In the first-line treatment of adults with uncomplicated acute sinusitis, amoxicillin-clavulanate offers no clinical advantage over amoxicillin, according to a large retrospective study.

Analysis of data from a propensity-score matched cohort of 234,608 patients (median age 43 years, 65.5 percent female) showed no significant difference in the primary endpoint of treatment failure within 14 days of treatment initiation, occurring in 3 percent of those treated with amoxicillin-clavulanate vs 3.1 percent of those who received amoxicillin (risk ratio [RR], 0.96, 95 percent confidence interval [CI], 0.92–1.01). [JAMA 2026;doi:10.1001/jama.2025.26902]

The most common reason for treatment failure was dispensation of a different antibiotic without an outpatient encounter, reported in 2.2 percent of patients in the amoxicillin-clavulanate arm vs 2 percent in the amoxicillin arm.

“This [outcome] is presumably something that’s called into a pharmacy and then that patient picks it up and fills it,” said lead investigator Dr Timothy Savage from Brigham and Women’s Hospital in Boston, Massachusetts, US, during his presentation at the ESCMID annual meeting.

The second most common reason for treatment failure was a return outpatient encounter with a new antibiotic dispensation, which occurred in 0.8 percent of patients in the amoxicillin-clavulanate arm vs 1.1 percent in the amoxicillin arm.

“At return visits, patients treated with amoxicillin-clavulanate most often switched to doxycycline, while patients treated with amoxicillin most often switched to amoxicillin-clavulanate,” Savage noted.

Emergency department visits (0.035 percent in the amoxicillin-clavulanate arm vs 0.022 percent in the amoxicillin arm) or inpatient visits (0 percent in both arms) for acute sinusitis and an inpatient encounter for a complication of sinusitis (0.003 percent vs 0.001 percent) were rare.

In terms of safety, the frequency of antibiotic-associated adverse events (eg, gastrointestinal symptoms, skin reactions, and hypersensitivity reactions, among others) was similar between the amoxicillin-clavulanate and the amoxicillin arms (1.3 percent vs 1.2 percent; RR, 1.04, 95 percent CI, 0.97–1.12).

However, the risk of secondary infections appeared higher with amoxicillin-clavulanate vs amoxicillin (1.2 percent vs 0.8 percent; RR, 1.42, 95 percent CI, 1.31–1.54), including yeast infections (1.1 percent vs 0.8 percent) and Clostridioides difficile infections (0.04 percent vs 0.02 percent).

“In the context of the association between broader-spectrum antibiotics and an increased risk of antimicrobial resistance, these findings do not support a clear preference for amoxicillin-clavulanate as first-line treatment [for uncomplicated acute sinusitis], recommended by the Infectious Diseases Society of America guidelines,” Savage said. [Clin Infect Dis 2012;54:e72-e112; Nat Rev Microbiol 2023;21:789-804]

He added that standard-dose amoxicillin may be a preferred empiric choice for such patients for whom an antibiotic is indicated.

When asked whether practice should be changed based on the findings alone, Savage acknowledged that the study captured healthcare use as part of the outcome measures and did not track symptoms on a daily basis. “So, there’s always room for randomized clinical trials and studies like this to complement each other.”

However, Savage also pointed out that the study used nonrandomized data from a huge number of patients across the US and that the findings remained robust across five different sensitivity analyses. “Those things give me a fair bit of confidence… I’m ready to change my practice based on this, but I could see how other folks may want some other symptom-oriented outcomes.”

For the study, Savage and colleagues used data from a nationwide healthcare utilization database and identified 521,244 eligible adult patients with uncomplicated acute sinusitis treated with standard-dose amoxicillin-clavulanate (875 mg-125 mg twice daily) or standard-dose amoxicillin (875 mg twice daily or 500 mg thrice daily).

The propensity-score matched cohort included 117,304 patients in each treatment arm. The median age overall was 43 years, 65.5 percent were female, and 11.8 percent were dispensed oral corticosteroids on the same day of the visit.