The website will be down at times for maintenance from 8:00 a.m. CT Saturday, March 2 through 10:00 a.m. CT Sunday, March 3.

brand logo

Am Fam Physician. 2015;91(12):877

Clinical Question

How frequently do harms of amoxicillin (with or without clavulanate) occur?

Bottom Line

The risk of diarrhea is significantly higher with amoxicillin/clavulanate (amoxiclav; Augmentin) than placebo, with a number needed to treat to harm (NNTH) of 8, which is not dissimilar from the number needed to treat for conditions such as otitis media. Contrary to popular belief, the risks of nausea, vomiting, and rash are not increased, but the risk of candidiasis is (NNTH = 23). (Level of Evidence = 1a)


Meta-analyses and systematic reviews are increasingly common in the literature, but most focus on benefit and give little attention to harm. This study is a good example of looking carefully at the harms instead. In this case, the authors searched for randomized, placebo-controlled trials of amoxicillin or amoxiclav for any indication. They did a careful search, had two authors review each article and abstract data, and evaluated study quality. Overall, the risk of bias was low, which is good. They found 45 studies of adults or children, but only 25 studies provided data regarding harms that were usable for the quantitative data synthesis.

The most common conditions being treated with amoxicillin were a respiratory infection or an ear, nose, and throat infection. Studies reported between zero and 10 harms, most commonly gastrointestinal and skin adverse effects. There was no increase in the likelihood of diarrhea overall (10 studies with 4,284 patients), but it was increased with amoxiclav (odds ratio [OR] = 3.3; 95% confidence interval [CI], 2.2 to 4.9). The pooled prevalence of diarrhea was 17.5% with amoxiclav and 5.6% with placebo (NNTH = 8). Only three studies (two of amoxicillin, one of amoxiclav) with 456 patients reported rates of candidiasis, and the likelihood was significantly increased (OR = 7.8; 95% CI, 2.2 to 27). The pooled rates were 4.4% in the amoxicillin groups and 0% in the placebo groups (NNTH = 23). Rates of nausea, vomiting, and rash were similar between the amoxicillin/amoxiclav groups and the placebo groups.

Study design: Meta-analysis (randomized controlled trials)

Funding source: Government

Setting: Various (meta-analysis)

Reference: GilliesMRanakusumaAHoffmannTet alCommon harms from amoxicillin: a systematic review and meta-analysis of randomized placebo-controlled trials for any indication. CMAJ2015; 187( 1): E21– E31.

POEMs (patient-oriented evidence that matters) are provided by Essential Evidence Plus, a point-of-care clinical decision support system published by Wiley-Blackwell. For more information, see Copyright Wiley-Blackwell. Used with permission.

For definitions of levels of evidence used in POEMs, see

To subscribe to a free podcast of these and other POEMs that appear in AFP, search in iTunes for “POEM of the Week” or go to

This series is coordinated by Natasha J. Pyzocha, DO, contributing editor.

A collection of POEMs published in AFP is available at

Continue Reading

More in AFP

More in Pubmed

Copyright © 2015 by the American Academy of Family Physicians.

This content is owned by the AAFP. A person viewing it online may make one printout of the material and may use that printout only for his or her personal, non-commercial reference. This material may not otherwise be downloaded, copied, printed, stored, transmitted or reproduced in any medium, whether now known or later invented, except as authorized in writing by the AAFP.  See permissions for copyright questions and/or permission requests.