicillin-clavulanate vs amoxicillin for acute sinusitis in adults. JAMA. 2026;335(19):1694-1705. doi:10.1001/jama.2025.26902 (published online April 18, 2026).
Background
Acute sinusitis leads to frequent antibiotic use, but it is unclear whether broader coverage improves outcomes for uncomplicated cases. This study compared common standard-dose options in routine outpatient care.
Patients
Adults aged 18-64 years with a new outpatient acute sinusitis diagnosis (2018-2023). Exclusions included chronic or complicated sinus disease, recent antibiotic use, recent hospitalization, and concurrent bacterial infections.
Intervention
Standard-dose amoxicillin-clavulanate (875 mg/125 mg twice daily).
Control
Standard-dose amoxicillin (875 mg twice daily or 500 mg three times daily).
Outcome
(Primary) Treatment failure 1-14 days after start (new antibiotic, or sinusitis-related emergency or hospital care). Also assessed antibiotic-related side effects and secondary infections.
Follow-up Period
14 days for treatment failure; up to 90 days for secondary infections.
Results
After matching, 117,304 patients were in each group. Treatment failure was similar (3.0% vs 3.1%), and emergency or hospital visits were rare (0.03% overall). Antibiotic-related side effects were similar (1.3% vs 1.2%).
| Secondary infection outcome | Amoxicillin-clavulanate | Amoxicillin | Risk ratio (95% CI) | Number needed to harm |
|---|---|---|---|---|
| Any secondary infection | 1.2% | 0.8% | 1.42 (1.31-1.54) | 294 |
| Yeast infection | 1.1% | 0.8% | 1.40 (1.29-1.53) | 313 |
| Clostridioides difficile infection | 0.04% | 0.02% | 2.14 (1.29-3.54) | 5000 |
Number needed to harm is based on absolute risk increase in the matched groups.
Limitations
Because this was not a randomized trial, unmeasured differences between groups could remain. Outcomes were based on insurance claims rather than patient-reported symptoms or lab testing. Results may not generalize to uninsured or publicly insured patients. Although some differences were statistically detectable, absolute risks were small for many patients (especially for Clostridioides difficile infection).
Funding
US National Institutes of Health and US Centers for Disease Control and Prevention.
Clinical Application
When an antibiotic is chosen for uncomplicated adult acute sinusitis, prefer standard-dose amoxicillin to reduce avoidable secondary infections without increasing treatment failure.
Journal Club
Discussion questions and an EBM quiz for this paper. How to run a journal club →
Discussion question (1)
In this propensity score–matched new-user, active-comparator retrospective cohort, amoxicillin-clavulanate vs amoxicillin showed similar 14-day treatment failure (3.0% vs 3.1%; RR 0.96) but higher yeast infection risk (1.1% vs 0.8%; RR 1.40)—how should we weigh residual confounding and these absolute harms when choosing first-line therapy?
Discussion
Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.